Emergency and Backup Power Sources: Preparing for Blackouts and Brownouts

Emergency and Backup Power Sources: Preparing for Blackouts and Brownouts i This page intentionally left blank Em...

Author: Michael F. Hordeski

108 downloads 1561 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Emergency and Backup Power Sources:

Preparing for Blackouts and Brownouts

i

This page intentionally left blank

Emergency and Backup Power Sources:

Preparing for Blackouts and Brownouts by Michael F. Hordeski

iii

Library of Congress Cataloging-in-Publication Data Hordeski, Michael F. Emergency and backup power sources: preparing for blackouts and brownouts/by Michael F. Hordeski. p. cm. Includes bibliographical references and index. ISBN 0-88173-484-5 (print) -- ISBN 0-88173-485-3 (electronic) 1. Emergency power supply. I. Title. TK1020.H67 2005 658.2'6--dc22 2005040624 Emergency and backup power sources: preparing for blackouts and brownouts/by Michael F. Hordeski ©2005 by The Fairmont Press, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Published by The Fairmont Press, Inc. 700 Indian Trail Lilburn, GA 30047 tel: 770-925-9388; fax: 770-381-9865 http://www.fairmontpress.com Distributed by Taylor & Francis Ltd. 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487, USA E-mail: [email protected] Distributed by Taylor & Francis Ltd. 23-25 Blades Court Deodar Road London SW15 2NU, UK E-mail: [email protected]

Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 0-88173-484-5 (The Fairmont Press, Inc.) 0-8493-3908-1 (Taylor & Francis Ltd.) While every effort is made to provide dependable information, the publisher, authors, and editors cannot be held responsible for any errors or omissions. iv

Table of Contents Chapter 1

Emergency Power ..................................................................... 1

Chapter 2

Power Stability and Quality ................................................. 31

Chapter 3

Standby Power Systems ......................................................... 61

Chapter 4

Emergency Generators ........................................................... 97

Chapter 5

Alternate Power Sources ..................................................... 139

Chapter 6

Distributed Generation, Clean Power and Renewable Energy ................................ 171

Chapter 7

Fuel Cells ................................................................................ 209

Chapter 8

Protecting Computer Data .................................................. 247

Chapter 9

Data Recovery ........................................................................ 285

Index ................................................................................................................ 311

v


Preface As the aging power distribution system fails more often to provide reliable power, emergency and backup power sources become critical. In the summer of 2003, 50 million users lost power in eight states and parts of Canada. A handful of commonplace, summertime trips (brief transmission line shutdowns) due to ebbing voltage set off the biggest outage in U.S. history. Extra power through a 345-kV line caused the line to sag into a tree and trip. Utilities in the U.S. and Canada saw wild power swings, several other lines tripped and stations started shutting down after losing power. The outage took down 21 power plants in 3 minutes. Another major outage in the U.S. occurred seven years earlier when high temperatures, sagging power lines and high demand caused a blackout that affected 4 million in nine Western states. The outage ranged from Oregon to San Diego and as far east as Texas and even affected parts of Mexico. Power lines in the Northwest became unbalanced and affected four main arteries that send power south. Major blackouts also occurred in 1997, 1998 and 1999 in the San Francisco and New York areas. Reliable and cost effective systems are needed that will take over during power interruptions and protect critical functions and data. These include standby power systems that employ batteries, kinetic energy storage, fuel cells, reciprocating engines and turbines. This book provides a view of the state of emergency power. An explanation of power disturbances and interruptions is given. Topics include surge suppression, voltage regulation, backup power sources, micro turbines, fuel cells, diesel engines, load management and power quality issues. Reliability and maintainability are critical issues along with comparisons of operating costs and environmental issues. Blackout planning is considered in detail along with emergency procedures and general energy preparedness. In the past the power system has depended on large power plants with long development times. Disruptions in operations at these plants can have major effects on regional supplies of power. In the near future, this crisis can be minimized by investments in new technologies that reduce the need to depend on the power grid. vii

The tools and technologies that are available for an energy user to minimize the effects of power interruptions include fuel cells that use the chemical reaction from a variety of fuels to create power and allow companies to generate clean high-grade electricity on-site without air pollution problems. Some of these units may use fuel processors while others would require hydrogen fuel in a hydrogen economy. Natural gas distributed generation uses gas turbines and gas engines. Modular power units are becoming smaller and more affordable, other modular generation sources include solar-voltaic, micro-hydro and micro-wind. One of the best ways to insure electrical supply reliability and reduce long-term costs is to utilize these smaller, clean, more efficient energygenerating technologies. Cogeneration systems are also available to small scale users of electricity. These modular systems produce electricity and hot water from engine waste heat. Home sized cogeneration packages are capable of providing most of the heating and electrical needs of a home. Cogeneration can produce a given amount of electric power and process heat with 30% less fuel than it takes to produce the electricity and process heat separately. This book stresses the role of contingency planning for reducing the effects of power outages. Frequent power quality problems can be overcome by innovative and practical approaches. This type of power management includes developing alternative sources, monitoring quality and load management solutions. Advances in metering and power monitoring are examined as well as safety and security issues. Cost effective power generation solutions include small modular power generation units. Advanced technologies and products are also available for energy storage options and lighting and cooling integration. Data backup techniques are described for systems and networks. Backup concepts continue or grow with technology and adopt new and innovative approaches to the process. Backup management software can automatically implement backup operations and data maintenance for increased efficiency and reliability. Chapter 1 examines past problems on the power grid. It considers energy usage trends and improving the power future. Basic concepts of emergency power, reliability, energy conversion and efficiency are discussed. Backup techniques and philosophy are explained. Emergency power planning programs, blackout preparation and energy conservation are proposed as solutions. viii

Chapter 2 discusses power stability, grid operation and quality. Virtual utilities are introduced. Grid stability is considered along with the effects of distributed power. Transient conditions and fault removal are explained along with power quality issues such as noise, grounding, ground currents, harmonics and power conditioning. Chapter 3 considers standby power systems and requirements. Backup power systems using chemical batteries, kinetic energy storage and super conducting magnetic energy storage are discussed. Other topics include battery construction and operation, UPS types and operation, maintenance and testing. Chapter 4 is concerned with backup generation. It considers AC generators and alternators, diesel backup and gas turbine generation. One of the newer power sources is the 20 to 60-kW, regenerated, gas turbine power package. This package, in combination with a battery pack, may also deliver low emission power in automobiles. One product is the Capstone 24-kW turbogenerator at a weight of 165 pounds. Chapter 5 describes alternate sources of power backup including solar systems and wind generators. Renewable energy is an important element that improves our energy security and preserves the environment. Renewable energy technologies can provide an important fraction of the nation’s electricity generation requirements and along with other generation sources provide more reliable power. Fuel cells, wind turbines and solar panels can provide power free from dependence on local grids. The search for alternative energy is not new, but the current focus is the goal of making clean and sustainable power a mainstream commodity. Chapter 6 considers distributed power generation including clean power, renewables and combined heat and power generation. Sustainable development is discussed along with environmental issues. Historically, combined heat and power (CHP) systems have been considered only for very large customers or specific facility types such as hospitals and municipal swimming pools. Solar PV systems and fuel cells were not considered at all. This attitude has changed as a new energy climate emerges. The terms used to describe these systems include customersited generation, self-generation, distributed generation, distributed resources, distributed energy, combined heat and power, cogeneration, renewable energy, clean power, and green power. Systems of about 50kW to about 1-MW are large enough to be cost-effective and small enough to be appropriate for many end-users. ix

Fuel cells can be an important source of power in the future as explained in Chapter 7. Topics include fuel cell technology and characteristics. The problems of different types of fuel cells for electric power production are discussed. Fuel cells are compared and their characteristics discussed. Chapter 8 examines the protection of computer data from grid interruptions. In the past, even large enterprises did not always implement backup and recovery procedures on a consistent basis. But, today businesses need to ensure they can recover from a power disaster. Larger enterprises are revising their backup procedures and expanding their emergency infrastructure and many smaller businesses are developing recovery plans. A company cannot afford not to protect its data. Hardware can be replaced but data is difficult or impossible to recover. Chapter 9 discusses the problem of data recovery and offers solutions for minimizing the efforts involved. Even with the use of specialized hardware and fault-tolerant solutions for clustering and replication, some data may be lost. Continued success for an organization that has suffered from a significant system or data loss does not depend just on the ability to replace hardware and rebuild infrastructure. In most cases, continued success depends on the ability to quickly and successfully recover business critical data. Topics include safeguards for data disaster protection, disaster recovery, reducing the risk of data loss, rapid database recovery, clustering and backup appliance storage. Many thanks to Dee, who did much in getting both the text and the tables in their final form.

x

Emergency Power

1

Chapter 1

Emergency Power Problems on the Power Grid In 2003 when Times Square blacked out, some thought it was due to a terrorist attack. In Manhattan, subway trains came to a stop, stranding hundreds of thousands. Toronto went dark along with Rochester, Boston, New York, and other cities. In less than 15 minutes, the computer-controlled power grid of the 80,000-square-mile Canada/United States Eastern Interconnection area went down. Most cities in the Northeast were now without traffic signals, television, airport landing lights, elevators, and refrigeration. As the lights went out, 50 million people in the U.S. and Canada were affected. General Motors shut down 17 of its 60 North American plants. Ford closed 23 plants out of 44. Lost business was estimated at $1 billion. As power returned, investigators focused on overloaded transmission lines around the Lake Erie Loop. Industry and federal officials ruled out a terrorist attack, computer hackers, lightning or the effects of 90° heat as causes of the outage that left almost 50 million people without power. Automated protective devices quickly shut down generating plants and distribution networks across more than a 9,000-square-mile area. About an hour before the main collapse, a section of the system in Ohio experienced problems and took itself off the grid. About 30 minutes later, a second section in Ohio also dropped off the main grid. Events inside the automated and computer-driven power system cascaded too quickly and there was not enough time for operators to react. The event took place in nine seconds according to Michel Gent of the North American Electric Reliability Council, or NERC, a private, standards-setting organization that monitors the transmission system. Even after electrical service had been restored to New York City 1

2

Emergency and Backup Power Sources

and most of the blacked-out areas of the East Coast, the upper Midwest and southern Canada continued to suffer. New York’s subway system slowly resumed service, but airline schedules were impacted and thousands of passengers were stranded. Officials in Detroit and Cleveland urged residents to boil drinking water because of possible contamination. Officials also warned that further rolling blackouts may occur before the system returned to normal in perhaps a week. Reactions to the emergency ranged from pride to finger-pointing. Some officials praised the lack of panic and disorder, along with the effectiveness of emergency-response systems put in place after the September 11, 2001 terrorist attacks on the World Trade Center and the Pentagon. New York Mayor Michael Bloomberg praised the orderly behavior of New Yorkers and the efficiency of firefighters and police officers. Bloomberg said the city’s water supply was safe and adequate, but he warned New Yorkers to stay away from the city’s beaches, which had been contaminated with unprocessed sewage during the outage. In Washington, officials from the departments of Homeland Security, Defense, Treasury and other agencies stated that the federal systems in place allowed them to quickly provide communications, National Guard troops and other resources if needed by local authorities. However, this assistance was mostly unnecessary. New York Governor George Pataki asked the president to declare the state a disaster area. That would make New Yorkers eligible for federal reimbursement for the extraordinary expenses that had been incurred.

CAUSE OF THE OUTAGE Politicians in the United States and Canada rushed to blame one another for failing to deal beforehand with the weaknesses of the power system. Officials in the Canadian prime minister’s office suggested that a fire at the Niagara-Mohawk power plant in upstate New York might have been to blame. Ontario’s provincial premier promoted the idea that the trouble started somewhere in the upper Midwest. In Congress, Republicans and Democrats blamed each other for failure to complete any action on pending energy legislation that contains provisions for improving the power grid. President Bush described

Emergency Power

3

the blackout as a wake-up call for the reform of an antiquated system. The White House announced the formation of a U.S./Canada task force to probe the cause of the outage. The inquiry centered on the Lake Erie loop which is a transmission path for the power that goes along the southern shore of the lake from New York west to Detroit, then up into Canada and back east to the Niagara area. This loop has been known to be a problem for years. There have been plans to make it more reliable but little has been done. Much of the power moving east from the Detroit area to New York would usually move through Canada. Shortly before the power failure, 300 megawatts of power were moving east, but the flow suddenly reversed itself with 500 megawatts going the other way. Such reversals in the flow of power around Lake Erie can cause transmission and generation problems in New York. The power system requires all its parts running at the same rate and the Lake Erie incident can cause transmission shutdowns in New York, which in turn can cause generation problems. The events tend to feed on themselves. Several transmission lines in Ohio went out of operation before the blackout. One system went down an hour before the main crash, and the other a half-hour before. These shut-downs may have triggered the problems around Lake Erie. Depending on the conditions, it could start a chain of events. The Ohio lines are owned by FirstEnergy which is the nation’s fourth-largest utility. The initial problems may have been the result of operator errors or a shortcoming in procedures, exacerbated by the failure of an alarm at FirstEnergy to signal the start of a fast-spreading event. The failed line in Ohio began a cascade that brought down 100 power plants including 22 nuclear plants in the U.S. and Canada as eight states and two Canadian provinces experienced failures.

ROLLING BLACKOUTS IN CALIFORNIA In January of 2001 California shut off the power due to massive power consumption by consumers and winter storms. These conditions forced California into a stage 3 power alert, where random blackouts throughout the state are used to conserve power.

4


Stage 3 blackouts affect businesses and residential areas alike. Only vital services like hospitals, police, fire and air traffic control are exempt from the blackouts. Up to two million residents faced the rolling blackouts, which can last as long as four hours. The San Francisco/San Jose area was the hardest hit. While some blamed the increase of technology companies and their 24-hour computing demands for being the problem, power officials said the problem came from excessive residential use, not businesses. If people had used one-third less power, that would have dropped some 5,000 megawatts off the grid and eliminated the blackouts according to the California Independent System Operator. Cal-ISO manages the California power grid and controls about 75% of the power in the state. The state had been facing power concerns for weeks, but it was made worse by a winter storm that brought rain and snow to the state. The Diablo Canyon nuclear power plant in San Luis Obispo County, CA, was also cut to only 20% efficiency. Much of the problem is due to the large amount of power needed in and around San Francisco and the surrounding area. Cal-ISO cannot get enough power from the southern part of the state to the north. The main route for this power, Path 15, is congested. For about a 100-mile stretch, the grid shrinks, which is similar to going from a four-lane to a two-lane highway. In July of 2002 the mercury rose into triple-digit temperatures in California dropping the state’s energy reserves to the lowest level in a year and sending the state into a one-day stage 2 emergency. Despite rolling blackouts and rising wholesale electricity prices, air conditioners continued to hum during the first intense heat of the year. Peak demand reached 42,441 megawatts, the highest of the year, according to the California Independent System Operator. It was about a year since the last rotating outage and the public had come to view the energy crisis as over. The state has been trying to improve the power supply with new power generation and emergency energy conservation measures. New power plants and improved hydroelectric production have helped, over a period of 18 months, California had a net increase of approximately 4,500 megawatts. Consumers have installed thousands of new, more efficient appliances and millions of energy-efficient light bulbs.

Emergency Power

5

Conservation measures have had some impact, California’s larger power customers used about 500 fewer megawatts during the summer. This is the equivalent of the output from a medium-sized power plant. The Real-Time Metering Program allows large utility customers such as retail stores, hospitals, office buildings and schools to monitor their hourly energy use on the Internet and in real-time to control their energy costs. Some municipal utilities have made voluntary, real-time rates available, where electricity is priced based on the wholesale market. This allows customers to adjust their production schedules according to the current electricity pricing. There are also voluntary demand response pay-back programs that pay to curtail prearranged electric loads when asked by the utility. The Energy Commission estimates that these meters will result in reducing peak electric demand by 600 megawatts per year. The cost for implementing real-time electricity meters is approximately $65 per kilowatt. The cost for typical peaking power plants using combustion gas turbine technology are several thousands of dollars per kilowatt. Building energy management systems (EMS) also play a role in demand response programs used by utilities to keep peak electric power usage low. A building energy management system allows utility customers to identify and program energy-consuming equipment and systems to shed loads as needed. Although the power plants and hydropower have helped increase the supply and the conservation programs have helped stiffen the grid, California’s energy supply remains vulnerable. Typically, the state’s energy load has climbed every year by a few percentage points. Although loads dropped below expectations a few summers, economic conditions and conservation efforts have influenced the demand for energy. Loads have increased and regional heat waves tend to thin reserve margins in the west. California has been unable to import electricity from other states in the same amounts as earlier done. To prevent another power crisis, California is asking utility customers to conserve 3,000 megawatts of electricity in the summer. Programs like real-time metering are expected to provide over 1,200 megawatts and consumers and businesses must provide the remaining 1,800 megawatts. The California ISO has been asking consumers to reduce power use in peak periods, between 3 and 6 p.m. These periods will probably occur for a few hours on the hottest days of the year when air conditioners are

6


running in large areas of the state. Consumers must reduce the lighting load in their home or office, set the thermostat up several degrees and avoid using any major appliances. They should also increase efficiency by adding insulation and changing windows.

NORTHEAST BLACKOUT OF 1965 The power outage of November 1965 hit New York especially hard as the Niagara power grid dropped out during the rush hour on November 9, 1965. At about 5:15 p.m., the power lines that connect Niagara Falls and New York City exceeded their maximum load causing a transmission relay to fail. The failure of that single part started a chain of events that cut power to more than 25 million people in eight states and two provinces. The power that had been heading for New York on that November evening took an alternate route to the power grid that feeds New England. The subsequent overload caused the entire grid to collapse, plunging Boston, Hartford, and most of the rest of the northeast in darkness. In minutes, utilities diverted their power northward, causing a shutdown of the grids in Ontario and Quebec. The situation became critical in New York, where the airports had no power, traffic lights were out and people were trapped in high-rises and in the subway system. As a result of what was called the Great Northeast Blackout, power utilities across the U.S. instituted fail-safe systems to blackout small areas to save larger portions of the grid. Many thought the problem was solved, but they would learn differently 12 years later.

1977 NEW YORK BLACKOUT The 1977 New York blackout took on a different hue compared to the 1965 outage. At 9:34 p.m. on July 13, 1977, in the middle of a midsummer heat wave, power was cut off to New York City, plunging nine million people into darkness. On Broadway, the show went on with candlelight and emergency power. The atmosphere in 1965 was congenial, but in 1977 it was malicious. In a few hours, the city was ablaze, as people in the Bronx and Brooklyn rioted. Police would arrest 4,000 people in connection with the

Emergency Power

7

looting and pillaging of more than 2,000 stores. Firefighters were called to more than 1,000 incidents, many of which were deliberately set. In 1989, Quebec had its own crisis as sunspots caused HydroQuebec’s power grid to switch off at 2:45 a.m. on March 13, 1989, cutting six million people off from electricity. Nine hours later, the power was restored. To most Americans, a vast blackout seems hard to explain. What do power stations in Ottawa have to do with the lights in New York or Cleveland? Many did not realize that over the years of restructuring regional power, companies have merged their generation capacity to become part of the largest infrastructure on the continent. A power sharing network stretches from Florida to Canada and acts much like a single electrical circuit. Electrical grids in Canada and the United States, link up at 37 major points so the two countries can trade power. When one utility has a shortage, it buys power from a neighboring utility. But, this overworked network also has a very old transmission grid of underground and overhead power lines that were last upgraded in the 1950s and 1960s. This system of 50-year-old lines cannot handle the rapid transfers of the new power-trading economy. One official stated that 80% of the generators that had been off during the initial 2003 blackout were soon running again at capacity but the transmission lines could handle only 20% of the output. Little attention has been paid to the grid’s vulnerabilities. Since the big blackout of 1965, the electricity-transmission system was supposed to have been redesigned with safeguards that would not allow such disruptive incidents. For years we have been warned of too little investment in a transmission grid that has been asked to do more and more. One major factor was deregulation. In the 1990s, many utilities were broken up, separating the transmission businesses from the power generators thus product electricity. Today the system is dominated by independent operators in a market-driven system with no link between generation planning and transmission planning. Operators can see few benefits in building power lines for other regions. Citizens’ groups have made local approval of new transmission lines difficult. In the past, power companies had to invest in transmission because it was part of their business model. Now, they may not own any part of transmission in the deregulated model.

8


The Federal Energy Regulatory Commission wants to give all electricity suppliers equal access to power lines. This would provide more power from generators in other regions. But power companies and politicians in the South and West have opposed the plan, stating that it would force prices up in their usually low-cost regions. Homeland Security called the 2003 blackout a test of the system. State and local officials took the necessary steps to be prepared for massive emergencies. In New York, the night of 8/14/03 was a graphic contrast from previous outages. There were no riots like those that crippled the city in the blackout of 1977. Officials recorded 800 elevator rescues, 80,000 calls to 911 and 5,000 emergency medical service calls. Based on initial information supplied by the power company, the event started in Canada. The office of the Canadian Prime Minister then stated that the blackout might have been triggered by a lightning strike on a major transmission line in upstate New York. Then industry officials became convinced that the problem that led to the blackout originated somewhere in northeastern Ohio. Officials were trying to determine why this situation was not brought under control after the first three transmission lines switched out of service according to the North American Electric Reliability Council (NERC). This agency was formed after the Northeastern blackout of 1965, to prevent another major breakdown. The Ohio lines were operated by First Energy, a transmission company based in Akron, Ohio. First Energy confirmed that their facilities in northern Ohio had suffered several mishaps during the afternoon prior to the blackout. These included a tree falling on one of the company’s heavy-duty 345-kilovolt high-tension lines and tripping off a generator at a company plant in Eastlake, Ohio. Another 345-kV line may have been so overloaded that it sagged into a lower-voltage cable below it, shorting out the circuit. But, First Energy said it believed its equipment had coped with these failures, which were not that unusual on a warm summer day.

TIMELINE OF THE 2003 FAILURE On Wednesday, August 13, 2003, at 3:06 p.m. the first of three transmission lines that are believed to have triggered the blackout trips off. The outage puts pressure on another line and the effect spreads.

Emergency Power

9

Cleveland, Ohio Power’s voltage drops to zero. An hour later, utilities in Canada and the Northeast experience major power swings. The Bruce Nuclear Station in Ontario shuts down and blackouts hit Toronto and southern Ontario, where most of the province’s 10 million residents live. A few minutes later the Campbell No. 3 coal-fired power plant near Grand Haven, MI, trips off, and then the Enrico Fermi nuclear plant near Detroit shuts down automatically after losing power. A number of transmission lines trip at this time including a 345-kilovolt line known as the Hampton-Thetford in upstate New York and Vermont. Much of New England is spared, when the region’s power operator disconnects its system from New York’s after realizing something is wrong. A transmission line between Pennsylvania and Toledo, Ohio, trips, and in a period of 15 minutes five nuclear power plants in New York state shut down. Parts of nine states including all of New York City are now affected.

POWER RESTORATION Restoring power to a massive area requires utilities to balance electricity coming from the restarted plants with load demands. An imbalance can trigger more blackouts. Supplementary power sources called black starters are used to re-engage the generators and get auxiliary systems on-line. Once the generators are up, they could flood the grid with too much power and shut it down again if there are not enough substations on-line to draw power. At the substations, operators must control the power distribution and gradually send more power to areas that need it. As new power plants are connected to the grid, those that are already up and running must drop back their output to stabilize the system. Essential facilities and services are the first to get their power back. These include hospitals, police and fire departments, water and sewagetreatment plants. As areas are brought up, they are connected with other nearby regions. This merging can cause destabilizing fluctuations for a while. By 10 p.m. on the day of the blackout, 50% of affected areas in New England had their power restored. By 5 a.m. the next day, 50% of Canadian areas were back online. New York was fully restored by 10:30 p.m.

10


FIXING THE GRID Economic growth and the proliferation of computers and other digital devices have strained the power arteries, while utilities and state governments debate over who should repair the problem. Deregulation allows local utilities to sell electricity where they can find a buyer. But the grids are still administered on a state-by-state basis. This is because states do not want to give up control, which stops new transmission lines from being built. State and federal commissions need to meet to consider upgrades of the grid across state lines. The Federal Energy Regulatory Commission (FERC) should order the construction of new lines, like the proposed Arrowhead-Weston line between Wisconsin and Minnesota. The highway system is planned on a federal level, the federal government should also direct expansion of the power grid. As the U.S. economic output doubled between 1975 and today, investment in the grid dropped from $5 billion to about $2 billion annually, according to the Edison Electric Institute. In a deregulated world, utilities use each other’s networks so no one wants to pay for an improvement that would also benefit competitors. Funding upgrades of the system with federal dollars is one solution. Financial incentives for power generators to build transmission lines for new plants is another. New plants could be paid by rate increases over several years. Like the infrastructure itself, the failure of support for long-range planning transcends national borders. As the global economy becomes increasingly dependent on the digital networks made possible by electricity, public funding worldwide for newer cleaner power sources and improving our infrastructure is decreasing. The U.S. spent one-third less on energy R&D in 1995 than it did in 1985. Germany, Italy, and the UK spent two-thirds less.

IMPROVING THE GRID The modern power grid is stretched over a vulnerable technological infrastructure. The Electric Power Research Institute (EPRI) was founded after the failure of the grid in 1965. EPRI believes we still have not fully heard the message of that massive blackout. One lesson of the current crisis, many believe, is that we need smarter methods of electric-

Emergency Power

11

ity generation, transmission and delivery not just more power and more lines. EPRI is the utilities think tank, an independent research organization funded by more than 1,000 power companies. EPRI was the first industry-wide R&D consortium in America. It is one of the largest consortiums in the world and represents utilities in 40 countries. EPRI’s members range from the older giants like Consolidated Edison of New York to the newer upstarts like Mirant and Dynegy. These members generate 90% of the electricity used in the United States. The Bush administration’s declarations about improving the power networks is familiar at EPRI. The institute has been laying the groundwork for this technology for decades. EPRI’s older members stand to gain more from an energy policy that favors the more traditional means of increasing power such as more fossil-fuel plants, more oil and reviving the nuclear power industry. Debate over the merits of solutions such as drilling in national wilderness areas, is a distraction from our ability to implement a practical blueprint for a new conception of the energy grid.

BLACKOUT PREPARATION The blackout that plunged one of every six Americans into darkness in 2003 triggered a national discussion of energy policy and preparedness. In Southern California, where earthquakes are an ongoing threat, concern over the flow of electricity and an awareness of emergency procedures is not new. Many are asking themselves about backup plans if the power goes out. At the Metropolitan Transportation Authority in Los Angeles, they believe that although the rail systems may not operate, the stations and all of MTA’s buildings including corporate headquarters and rail and bus facilities will still have emergency power systems for mission-critical operations. Buses would not run normal routes, but they would continue to run to the extent possible. In the rail stations, emergency power systems would generate power for lighting to allow station evacuation. Some food stores such as Gelson’s in Marina del Rey, California, plan to close retail operations and cover refrigerated cases. If the loss of

12


refrigeration is more than two hours, they must order dry ice to keep refrigerated items cold. Some facilities such as hotels use a mixture of procedures. Hilton Hotels has emergency procedures in place for blackouts, fires or earthquakes. In case of a power outage, their hotels keep a supply of flashlights, glow sticks, bottled water and AM-FM radios to distribute to every room. There are back-up lights that come on in the stairwells, hallways, lobby areas and public areas. Key locks are battery operated so guests can get in and out of their rooms. The Westin Hotel in Long Beach has a 1,500-kilowatt generator that supplies the entire hotel in case of emergencies. They do not have to worry about seeking out any other power. During the power outage on the East Coast, they ran a scheduled test and everything worked perfectly. Plastic manufacturers such as Plastic Services and Products PMC Global Inc. use plastic extrusion molds which use a lot of electricity. They run around the clock, so their electricity costs are large and they would need a large amount of backup capability. Without backup generators, if the grid goes down they would have to shut down. Large office buildings of over one million square feet, should have well thought-out emergency plans. But, even here if there were a massive power outage, the emergency power systems might power 20 to 25% of each building. Then, they could run at least one or two elevators in each building and have minimal lighting, enough for people to get in and out. At Universal Studios Hollywood back-up generators would provide power to handle lighting and safe exit from all rides and shows. Emergency power systems would be able to handle audio communications for the guests. In Silicon Valley, technology companies like Oracle are constructing their own local energy network using substations, diesel generators, and power-conditioning systems. In many technology installations a supply of fluctuation-free electricity is critical. Chip fabrication plants and server farms must balance the expense of building independent electricity resources with the cost of equipment failures and network crashes caused by unreliable power. Hewlett-Packard has estimated that a 15-minute outage at a chip fabrication plant cost the company $30 million, about half the plant’s power budget for a year. Surveys of technology companies show that while nearly all are

Emergency Power

13

trying to reduce electricity consumption, preparedness for potential blackouts is spotty. One survey by the Washington Software Association (WSA), showed that of 48 members responding, only four had developed backup power sources, three used generators and the other solar. There is concern about rolling blackouts and brownouts and companies should conserve, to protect themselves from power failures, and to act politically to guarantee future energy supplies. In the present climate now, companies are recovering from stagnant economic conditions any disruptions in their ability to generate products and revenue will impact their recovery. Another survey of members of a technology industry association indicated about 70% increased their blackout readiness. Some of this can be attributed to preparations for Y2K. Many manufacturers installed their own generators as a part of Y2K preparation. In spite of a lessened sense of crisis, the power system has little extra capacity, so companies are still vulnerable. We have seen a growing crisis in California and the rest of the country. Companies need to be prepared and awareness is increasing in the face of the massive eastern blackouts and the rolling blackouts in California. Increased efforts are needed to reduce energy use including educating employees on how they can save. Companies are looking at modifying workloads and shifting energy-intensive operations to off-peak times. Companies should institute long-term reductions in energy use to ease demands on the system. Among the organizations benefiting from Y2K planning is the Fred Hutchinson Cancer Research Center in Washington state, which has several contingency plans to cope with an emergency power shortage. These strategies include prioritized plan for load-shedding, dual feeds from Seattle City Light and diesel generators for emergency power. The facility has also been reducing consumption. It has been a test site for a new energy-saving system using variable volume and air pressure in its air circulation system. This system and 23 other measures implemented in the last two years, has helped the facility cut energy use by 6% even though its activities have increased. Immunex Corporation also planned for Y2K in 1999, and updated those plans to meet emergency power needs. Immunex put together a

14


task force to plan strategies to cope with rolling blackouts. It installed generators to keep most vital systems functioning, but a rolling blackout would still be a business interruption since not all departments are on generator power. Birmingham Steel is a large mini-mill in the West Seattle area. It has been implementing new energy-saving technologies in its steel-making processes and has cut its energy by almost 12%. The company has week-long shutdowns during the summer in order to reduce the load on Seattle City Light. They are also working with Seattle City Light in a joint venture to use the plant’s waste heat, a by-product of producing steel, to operate steam-driven generators. Some of the companies well prepared for power failures are Internet server farms, since they must provide stable, uninterrupted services. For example, Zama Networks has a double power backup for its 30,000-square-foot server farm near Boeing Field. The entire facility runs on several hundred rechargeable batteries. The batteries act as a buffer and keep the power cleaner during normal operation and can keep the plant running for short blackouts. Backup for the batteries is a two-megawatt diesel generator with fuel for three days. They also have processes in place to respond to a long-duration power outage and expect to have a power event during the hotter months. Zama has been reducing its power consumption and now uses 5,600 kilowatts a day, compared to the 7,200 kilowatts a day it used earlier.

BACKUP POWER The massive power failure that darkened much of the eastern United States and Ontario in the summer of 2003 did not reach Indianapolis. If it had, only a small group of local companies and facilities would have been capable of operating through the blackout. In spite of that, most Indianapolis companies, industrial parks and other facilities have not rushed to install emergency power supply systems. Facilities managers indicate that a majority of businesses are holding back on purchasing backup systems that would allow them to function in the event of a major failure of the electrical grid. In Indiana, the electrical infrastructure has a good record for reliability compared to other areas. The cost of backup power may seem

Emergency Power

15

high when you consider the overall cost of investment and the price of maintaining it. Although it may be hard to create an EVA (economic value addition) that would justify it, an extended downtime could be crippling to many facilities. Large users may be reluctant to build a stand-alone power plant solely for emergency use, but almost all major office buildings have some emergency power backup systems. Normally, these provide enough power for emergency functions such as basic life safety. This kind of system is not very costly. Moving up to a system large enough to keep a major office park going is not inexpensive. Many large companies become interested in a large-scale backup system, because of concerns about the reliability of the power grid in their specific area. For example, a building site might be served by an electrical substation that has a history of reliability problems. It may be vulnerable to storm damage or have maintenance problems. In this case, a company should look at alternate sources of power. Although a company may not wish to build its own generating plant, it can arrange its electrical system so it can be connected to a different substation in case the regular system goes off. In Indianapolis, industrial parks like the Intech Park, a major business park on the northwest side use this technique. The power feed serving the park is looped so that it can be switched to another substation if needed. This system is used because the high-tech companies at the park do not want a lengthy power outage that could hurt them. Some major manufacturers have a critical need of backup power capacity. Pharmaceutical giant Lilly produces large quantities of medicines and vaccines that can start deteriorating after a few hours on a stopped conveyer line. Lilly has a high capacity backup system because of the nature of its products. If a widespread power outage were to occur in Indianapolis, Lilly would be unable to maintain its critical operations. Roche Diagnostics Corp. in Indianapolis also has a backup system in place to allow critical functions to continue in the event of a major power failure. There are other facilities that must have backup power systems like hospitals. These elaborate backup systems are in place and seemed to operate well enough during the blackout. In spite of the problems caused by the 2003 record blackout, none of the hospitals and other emergency facilities in the affected areas reported any trouble in operating.

16


MANDATED BACKUP SYSTEMS Hospitals and other vital emergency centers are required by law to have backup systems. This is usually with a double and in some cases, triple level protection. St. Vincent in Indianapolis is typical of what hospitals use. The hospital complex has three separate emergency power systems, each with its own diesel generator. Each of the systems is connected to the hospital’s electrical system. If the external power fails, one of the generators automatically goes online and provides power for the hospital’s critical operations. The switchover must take place in 10 seconds. If the first generator does not function, the second will be started up. The third unit is available if something happens to the second. The generators run on diesel fuel, and the hospital has about a 6hour supply. If necessary, the hospital can run on its emergency generators for as long as it can obtain diesel fuel. The system is not designed to make up for all lost power in case of a failure. It supplies enough electricity to run operating rooms, intensive care units and neonatal units at full power. Nursing stations and staffing areas get full power, as well as hospital entrances, stairs and elevators. Lighting is cut back with only every third hall light getting power. The emphasis is keeping the most critical elements of the hospital operating. A blackout should not affect the quality of patient care. Hospitals do not use their backup systems to replace full power because it is too costly. The redundant systems in use are expensive enough. St. Vincent Hospital recently replaced one of its three generators, at a cost of $500,000 for the generator and its control system. St. Vincent tests the system every month and inspects it every week. The system has functioned well through several outages caused by ice storms and other inclement weather. Other critical institutions that have backup systems include police and fire stations and many call centers and data centers.

PLANNING FOR EMERGENCY POWER Planning cannot prepare for every calamity, but it can provide emergency personnel with valuable information such as what essential

Emergency Power

17

facilities and services will need power, the amount of power and how to provide it. Expecting the unpredictable is part of the emergency response plan. Instead of trying to plan for a specific event, such as a wind storm, fire or flood, concentrate on the common results of any disaster. Vital among these is the loss of electric power. Electricity is scarce after a disaster. Lights are out, telephones may be disabled and many businesses will be shut down. Essential services are often impacted. People may require food, water, heat and emergency medical attention. True recovery requires the power to be on, but it may be difficult to predict when utility service will be completely restored. A backup power system plays a critical role in recovery from all types of disasters. Permanent backup systems are used by facilities that must maintain some continued level of public health, safety and welfare, even in extended blackouts. Mobile generators are available in all sizes for powering schools, stores, offices, factories and homes while work goes on and the grid is restored. The speed of recovery depends on how well the enterprise has planned for emergency power resources. The provision for electricity is critical in a disaster management plan. Procedures should be clear and well thought out. Global supplies of mobile generator sets have almost quadrupled in the last decade. This allows a facility that has planned ahead to readily secure almost any amount of short-term emergency power from units for small offices or homes to 2-megawatt power modules to supply large buildings. An extended power failure can have many causes; some will be natural and others are not. Some may be more predictable than others. For example, it is difficult to imagine that many could have foreseen the April 1992 flood that shut down power for several weeks in the heart of Chicago. The flood was triggered when construction workers were installing support pillars in the Chicago River bottom. The pillars punctured the roof of a freight tunnel under the city. Water then flowed through the complete system of tunnels and into the city’s building basements that contained the building’s main power distribution systems. Another extended power outage occurred in central Auckland, New Zealand, in February 1998. The four main power feed cables for the city failed because of an overload. The blackout affected 50,000 city

18


workers and 6,000 residents. Butter, meat and other perishable exports in thousands of refrigerated containers were at risk. They were waiting to be shipped from the city’s port at the height of the export season. In September 1998, weather forecasters did predict the landfall of Hurricane Georges on certain Caribbean islands. But, the force of the hurricane was underestimated. Puerto Rico and several nearby islands lost all power. Mobile diesel-powered generator sets were important in the recovery from all these events. The demand for generators exceeded local supplies, and units were sent from significant distances. New Zealand and the Caribbean Islands received units by airlift. The logistics of supplying such power equipment in a hurry are difficult, but effective planning makes the task go smoother and recovery is faster.

PERMANENT BACKUP It is vital in emergency power planning to provide essential facilities with permanent backup power. The backup equipment must be properly sized and in good repair. Essential post-disaster services include medical care, drinking water supplies, police and fire protection, refrigeration, communications, wastewater treatment, transportation (including highways, rail, airports and seaports), weather forecasting, temporary relief shelters and emergency response command and control. Backup systems should be sized to carry critical loads such as the power to deliver the facility’s necessary public services. Some facilities, including wastewater treatment plants and hospitals are important enough that backup systems must be sized to at least some level of reduced operation. Backup systems should be included in a planned maintenance program that includes regular inspection and operational testing.

MOBILE POWER A major storm or flood may cause damage to permanent backup power systems. In the 1992 flood in Chicago many backup systems were installed in office building basements and sub-basements that filled with water.

Emergency Power

19

The true test of planning is how well it functions in actual practice. We live in a world that increasingly depends on electricity. A facility should give some priority to mobile power. The sooner power is available, the more efficiently materials and services can be delivered. Mobile power equipment should be sized in the same manner as permanent backup power. Power outage can create major logistical problems as public agencies and private businesses rush to provide temporary power. An outage affecting a major city, such as Chicago, may require thousands of mobile generators. The challenges are even greater if the power outage is caused by a natural disaster. Then, the delivery of power may be affected by transportation breakdowns and the distribution of fuel and other supplies. Space must be available for parking the generators outside the buildings. A facility with a large power requirement may need one or more of the large power modules that are 8 feet wide and 40 feet long. Extension cable are needed to connect the generators to the building’s electrical system. Transformers, distribution and panels, feeder panels, and other accessories may also be needed. During an emergency, diesel fuel supplies and delivery may be difficult to obtain. An on-site fuel tank should have the capacity for at least 24 hours of run time. If on-staff personnel are not experienced with power-generation equipment, it is necessary to arrange for professional assistance to install and operate the mobile units. Many suppliers offer permanent backup systems for sale or lease, as well as mobile power units for rent. The supplier should be able to deliver the power generating sets with all required equipment including cables and transformers. Suppliers should be able to offer training for equipment operation or provide operators along with service and maintenance. When renting power units for emergencies, it may be difficult to obtain a contact that guarantees equipment availability. However, many suppliers offer contracts with a right of first acceptance. Under this arrangement, you pay the supplier a reservation or retainer fee for an allocation of certain equipment. The supplier then agrees to reserve the equipment and not release it to another party without your consent. In Puerto Rico after Hurricane Georges, relief efforts were stalled by trees and power lines blocking roads and preventing transportation. The storm also blew down one of the large cranes in the port at San

20


Juan. This created delays in off-loading emergency generators arriving by ships. The mechanics of power delivery are important, especially when equipment is not available locally. Provisions may be needed for staging areas for generators at airports and seaports. Slowdowns in customs can delay the delivery of power. Special legislation can allow generators to be imported in emergencies. Provisions allowing temporary, duty-free imports of equipment can expedite delivery. Contacts established with freight companies during planning can increase the availability of ships or air transports when a disaster occurs. Finances are another issue. As a part of the planning, there should be agreement on payment terms with mobile power suppliers. This may be a letter of credit from a financial institution or budgeting the necessary funds. TESTING THE PLAN An emergency plan is a living document and should be revisited and updated periodically. The plan should also be tested through simulation drills. In a typical drill, participants are presented with a specific scenario and asked to respond to it according to the procedures outlined in the plan. It is useful to involve the local electric utility in drills. During an actual emergency, coordination between utility staff and emergency personnel may improve the utilization of mobile equipment. If emergency personnel know when utility power is about to be restored in a given sector, they can plan to release mobile power units to other areas where they are needed. Disasters are by definition unpredictable and even the best plan will not eliminate the need for good judgment and resourcefulness. However, a plan immediately moves disaster recovery several steps forward. It makes critical actions nearly automatic and provides a basis for sound decision making as the event unfolds. NETWORK OF THE FUTURE The development of backup systems of distributed energy resources is a part the network of the future that will be required to meet

Emergency Power

21

the needs of the digital economy. This marriage of the network that started in the late 19th century with the more recent innovations of the late 20th will produce a future energy providing system that has been called an intelligent grid, energy net and Energy Web. Parts of this network are already appearing in many sectors of the power industry. They have their own momentum and are colliding with regulatory and market barriers. One project, called InfoWatt, would replace the core wiring used in power lines from steel to fiber-optics. This would allow more power on existing lines. The new wiring would allow additional power plants to be brought online along with the distribution of this power. The University of Southern California’s department of material science was involved in the mechanical testing of the, design. InfoWatt also would provide additional computer bandwidth to transmit data between Internet backbones. Power lines are made of aluminum on the outside to carry the current, which tends to concentrate on the outer surface. A steel core is used to support the wires. Aluminum is not strong enough to support large cables, so the steel core is used. However, steel has some drawbacks. It sags in the heat, which can trigger an outage. There is also some loss of transmission capability, because the amount of steel needed is thick. This means less aluminum in the cable and less power capacity. If the lines are thicker, they become subject to environmental problems due to wind and ice. One solution is to change the core. Southern California Edison, the University of Southern California and several government agencies have been working on a fiber-optic core that is lighter than steel, thinner and can transmit data. In the past 20 years, the power industry has increased production by 30% but they have only increased the distribution capacity by 15%. There was excess capacity on the grid when power was generated and sold regionally. Deregulation and shipping power over long distances are stressing the capacity of the grid to carry the needed power. Power transfer has been a major problem in California. Northern California has been subject to blackouts. Southern California may have power to provide to the North, but a problem with the grid, called Path 15, does not have enough throughput. Building more power lines is expensive, an alternative is higher capacity cables on the existing towers. Replacing steel-core wires with fiber allows more aluminum to be

22


used in the same thickness of cable since the fiber core is thinner than steel. This results in 15% more capacity during normal loads and up to 200% higher capacity during peak loads. The smarter energy network of the future should utilize a diverse group of resources located closer to the consumer, providing low- or zero-emissions power in backyards, driveways, downscaled local power stations, and even in automobiles, while giving electricity users the option to become energy vendors. The front end of this new system will be managed by third-party virtual utilities, which will bundle electricity, gas, Internet access, broadband entertainment, and other energy services. This is similar to Edison’s original vision of the industry, which was a network of technologies and services to provide lighting. Digital networks will be used to remake the grid. Embedding sensors, solid-state controllers and intelligence in this new supply chain, the grid should be more robust and adaptive with more photovoltaic arrays and wind turbines. The new grid will have to extract the maximum value from limited resources.

DISTRIBUTED GENERATION Distributed generation was Edison’s first plan for universal electrification. Today, this energy mix includes photovoltaic arrays, gas turbines and variable-speed wind turbines that help to make the price of wind power competitive with fossil fuels. The development of renewable resources and increasing energy efficiency are essential to securing a sustainable future. A long-range prescription is found in EPRI’s “Electricity Technology Roadmap” developed when 150 organizations brainstorms a set of goals for the next 50 years of energy R&D. It brought together representatives from the Department of Energy, the Natural Resources Defense Council, Rand, MIT, the New York Power Authority, General Electric, AT&T, Motorola, the Nature Conservancy, Exxon, the World Bank, Royal Dutch/Shell, Oracle, Microsoft and others. Meeting the energy needs of the next century, the Roadmap suggests a substantial overhaul in how we think about electricity. The industry’s most basic assumptions will have to be put on the table, including the hub-and-spoke hierarchy of the existing grid based on large central power stations with long distance transmission lines radi-

Emergency Power

23

ating outward which has been the backbone of the business since in the 1920s. The current power infrastructure is incompatible with the future. The EPRI’s Roadmap indicates that an energy revolution is under way. In 1999, the Swiss engineering giant ABB announced that it was offloading the building of nuclear plants to concentrate on renewables and distributed generation. Smaller-scale methods for producing electricity closer to the consumer are not a new idea. It was Edison’s first plan for universal electrification, where neighborhood steam plants would provide power and heat for 1-mile-square lighting districts.

MICROPOWER Distributed generation is also called micropower. Renewables such as photovoltaic arrays and wind turbines are micropower resources along with reciprocating engines, fuel cells, Stirling engines and gasfired microturbines. Micropower is surging on world markets, both in industrialized countries and in regions with no electricity. Here distributed generation offers these rural communities access to power without costly grid extensions by the major utilities. Wind power has been the fastest-growing energy source, moving at an average of almost 25% per year. Freestanding windmills and wind farms are going up all over, especially in Europe. Denmark gets almost 15% of its energy supply from renewable resources, and about half of the wind turbines in the world are made by Danish companies, such as Vestas Wind Systems and Bonus Energy. These units have been going to Germany, Spain, and the UK. One wind farm constructed in Texas, using Danish turbines, can provide enough electricity for almost 140,000 homes, while avoiding 20 million tons of carbon dioxide emissions from conventional power plants. EPRI has been involved in designing turbines that provides a steady flow of power under varying wind speeds. In 1989, the institute started a 5-year program to upgrade the technology. EPRI and the Department of Energy promoted the capability of wind power, while utilities and federal and state agencies mapped out promising areas for high-turbine sites. By 1995, variable-speed wind turbines designed by EPRI were generating 3 billion kilowatt-hours per year.

24


PHOTOVOLTAICS These systems which make power from sunlight are growing internationally. In 2003, the largest solar energy project in the world was started in the Philippines. It involved the Spanish government, the Philippines Department of Agrarian Reform and BP Solar. The solar wing of British Petroleum provides more than 10% of the photovoltaic cells used in the world. The $48 million project on Mindanao Island will bring electricity to 400,000 residents of 150 villages on an island that is home to one-third of the nation’s rural poor. The project will produce enough electricity for new irrigation and drinking water distribution systems, as well as lights and medical equipment for schools and health clinics. Seventy-nine power systems will be built. The Mindanao project illustrates the potential for micropower to raise the quality of life in developing countries without building large power plants or relying on expensive fossil fuels. Micropower is being demonstrated over the globe. Just as developing countries are jumping straight into mobile phone service without laying expensive land lines, micropower technologies are enabling those historically left in the dark to leapfrog hub-and-spoke grids.

POWER SENSORS The Electric Power Research Institute envisions digital sensors along the grid and in homes, sending load information back to planning centers. Software could then analyze usage patterns and reroute power to anticipate outages. The blackouts indicate that the current grid technology lags behind our air-traffic-control network. When airlines have a storm and are shut down in one part of the country, they bypass this area. A real-time monitoring system would inform customers of real fluctuations in electricity prices. If prices are rising on a hot afternoon, facilities might choose to save money by turning off equipment. This would reduce the grid load and bring prices down, providing a selfcorrecting action that helps to satisfy the demand. Conservation is also important. During the California blackouts of 2001, parts of the state cut energy consumption by 20%. Consumers can reduce power during peak hours.

Emergency Power

25

POWER RELIABILITY Electricity is produced at most power plants, by large spinning generators driven by moving water, reciprocating gas or diesel engines, or gas or steam turbines. The steam may be created by burning coal, oil, or natural gas or by a nuclear reactor. At a transmission substation, large transformers increase the voltage from thousands to hundreds of thousands of volts so the power can be transmitted long distances with minimal losses. The electricity travels along these high-voltage lines to a power substation, where the power can be redirected to other high-power lines or stepped down to a lower voltage that is sent to area power lines. In the power grids, there is a tender balance between supply and demand where sudden fluctuations can cause portions to fail. If a transmission line breaks, the system is designed to isolate the problem and disconnect it from the grid. If the control mechanisms in place, the computers, circuit breakers and switches fail to contain the problem quickly enough, there can be rapid fluctuations at substations elsewhere in the grid, triggering more shut-down mechanisms. The problem can spread back to generating plants that may be producing too much or too little electricity, causing more shutdowns. Eventually, the problem is contained or it spreads causing a blackout. The 2003 blackout affected 50-million people in the U.S. and Canada located in eight states and two Canadian provinces. The power failures were attributed to three deaths. The blackout caused 22 U.S. and Canadian nuclear plants to shut down along with 10 major airports and 700 flights were canceled nationwide. In Cleveland 7600 gallons (29,000 liters) of drinking water were distributed by the National Guard in Cleveland after the city’s four main pumping stations went down. In New York 350,000 people were stranded on the subway systems when the power went out. Nineteen trains were stuck in underwater tunnels. The reliability of the system depends on procedures that were meant to isolate problems and keep them from spreading. Since supply and demand needs allow the system to work, mechanisms on the grid monitor the flow of power. If there is a sudden failure, like a lightning strike on a transmission tower, circuit breakers open, and the sector releases itself from the grid. The process is called islanding and the goal is to contain the fault by sealing it off from the rest of the network. When this happens, it creates a hold and the network is pro-

26


grammed to either pick up power from other sources or shed load, purposely shutting off power in one part of the grid to protect the rest of it. This is why suburbs are more likely than urban area to suffer brownouts. The system is programmed to protect essential services like hospitals in the most densely populated areas. Getting a power system up and running after a blackout is called a black start. Power-generating units must be brought online slowly. Cities are divided into electrically isolated areas and brought back one by one in parts that the system can handle. As they build up the system, they can start to stringing them together, but if they move too fast, the whole system may go down again. Nuclear power plants usually take at least 24 hours to restart. Any load shift has to happen fast, and no one likes to drop load since it will have to be reported to the regulatory commission. It is also difficult to start up again. In the 2003 blackout, there may not have been time to make a decision. In past blackouts, there was a window of as much as 45 minutes to attempt a deviation. This time events moved across hundreds of miles in seconds. However, Vermont unplugged itself from power feeds from New York and the state was spared from the blackout. In Chattanooga, engineers at the Tennessee Valley Authority operations center saw the transmission flows spike and they got their generators to slow down and stabilize the flow of electricity in their area. A system of circuit breakers in Ohio also stopped the spread of the blackout south. Large amounts of money and time have been spent on the problem of spreading breakdowns since the blackout of November 1965, when an overloaded relay switch in Toronto left 30 million without power through New England and New York City. That event triggered the creation of the NERC., an industry group that sets standards for the transmission system. The NERC set up the system for isolating plants so that if one failed, it would not affect the others. Even with the new safeguards in place, there were still problems in July 1977, when lightning storms in northern Westchester County, N.Y., shut down two transmission lines. Within an hour New York City sealed itself off from the larger grid, but since the city does not generate enough power internally to sustain itself, nine million people lost power, some for more than a day. In July 1996, about two million customers from Nebraska to Washington State to Baja California Norte in Mexico lost power when a 345

Emergency Power

27

kilovolt line was shorted by a tree in Idaho. A mechanical problem shut down a parallel line, setting off a wide collapse. A combination of market forces, politics and a lack of welcome for high-voltage lines has hindered a transmission system that can keep up with demand. There is little incentive for utilities to construct new lines, especially after new federal rules in the late 1980s effectively capped the return on such investment at around 11%. By 1999 transmission investment was less than half the $5 billion it had been 20 years earlier. It has been known for over a decade, that the utility industry was not investing enough in the reliability of the grid. Congress has neglected the operation of the system.

BUSINESS LOSS Many small commercial property insurance policies typically exclude coverage for damage from power outages. This is not true with some large commercial policies because many large companies buy special endorsement to cover service interruptions. However, policies are triggered by different factors, including the duration of the outage, and its cause. Most types of outages are not covered, but exceptions sometimes include damage caused by effects of an outage, such as a fire, or terrorism, if that type of coverage has been purchased. In the August 2003 blackout, small businesses suffered outage-related losses from $100 to $125,000 according to a Detroit Regional Chamber survey from 150 members in the retail and food-service industries. The Associated Food Dealers of Michigan estimates that grocers lost more than $50 million in perishable food in the outage. Oakland County in Michigan estimates the loss at more than $90 million with $82 million in lost wages, $5 million in lost business and about $4 million in damages to utilities. The power outage caused numerous lawsuits for lost inventories, productivity and even information stored in computers. The Michigan Lawyers Weekly stated that not only power companies but others that allegedly failed to provide proper emergency backup systems could be targets. If companies failed to provide appropriately and demonstrated an inability to provide proper backup systems for critical functions like computer data, they could be subjects for lawsuits. The New York City firm Cauley, Geller, Bowman & Rudman L.L.P.

28


filed a class-action lawsuit on behalf of all those who lost power in the Court of Common Pleas in Cuyahoga, Ohio. The complaint against FirstEnergy Corp. charges that the company recklessly caused the power outage, failed to have a functioning alarm that could have warned of problems, failed to cut back problem tree limbs and failed to maintain a fail-safe system that could have separated the local system from the rest of the power grid. The complaint seeks damages for injuries as well as punitive damages.

BLACKOUT LESSONS Hospitals hit by the massive August 2003 blackout learned some important lessons that can help prepare for similar disasters. One was to acquire multiple power generators. Another was to test backup procedures. Like many other health care organizations, William Beaumont Hospitals in Royal Oak, Michigan, and Memorial Sloan-Kettering Cancer Center in New York have diesel generators for running information systems during a power outage. These generators automatically kicked in during the 2003 outage that hit parts of the United States and Canada. But, these hospitals encountered unexpected problems. At Beaumont, information technology personnel had been fighting the MSBlaster worm virus and had just started making patches to workstations. Many of the workstations normally connected to both regular and emergency power outlets could not be used until they received the patch. Two generators dedicated to information systems at Beaumont Hospitals’ Royal Oak and Troy facilities kept laboratory, pharmacy, clinical and registration systems running during the blackout. The workstations were cleansed of the MSBlaster virus, with registration and laboratory workstations cleaned first. Registration allowed patients into the information systems for tracking. The hospital was short of PCs because only a distinct amount are connected to emergency power and not all of these were yet virus-free. Moving to backup paper forms was not a problem. When systems are taken down for software upgrades, the staff must go to paper processes. The downtime during upgrades provided the drill-time needed. Simulated drills are hard to do, while the downtime from upgrades is real.

Emergency Power

29

At Memorial Sloan-Kettering, one of three auxiliary units that protect against power surges burned out during a large power surge just before the blackout. Circuit breakers blew, as designed, disconnecting the generator’s link to several servers running critical computer applications. Both problems took time to correct in an emergency environment. Hospitals do not operate on full power with generators during blackouts. Generators run only critical patient care devices and information services. Food preparation is minimal, serving patients and not always employees. Most hospital air conditioning units do not work during a power outage and the workplace gets hot. Beaumont had to truck in water because the blackout shut down the Detroit area’s water distribution system for four days. After the circuit breakers blew at Memorial Sloan-Kettering, the hospital’s e-mail system was down for 30 minutes and its primary clinical system was out for about 90 minutes. The human resources system was off and came up about five hours later. The auxiliary unit that burned out was an uninterruptible power supply (UPS). This battery back-up unit is used to handle power during the transition from normal electrical power to generator power. Having multiple UPS units gave Memorial Sloan-Kettering multiple power sources for information systems. Therefore, most information systems were not affected by the failure of one UPS unit. At Memorial Sloan-Kettering, which had undergone recent renovations, some floors did not yet have emergency outlets connected to a generator, so workstations that are supposed to be connected to normal and emergency power were down. Workstations were plugged into emergency outlets in telephone closets. The hospital also had some difficulty contacting employees at home because so many homes have cordless phones. These phones plug into normal outlets for electrical power to recharge batteries and transmit signals between the base unit and the cordless receiver. A hardwired phone with a handset connected by cord to the unit, receives all of its power from the telephone line. Hospitals and other facilities considering the purchase of generators can purchase more than one unit to split the load and have redundancy in emergency power. Generators should operate at a minimum level of 30% capacity because a generator running below its minimum capacity level may not burn all the fuel in the cylinders, which washes oil off the cylinder walls and cuts engine life.

30


Facilities with generators should run them for one to four hours under load quarterly as part of a regular maintenance program. A dummy load can be created by connecting a generator to a resistive load bank (an electrical unit load with resistors). Allow the water and oil temperatures to stabilize, then monitor performance over the testing time. References Dietderich, Andrew, “Insurers Denying Claims Until Cause of Outage Found,” Crain’s Detroit Business, Vol. 19, August 25, 2003, p. 22. Goedert, Joseph, “The Blackout as a Learning Experience,” Health Data Management, Vol. 11 No. 10, October 2003, p. 14. Hirsh, Michael and Daniel Klaidman, “What Went Wrong,” Newsweek Special Report, August 25, 2003, pp. 37-41. “Trouble All Down the Line,” Time, August 25, 2003, Vol. 162 No. 8, pp. 35-38.

Power Stability and Quality

31

Chapter 2

Power Stability and Quality GRID OPERATION The electric power grid suffered a massive blackout on the afternoon of August 14, 2003 when lights went out from Ohio and Ontario to New York. It seems that a local system failure cascaded over a wide area, but many have long seen major problems in the grid and search for ways to minimize the effects of blackouts. The grid is always involved in a balancing act. The amount of electricity taken from the lines (the load) has to match the electricity being generated. If the power generation drops too much, system controllers have to shed load, causing brownouts or blackouts. The electricity flows through the grid as alternating current so AC frequencies at each station must match. Partial deregulation during the early 1990s allowed some states to separate their generation and transmission industries. Generation systems boomed, but transmission lagged behind due to a patchwork of interstate regulations and jurisdictions. Nationwide policies covering transmission system operation, capacity and investment would force transmission owners to implement a stronger and more resilient grid. Currently protective relays shut down power lines if high currents threaten to make them overheat and sag, but those lines could be kept functioning with more heat-resistant lines, which are already available. Generators switch off if the AC frequency or phase changes rapidly because the generators can damage themselves trying to respond to these changes. The use of breaking resistors, which exchange electricity for heat, could help generators make smoother transitions. Better communications among power stations would also aid in stabilizing the grid. Protective relays rely on local information and may disconnect a line unnecessarily. Dedicated fiber-optics would permit 31

32


comparisons of conditions at adjacent stations, reducing needless shutdowns. The Global Positioning System (GPS) could be used to put time stamps on station readings, allowing operators to make better decisions by using successive snapshots of grid conditions. The Bonneville Power Administration, based in Portland, Oregon, and Ameren Corporation, a St. Louis utility, use GPS time stamping. Once operators get a snapshot of grid conditions, they could transfer the information to faster, smarter switches. Flexible AC transmission devices could tune the power flow. Superconducting valves called fault current limiters would allow circuit breakers to disconnect lines cleanly. Installing more AC lines or more powerful lines would increase transmission capacity but could lead to bigger transients in the grid. If something goes wrong, there has to be a way to contain a disturbance, and the most common way to do that is to disconnect lines. A master computer with a total view could serve as traffic control for the grid. Studies indicate that such a global view would have prevented about 95% of customers from losing power during the 1996 blackouts in the western U.S. One technique to improve control would automatically quarantine trouble spots and divide the remaining grid into islands of balanced load and generation. EPRI has commissioned computer-modeling studies of the technique, called adaptive islanding. These studies concluded that it could preserve more load than conventional responses. Adaptive islanding would take about five years to implement, but blackouts would not disappear. The chance of a cascading failure is real in stressed or highly interconnected systems. With every incremental increase in grid reliability, the cost of the next increment goes up. Over the long term, some improvements in grid complexity could occur. Direct current lines, which have no frequency associated with them, tend to act as shock absorbers to disturbances in AC systems. DC lines separate the Texas power grid from the eastern and western grids. Adding more could help make the grid system more stable, although high-voltage DC is expensive.

VIRTUAL UTILITIES Deregulation of the power industry has changed the way the utility business operates, but changes in the future will be more visible.


33

Most noticeable will be a reduction in the number of high profile, monolithic power stations and their replacement by small, localized generators. While this may improve power availability, it will mean an addition to the number of transmission lines already in place. But, it will also make it more economically viable and practical to incorporate less common forms of power generation. While these developments may lead to the need for more local transmission lines, advances in reactive compensation will allow more lines to be run underground. These changes are currently driven by the needs of manufacturers and organizations that could benefit by generating their own electricity and selling any surplus. The benefits of localized generation will grow and this may reduce the need to transmit electricity over very long distances. This increased volume of privately generated electricity requires more control and management and this means the development of virtual utilities. These are organizations that do not own generating capacity, transmission lines or distribution equipment. They control a power network by paying those who supply electricity to the system and collects from those who use power. They maintain the infrastructure through subcontractors. Microturbines, wind power generators, solar energy and fuel cells do not naturally produce electricity at 50 or 60-Hz. More flexible AC transmission systems would help the efficient connection of these resources to grid systems. Grid stability will become an even more complex issue for control and protection. One technique is to use high voltage direct current (HVDC) systems as isolating links between grids as a way to reduce the need for large-scale synchronization.

DC TRANSMISSION The use of direct current on power networks avoids the problems of instability which can occur on long AC transmission lines and cause surges and blackouts. When connecting isolated grids, HVDC back-toback stations allow power interchange while blocking power line problems. DC transmission means lower line costs with no need for frequency control equipment and lower line losses and costs. But, HVDC stations are more expensive than AC substations and may interact in an adverse manner under certain conditions.

34


ABB Power Systems’ HVDC Division in Ludvika, Sweden, has been involved in several developments in HVDC. These include using insulated gate bipolar transistors (IGBTs) instead of thyristor valves for control. This improves the cost range for dc transmission, making it more feasible for local distribution. Another development is deep-hole ground electrode technology, which drops cabling costs by replacing one wire with an earth return. The electrode can be located close to the station, with reduced power loss and interference and provides opportunities to use monopolar HVDC transmissions. Land electrodes usually require a large area, especially where the earth resistivity is high. But, lower resistivity can often be found between 100 to 200m below the surface due to a higher salt content. This means lower electric potentials and potential gradients than at the surface.

DISTRIBUTED POWER Distributed power generation means placing energy generation and storage as close to the point of consumption as possible with maximum conversion efficiency and minimal environmental impact. Typically, centralized power stations are over-designed to allow for future expansion and so they run for most of their life at a reduced efficiency. They also represent a higher financial risk to the owner because of the greater amount of investment in a single plant. Distributed generation involves dispersed generators, which are customer sized and usually in the service transformer range of 5-500 kW. These are connected at low voltage to the network. Larger distributed generators are about the size of primary distribution equipment such as feeders or substation transformers in the range 2-10 MW and connected at medium voltages to the network. Thermal generators may be used, where heat is the main energy requirement. In one virtual utility project ABB has a partnership with Progress Energy of the USA, which supplies about 3 million customers in the Carolinas and Florida. This project allows the connection of combinations of energy sources, including microturbines, CHP plants, wind power and fuel cells. Internet-based links are used to connect the sources to a central control center. This allows Progress Energy to monitor about


35

10 MW of distributed generation from a single location. The control of virtual utilities requires monitoring and supervision software with aggregation and reporting software to aid decision-making functions. This software interfaces with other packages that link users to trading and forecasting packages. The market rates are continually available and compared with generation capacity. Unit control and dispatching packages complete the software functions. An increase in virtual utilities should result in increased trading competition and lead to further technical developments. This will improve efficiency in system operation along with forecasting and scheduling. Weather forecasting will also be used as a crucial factor in predicting power usage. Improved simulation software should result in improvements in forecasting and scheduling decisions. The methods used for managing a network will also change. Currently, most power networks use a top-down structure, with centralized controls for energy sources. In the future, it should be possible to have a bottom-up integrated optimization of energy supply based on the availability of generators, with a decentralized control. This would allow more resource optimization and reduce the need for high levels of standby capacity. In transmission and distribution, optimum routing will become more important to match energy usage to local sources. The key objective will be to route electricity by the shortest path and to optimize the energy imported from outside regions. In the future environmental and resource conservation pressures will lead to an increased regional and intercontinental energy exchange. This will require enhanced management systems. By 2060, the World Bank estimates that developing nations will consume over twice the power used by established countries.

GRID STABILITY Stability is a property of a system that is disturbed and returns to its original condition. In a power distribution system consisting of transmission lines connected together with two or more generators, the rotors of the generators rotate at a constant speed and are in step as a normal condition. When a fault occurs on one of the lines, the generator closest to the

36


fault will supply the largest portion of the fault current while the other generators supply smaller parts of the current depending on their distance from the fault. The sudden load on the generators will cause them to slow down but not equally. The generator closest to the fault will slow down more than the others and the generators will no longer be in step. The governors connected to the generators will attempt to bring the generators back to normal speed. There will be an angular displacement or difference between the rotors of the generators. The rotor of the generator that has slowed the most will attempt to return to normal while the rotors of the other generators may have already returned to normal speed. The generators will tend to slow down as the load on them is reduced with the slowest generator picking up load. This causes a rocking motion to take place on the rotors. As the fault is removed by taking out of service the line section on which it is located, the rocking motion will decrease and the generators will get back into step. If the fault persists, other generators will try to pick up the load and may fall out of step until a complete shutdown occurs. If the fault is removed quickly enough, the rocking motion will decrease as the generators return in step to their original condition. The system except for the faulted section out of service then returns to normal.

SYNCHRONOUS MACHINES Synchronous machines for power generation are usually the round rotor type or the salient pole type. The large round rotors are generally found in steam turbine generator installations, while the salient pole units may be found in water-turbine generators, synchronous capacitors and synchronous motors. A synchronous capacitor or condenser is a machine which is allowed to float on the line and draw only leading current from the line. It is not intended to supply any mechanical load and is built with a lighter construction than a regular synchronous motor. They are rated in kVA on a zero-leading power factor basis and are capable of receiving capacitive kvars equal to their rating. Synchronous capacitors range in size from a few hundred hundreds of thousands of kvars. The larger units are hydrogen cooled and installed outdoors.


37

The salient pole unit has a relatively large airgap between the adjacent pole sides. Each pole winding is concentrated on the pole and the magnetic field or flux distribution in the air-gap is adjusted to produce a sine wave output by the saturation of, and by chamfering the tips of the poles. The round rotor type, has a distributed field and magnetic flux passes between adjacent poles. The flux decreases in steps going from coil to coil of the distributed winding.

OVERHEAD LINES Overhead transmission lines employ large conductors that may be made of stranded copper conductor or aluminum conductor steel reinforced (ACSR). The conductors must have enough mechanical strength to support long spans under normal conditions and also under the conditions of ice and wind loading. Hard drawn copper for producing the highest strength copper conductors is used as well as aluminum. In 8conductors, the steel core is considered as taking all the mechanical tension. The AC resistance varies with the amount of alternating current flowing, but it is often determined for a current density of 600 amperes per square inch. There is an increase in the AC resistance due to the skin effect eddies which increases as the diameter of the conductor increases. The skin effect for the ACSR conductor is generally greater. An approximate rule for the voltage of transmission lines is 500 to 1000 volts per mile of the line. In the early days of AC power in the United States, the operating voltage increased quickly. In 1890, the Willamette-Portland line in Oregon operated at 3,300 volts. By 1907, a line was operating at 100-kV and in 1913 the voltage rose to 150-kV. In 1926, the voltage was 244-kV. By 1953, lines operating at 345-kV were being constructed. The power that can be safely transmitted over a transmission line at a specified voltage varies inversely with the length of the transmission line. The cost of the transmission line increases directly with its length so that the cost per kW transmitted increases more rapidly than the first power of the length of the line. The series line impedance is mostly reactance and may be halved if the line frequency is changed from 60 cycles to 30 cycles, then the

38


power that may be transmitted over a given distance would be doubled. But, costs and size of the equipment at the lower frequency are much greater and the possibility of flicker in lighting loads, makes these lower frequencies generally impractical. In a 60-cycle transmission system if the line voltage is doubled without altering the line, the amount of power that can be transmitted is about four times greater.

SYNCHRONOUS CONDENSERS In long distance alternating current transmission one or more intermediate synchronous condensers are normally used. They allow the transmission of more power over a given transmission line or system at a specified voltage, resulting in a lower cost per kW transmitted. The reliability of the transmission system is also improved since an intermediate synchronous condenser also aids the stability of a transmission system. For a given size and spacing of conductor the series impedance of a line increases directly with the length of the line. For a given line voltage and a specified operating angle between voltages the line drop is constant in magnitude. The longer the line length, if intermediate synchronous condensers are not used, the smaller the line current will become. If the operating angle is increased, the current will increase but stability is reduced. This prevents increasing the operating angle above a certain value.

TRANSMISSION CIRCUITS Transmission lines supply power in large blocks to load centers. This power may originate from several generating stations and other sources that may be part of a large power pool or grid. When a fault occurs on a transmission line that supplies power to a load that has other sources of power supply, that line will be removed from service and the load served from other sources. Each of the other sources may have to increase its output to take care of the loss of supply from the faulted line. In some cases, the other sources may not be capable of the sudden increase in demand so a part of the load may have to be dropped to prevent the complete loss of all of the load.


39

To maintain a high degree of continuity of service, the system is usually designed so with the loss of the largest source of supply, the rest of the system is capable of picking up the load. Ideally, this should be done without interruption of service or loss of any portion of the load. It is desirable to have power transmitted from a generating station over more than one transmission line in order to provide the continuity of service desired. These transmission lines from a source are often operated in parallel.

OPERATING CONNECTIONS The important operating connections employed in transmission lines include the line step-up and step-down transformers and the high voltage circuit breakers. It is not economical to generate voltages much higher than about 20,000 volts. The power is generated at medium voltages and stepped up by transformers to several hundreds of thousands of volts. At the end of the transmission line, transformers step the voltage down to the distribution voltage which may be as high as 70,000 volts. The distribution transformers then step this voltage down to the distribution voltage which is usually 2300 in suburban communities. In larger communities, the distribution may be done in underground cables at voltages of 6600, 13,200 or 26,400 although higher voltages are also used. Power may be sent 10 or 20 miles at these voltages, although they are not classed as transmission line voltages. For amounts of power greater than 25,000 kW, the voltage used increases directly with the distance involved. This voltage is 500 to 1000 volts per mile so a line of 150 miles would be designed to use 100,000 to 150,000 volts. The transmission system includes the high voltage buses and structures along with the generators and their controls and also the receiver synchronous condensers. There may be two or more transmission lines from the same generating station. They are not paralleled except through the receiver end low voltage load center. The interrupting duty of the low voltage circuit breakers at the sending end of the system may be relatively low. For the continuity of service, the source should have enough capacity to immediately pick up the load being carried by the line which may be removed from service. When there is more than one source, the

40


sources remaining in operation may pick up parts of the load and the remaining capacity of each source and associated transmission line may become smaller. The sources and lines may operate at a fairly large phase angle difference under normal conditions provided that the loss of the largest source of power supply in the system would cause a relatively small percentage increase in line load and operating angle. But, if the increase in line load is relatively large, then the normal operating phase angle may have to be smaller. If there is more than one line, the unit system connection, lines and transformers, may be used with the lines bused at each end on the low voltage side of the transformers.

TRANSIENT OPERATION OF TRANSMISSION SYSTEMS The alternating current electric system is essentially a constant speed system where the speed is controlled within very narrow limits. The governors for the generators operate to keep the machines rotating within these limits. If an increase in load occurs on the electrical system, the governor does not start to operate until a small but definite increase or decrease in speed has occurred. In general, the speed will increase or decrease still further before the governor’s action on throttle mechanisms operates sufficiently to balance the increase or decrease in the electrical load. Up to the time that the balance is obtained, the electrical output or input may be greater or lesser than the input to the prime mover, and the difference between outputs and inputs (neglecting machine losses) is supplied by a decrease or increase in kinetic energy of the rotating parts of the generator and prime mover. Transient operation following a switching operation follows a set sequence. When a generating system is supplying a load through two paralleled transmission lines and a receiver end condenser is operating, the operating angle between sending end and receiver end low voltage buses may be about 10 degrees. If one transmission line with its transformers is suddenly removed from service, then the speed of the generators should increase enough to cause the throttle to operate and the prime mover input would decrease. The mechanical moment of inertia of the generator and prime mover rotating parts, prevent an immediate change in the operating angle.


41

Part of the load is initially supplied by the stored kinetic energy in the rotor of the condenser which begins to drop back in phase angle. The generator rotor begins to advance in phase angle until the electrical power output of the generator is equal to that of the load. At this point, the rotor of the condenser is running at a slightly slower speed than the rotor of the generator, and the condenser rotor drops farther behind in phase angle and the line transmits more power than required by the load. The excess power transmitted becomes available to increase the speed of the condenser rotor back to that of the generator. When this is accomplished, the speed of the two machines is equal but the power output of the line is greater than the load requires and the condenser rotor is accelerated until it reaches a new normal steady state operating angle. But, it is moving faster than it should be and it will move to a lower operating angle where the power output of the line is less than that required by load.

SWITCHING OPERATIONS A switching operation that removes one line from service produces a large increase in load on the line remaining in service and produces electrical-mechanical oscillations between synchronous machine rotors at each end of the system until these oscillations are damped out. The magnitude of the first quarter cycle of oscillation is the difference between the new steady state operating angle and the old steady state operating angle before the system was changed. Since the power output varies as the size of the operating angle between machines, greater angular changes may be produced during the second quarter cycle of such an oscillation than during the first quarter cycle. This type of oscillation normally occurs at the rate of about one to two cycles per second. At the rate of two cycles per second, the percent change in angular velocity is about 0.5 percent. The small percent change in speed produced during the oscillation shows an electrical system under transient operation which is readjusting itself to a new steady state operating condition. There is almost a constant power input through each prime mover during the initial stages of the transient operation. The governors on the prime movers do not act until after the worst

42


portion of the transient operation is over. If a three-phase fault should occur, during the time the fault is on the transmission system, it will reduce the line voltage at the fault to zero and prevent all power from being transmitted between the generators and the load. The time of clearing of a fault may be less than ten cycles on a 60 cycle system or less than 0.2 seconds. If the line was operating at full load, this power is available during the time of fault to accelerate the rotating parts of the prime movers and generators above synchronous speed. At the receiver end the only source of supply is from the receiver synchronous condensers that hold up the load voltage. At the time of fault, the condensers act as generators and supply power to the load as well as to the fault. This causes a decrease in the stored kinetic energy in the rotor of the condensers causing them to slow down. The generators at the sending end of the system are accelerating while the synchronous condensers at the receiving end are decelerating. Both effects produce an increase in the operating angle between internal machine voltages. As the fault is cleared, the operating angle between machine rotors increases. If the operating angle should increase beyond 90°, the maximum power output of the transmission system is reached and the system will pull out of step. If the system does not pull out of step during the fault, then when the fault is cleared the remaining line in service may again transmit power, but the operating angle has increased and the rotors at each end are no longer operating at identical speeds. Even if the new steady state operating angle with one line out of service is reached at the time the fault was cleared, there will be an overswing before the rotors again reach identical speeds. If this overswing reaches beyond 90°, then the system would fall out of step before the machine rotors reach identical speeds. During the fault and overswing, after the switching operation to remove the fault is complete, the actual change in speed of the system may not be sufficient to cause the governor of the prime mover to operate and to change the input to the generators in the short period of time involved. The stored kinetic energy in the machine rotors, along with the power angle determines any acceleration of the rotors above or below synchronous speed. During the fault the system as a unit begins to increase in speed above synchronous speed and there is also a net acceleration between machine rotors. This net acceleration between rotors


43

determines if the machines stay in synchronism or pull out of step. The machines normally do not lose synchronism during the fault, if it is cleared in a reasonable time by automatic breakers and relays, but the difference in angular velocity acquired during the fault is often sufficient to cause the machines to pull out of step after the fault is cleared. This becomes a function of the amount of synchronizing power which may be transmitted after the fault is cleared, so that there is also a net acceleration after the fault is cleared which tries to change the rotor velocities so that they are again equal. The line section removed when the fault is cleared reduces the voltage at the load so that the load does not take as much power as formerly. The system as a whole continues to increase in speed until the governors act to reduce the input but this occurs after the worst of the transient oscillations takes place. If the machines do not lose synchronism at the end of the first half cycle of the oscillation, then they will probably not fall out of synchronism, because during the next cycle there is more time for the voltage regulation to increase the transient internal voltage and the oscillations are being damped out by losses in the field circuit. These losses are produced by the oscillations. Damper windings are used by synchronous condensers while other machines use an approximate equivalent in the solid rotor and metal wedges in their construction. These winding make the damping effect more rapid.

FAULT REMOVAL The protection of transmission systems from faults must be accomplished by removing the fault as quickly as possible. The usual technique involves high speed circuit breakers actuated by relays. The circuit breakers usually operate in oil. Smaller sizes operate in air. In some air breakers the arc that forms when the contacts separate is controlled by the magnetic field into a series of small gaps. These gaps aid deionization and extinguish the arc quickly. In some types of circuit breakers, compressed air is used to extinguish the arc. In oil circuit breakers, the mechanism is immersed in an oil similar to transformer oil. When the circuit is opened and an arc forms, some oil is vaporized and a pressure is built up. The pressure of the oil helps to

44


extinguish the arc. Some circuit breakers are mounted in the same oil with the transformer. The relaying of faults depends on the types of relays used. These include overcurrent and directional overcurrent relays as well as pilot systems. These types of relays are not used for the protection of long single lines or loop systems. Here, distance relays are employed. They compare the voltages and currents during a fault which are a function of the circuit constants between the relay and the fault along with the distance to the fault.

OVERCURRENT RELAYS During a short circuit the current flow in the affected conductors is greater than the load currents. This difference allows the relay to discriminate between loads and faults. The fault current will depend on the fault location, type of fault, connections and amount of generating capacity. The measurement of fault current alone does not determine the location of the fault. A system can be sectionalized with relays set to operate on a minimum fault current for a fault in that section. To prevent tripping of corresponding breakers, time delays are used in the operation of the relays. The time delay increases as the source is approached so that the breaker nearest the fault clears before other relays operate. When fault currents can flow on either side of a bus, a direction element is used which allows the relay to trip when only current is flowing away from the bus. An inertia factor affects the time required to clear a fault. This is the ratio of stored kinetic energy in the machine rotors at the sending end to the stored kinetic energy in the prime movers at the sending end. In a water wheel generator, the allowable time for clearing a fault varies from about 0.04 to 0.14 seconds. This takes about 2.5 to 8.7 cycles. A large moment of inertia is present in water wheel generators. Steam turbogenerators run at a higher speed and have a higher amount of stored kinetic energy. A transmission system using steam turbo-generators will have an inertia factor of about 10-12 and the allowable time for clearing the fault of more than twice the time if the generators were water wheel machines having factors of about 3.


45

Circuit breakers and relays are designed to provide a high speed interruption which improves the stability. A total relay and circuit breaker time of less than three cycles is typical. This is improved with newer electronic relays. When several breakers are in series, the cascaded time settings require that the relays nearest the generators have longer time delays. When more than a single source of supply is connected to the system, especially in a ring bus system, the appropriate time settings must be calculated. The short circuit currents which exist under the known circuit conditions are calculated by the method of symmetrical components and the currents through each of the relays determined from these calculations. Differential, pilot and balanced protection all depend on the comparison of currents which have defined relations to each other under normal conditions, and much different relations during the fault conditions for which the breakers should trip. Differential protection is used for protecting generators, transformers and bus structures from internal faults. The vector sum of all of the currents in each phase, flowing out of the equipment being protected, is sensed by the relay. If the currents in the relays are not at the same voltage base, as is the case for transformers, they are brought to the same base by current or auxiliary transformers in the relay circuits. In some cases, the relay itself is used. For generators, the vector sum of the currents flowing out of each phase winding is sensed by the relay. For transformers the vector sum of the currents in each winding on the core, after correction in instrument transformer circuits is sensed by the relay currents on a common voltage base. The vector sum of all the currents in each phase of each line connecting to a bus is sensed by the relay. Under normal conditions and for faults outside the equipment being protected, the vector sums should always be zero. When a fault occurs in the equipment, a current should flow which is not included in the vector sum. In transformers, a fault between turns of a winding will change the effective transformer ratio making the sum no longer zero and a current flows through the relay. The relays are given the most sensitive setting which will not cause operation for faults outside the apparatus. The main causes of faulty operation are saturation in current transformers from high values of current, and transformer energizing transients, which cause currents to flow momentarily in the winding which has just been energized.

46


When differential protection is used for polyphase transformer banks with delta connected windings, all the corresponding currents in the windings must be considered. The current transformers are inserted in the leads of the individual transformers before the delta is closed. Pilot protection is a similar type of protection used in transmission lines where the ends of the circuit are some distance apart. Auxiliary circuits using pilot wires are employed to obtain the vector sum. Pilot protection involves the simultaneous opening of circuit breakers at the terminals of transmission lines using the pilot wires as a communication link between the circuit breakers. These physically separate pilot wires are connected in various ways. In the circulating current pilot wire technique, current transformers are used for sensing the load currents and fault currents which circulate over the pilot wires. This technique is also used with current balancing relays. Another scheme uses pilot wires with percentage differential relays. The directional comparison pilot wire scheme uses direct current over a pair of wires with polyphase directional relays. The link may also use the transmission line conductors themselves with a very high frequency current superimposed on the line. The carrier current signal operates the relays to keep the circuit breakers closed. A fault on the line interrupts the signal and opens the breakers, causing the line to fail safe. Pilot systems also use radio signals with microwave channels operating from relay stations located in a line-of-sight along the route of the transmission line. The relays operate to de-energize the line in the same fail safe manner. Balance current protection involves the difference of corresponding currents in two similar parallel lines. As long as the lines are alike, the relay current is zero. A fault on one of the lines will cause a difference in the two lines and unbalances the currents unless the fault is near the far end. In this case, relays at the distant end will operate to open the breaker in the faulted line. Then, a difference in current occurs at the near end, since the lines are no longer tied together at the far end.

POWER QUALITY MONITORING In the past, most electrical equipment represented a linear load. This was usually a resistive or inductive load such as lights or motors.


47

This type of equipment can tolerate disturbances on the electrical transmission and distribution systems. As electronic controls and other circuitry was added to these linear loads, a consistent sinusoidal voltage became more important. These electronic loads can cause the sinusoidal voltage to be deformed. Some of the early electronic equipment were not tolerant of events on the power system and disturbances would often cause the electronics to fail. Electronic equipment has improved significantly and most now have some type of built in protection from transients. However, the protection does not always protect the internal devices. Power quality monitors are available with a variety of features and methods of capturing data. Low end power quality monitors usually capture voltage, current, and detect voltage sags. Mid-range monitors capture power factor, energy, and waveforms along with the features of low end power quality monitors. High end units can also capture high speed transients and harmonics. Some equipment has a range of these features.

RMS VOLTAGE The root mean square voltage is a basic measurement provided by a power quality meter. But, not all meters capture RMS voltage the same way. They use a variety of methods of calculating the voltage and storing the data. Important factors are the sampling rate and the type of data that is captured. Most low cost hand-held meters do not actually calculate the RMS voltage. They measure the peak voltage and divide the voltage by the square root of two, which assumes that the voltage is sinusoidal. Other meters known as true RMS meters calculate the RMS value by sampling the waveform multiple times per cycle, square the values, average the square voltages and then calculate the square root. Most meters that are referred to as power quality meters actually measure the true RMS value, but low grade meters use the peak detection method. The sampling rate can refer to both the number of samples per cycle the meter uses to calculate the RMS voltage and how often the meter samples a cycle of data. The number of samples that a meter uses per cycle is generally from 16 to 512. Low end meters sample the AC waveform 16 to 64 times per cycle, mid-range meters usually sample the

48


AC waveform 128 samples per cycle and high end meters use 256 or more samples per cycle. A higher sampling rate allows the RMS voltage to be more accurate. The frequency that a meter samples a cycle of data is also important. Most power quality meters calculate the RMS voltage every cycle but some meters only perform the calculation once a second, minute, or even 15 minutes. The meter should calculate the RMS voltage for every cycle because variations that occur between cycles will affect the RMS data that the meter captures. The type of data that a meter captures can be the average voltage, minimum voltage, or maximum voltage. The meter needs to capture average voltage, but the minimum and maximum voltages are also important. The minimum and maximum voltages indicate how much the voltage varies from the average during an interval. The long-term minimum, maximum, and average RMS voltages indicate the long term voltage regulation. Monitoring the average voltage over a period of time can indicate if the utilities’ voltage regulation equipment is operating properly. The minimum and maximum voltages will also indicate if a sag or swell occurs in a certain time period when the monitor calculates the RMS voltage every cycle. At one facility, a chiller would trip off at night with an overvoltage error code but spot checks of the voltage showed that the voltage was within tolerance. When a power quality meter was used, the RMS voltage showed that the voltage would reach 512V on a 480V service during the night. While this was within the utility’s allowable tolerance of 480V +/-10%, the voltage was over the 460V chiller’s 10% tolerance. The utility investigated and found a setting on a voltage regulator was incorrect.

RMS CURRENT Most power quality meters are able to monitor current in addition to voltage. The basic measurement of AC current is the RMS current while the important factors are the sampling rate and type of data collected. Most meters that capture both RMS voltage and current usually either have the same sampling rate or one half the sampling rate for the current waveform. The RMS current indicates the load on electrical equipment such as transformers and breakers. Recording the RMS current over periods of


49

time can be used to determine the peak and average load of electrical equipment.

POWER FACTOR Power quality monitors that measure both voltage and current also provide the power factor which is a measurement of the phase angle difference between the current and voltage. It indicates the ratio of real power delivered to the total power being delivered. The power factor is calculated by dividing the real power (watts) by the apparent power (voltamps). The apparent power is found by multiplying the RMS voltage by the RMS current. The real power is calculated by multiplying each voltage and current sample together for a cycle, then summing them and dividing by the number of samples per cycle. This method of calculating the power factor takes into account the harmonics and is called the total power factor. Another method uses only the fundamental 60Hz contribution to real power and is called the displacement power factor. Most power quality monitors calculate both the total and displacement power factor, but some only calculate one and call it the power factor. Some utilities charge a penalty if a facility has a low power factor or they may charge for reactive power, which does no real work. A low power factor uses up the capacity in transformers, breakers and conductors since they are delivering reactive power. If a utility does not charge for kVARh or a low power factor, then they will be charging for extra facilities if a large transformer is needed to deliver the kVAR. If the power factor is not monitored, the spare capacity of a facility’s electrical system may not be known, which effects decisions about new loads. A plant may need to expand, but their electrical system may not have enough capacity to add the additional load from the expansion. After monitoring the power factor and current, there may not be enough capacity because their power factor is 65%. The plant would need to have a larger transformer and switchgear installed or they can improve their power factor using capacitors. The low cost solution is usually to install the capacitors and free up the capacity in the existing electrical system.

50


POWER AND ENERGY Power quality meters that monitor both voltage and current usually have the ability of measuring energy. Some of these meters are even certified as revenue accurate. This means the meter can accurately measure billing parameters such as kWh, kVARh, and demand. This information is useful for the daily operations of a facility and has the potential to provide substantial savings. Electric utility charges may be based on peak demand, kWh and power factor. Peak demand is the maximum kWh consumed over a 15 or 30 minute interval. The interval can be either a sliding window or fixed window. Knowing when the peak demand occurs can point to ways of lowering the peak period. Peak demand can be lowered by staggering the start of large loads like chillers outside the demand window. Detailed kWh information allows the facility to take advantage of off-peak hours when the price per kWh is lower.

VOLTAGE SAGS Voltage sag detection is a basic feature of a power quality monitor. A power quality meter can monitor the cycle by cycle RMS voltage for periods when the voltage is below 90% of nominal. The monitor can record the lowest RMS voltage, the duration of time that the voltage is below 90% and a timestamp of the data. This type of measurement is called magnitude and duration or Mag-Dur. The method provides an exact timestamp and duration of events. The data captured with voltage sag detection can be used to analyze equipment misoperation from voltage sags. Mitigation can be applied to the most sensitive critical loads first.

WAVEFORM CAPTURE The more expensive power quality meters are able to capture the waveform when a voltage sag is detected. This is an definite advantage over a meter that only captures magnitude and duration. Voltage sags with the same RMS voltage can effect equipment in different ways if the waveforms are different.


51

The waveform can also be used to analyze the effect of capacitance on loads that have enough energy to ride through the voltage sag. If the same event happens at a different time, equipment can be affected differently depending on loading conditions. Analysis of the waveform can be used to determine why certain equipment tripped off and other equipment did not. Waveform capture is also useful in determining if a voltage sag was upstream or downstream of the meter by comparing the current and voltage profiles. If the current increases at the same time as the voltage decreases, the sag is caused by an event downstream of the meter. When the current increases after or decreases during the voltage sag, then the event was caused by something upstream of the meter. This can be used to determine where a problem occurred.

TRANSIENT CAPTURE High-end power quality meters have the ability to capture transient events that may not affect the RMS voltage enough to trigger a voltage sag or swell. This type of event is captured by a snapshot of the deviation of the waveform from the expected or normal waveform. Transients may be caused by equipment switching or lightning. Switched capacitor banks will cause transients when they are switched on, they draw a high initial current, which reduces the voltage until the capacitor is fully charged. An overvoltage can be caused by the inductance in the system from changing currents. This event is not captured by the voltage sag detector because the RMS voltage does change enough. Other electrical equipment can also cause transients due to switch or relay contact noise. At one facility, several motor drives were tripping off and the monitor showed an overvoltage error code. A check of the voltages indicated that the voltage levels were normal. Monitoring the incoming voltage indicated that a capacitor switching transient was causing an overvoltage on the DC bus of the motor drives. A capacitor clamp was installed along with line reactors on the drives. The capacitor clamp and reactor reduced the transient and the drive operated properly.

52


HARMONICS High-end power quality meters are also able to monitor harmonic frequencies. Most meters can monitor the total harmonic distortion (THD) and some can monitor individual harmonic frequencies. The meter usually monitors individual harmonics from 2 to one half of the sampling frequency minus 1. A monitor taking 128 samples per cycle would monitor harmonics 2 to 63. This is done by taking the Fourier transform of the voltage waveform to determine the magnitude of the individual harmonic components. The harmonics with the higher magnitudes are the 3rd, 5th, and 7th. The 3rd harmonic is caused mainly by single phase electronic switching loads while the 5th and 7th harmonics are more likely to be caused by three phase electronic loads like motor drives. Harmonics are important because electrical equipment such as transformers, breakers and conductors should be derated to allow for the heat caused by the harmonic currents. Electrical equipment is designed and sized for 50 or 60Hz AC current. As the harmonic currents flow through the resistance in the equipment, they cause heat. In transformers and motors, the higher frequencies cause higher core losses from eddy currents. Circuit breakers may trip early from the generated harmonics. One facility was experiencing tripping of a new 800A breaker. After monitoring the harmonics at half load, it was found that the peak current was allowing this peak sensing breaker to trip. The harmonics added to the peak of the current while the RMS current was not high enough to trip the old thermal breaker. At half load on the breaker, the RMS current was measured at 509A, with a peak of 720A. The actual RMS current was only 321A. At full load, the peak current was enough to cause the 800A breaker to trip. Solving problems with transients and harmonics requires power quality equipment to record the types of events that are seen by the facility and affects equipment. Event notification is available in some power quality meters. This feature starts the process of sending text messages to any device with an e-mail address. These events range from peak demands to voltage sags. The notification of peak demands can result in significant cost savings. A voltage sag can indicate a brownout has occurred. A full power backup with some type of uninterruptible power supply to ride through voltage sags is expensive.


53

A power quality issue may be misdiagnosed based on incomplete monitoring. This can lead to incorrect solutions being applied or the solution being incorrectly sized. The voltage sag’s depth and duration needs to be known along with the equipment response. POWER QUALITY ENVIRONMENT In the past, sensitive data processing equipment was generally located in a controlled environment like a data center. This facility was usually designed with special attention to the electrical and environmental support systems. Publications such as Federal Information Processing Standards Publication 94 were used to guide the design and installation of the data center’s electrical system. Equipment that was used in the general office environment or on a factory floor generally did not contain sensitive electronics. The advent of microprocessors, personal computers and networked computer systems changed all that and electronic equipment might be installed almost anywhere. Personal computers and microprocessor-based equipment are now used in offices, warehouses, factory floors, and other locations. These facilities can have equipment problems due to poor wiring and other power quality related issues. Uninterruptible power systems (UPS) are important tools in protecting computer systems from power failures. UPS equipment uses a battery system to provide energy during the power failure, providing enough time to save data and properly shut down the system. Many users also have backup generators if the power failures are extended. One benefit of using an UPS is that sensitive equipment gets conditioned power. An UPS takes AC power, converts it to DC for the batteries and then converts it again to AC. The circuitry has enough feedback to keep the power output of the UPS isolated from most input fluctuations. The UPS output voltage is solid and does not change with the input voltage. Most electrical noise at the UPS input is filtered out.

ELECTRICAL NOISE Electrical noise refers to any undesirable electrical signals that could affect an electrical or electronic circuit. Noise can cause server

54


reboots, data corruption, lock-ups and slow down network operations. Electrical noise may be defined as common mode, transverse mode or interference. Common mode noise refers to the electrical signals between a circuit conductor and the grounding conductor. In a balanced three-phase system common mode noise should be equal in magnitude and in phase to all conductors and ground. The transverse mode is also called normal mode or differential mode. It refers to the signals that exist between a pair of circuit conductors. Electrical noise can affect the data stream in a data communication signal traveling between two computers or between a computer and printer. The data at the receiving device can be different from the data the transmitting computer sent. Computers have the capability of detecting and correcting these data errors. But, this uses some of the data transmission capability of the link and causes delays. If the errors occur too frequently, the error correction system may become swamped and inoperative. There can be radio frequency interference and electromagnetic interference. Radio frequency interference is typically higher in frequency and is capacitively coupled into the electrical system. The electrical wiring acts like an antenna for radio frequency interference. Sources of radio frequency interference include radio and television transmission. Electromagnetic interference is inductively coupled through the electrical system through the conductors in transformers, motors and other wiring. Computer manufacturers expect neutral to ground voltages to stay below a specific level. In some cases this may be 500 millivolts. Erratic operation can occur if the neutral-to-ground voltage exceeds this level. The neutral-to-ground voltage can be kept low by using dedicated circuits for critical equipment, isolating high current circuits, keeping circuit runs short and using larger conductors. Wiring in a cable tray allows magnetic fields from adjacent current carrying conductors to be introduced into connected circuits, resulting in noise in these circuits. If the adjacent circuits are spaced further apart, this minimizes the effect of each circuit’s magnetic field on the other. The ground wire for each circuit should be located equidistantly from each circuit conductor. Then each conductor’s magnetic field influences the ground wire equally. The magnetic fields on the ground wire


55

cancel and the result is smaller voltage or current flow in the ground. If conduit is used, it acts as an electromagnetic shield and limits electrical noise from magnetic fields. The closer proximity of conductors in the conduit also minimizes the effects of magnetic fields on the conductors. Rigid metallic conduit can be used as a ground conductor in certain applications instead of an actual ground cable. The National Electrical Code should be used to make sure the electrical installation is safe. Power quality measured are addressed in The Institute of Electrical and Electronic Engineers publication Standard 1100-1999 IEEE Recommended Practice for Powering and Grounding Electronic Equipment.

DISTRIBUTION PANELS Power quality related problems can also originate at the distribution panels. Using a single conduit for distribution panel service and a single conduit for the load or branch circuits does not aid power quality. Running all the branch circuits in a single conduit allows each conductor to produce a magnetic field that gets induced into the other conductors. Noise proliferates and there is no shielding. A neutral-ground bond should only be made in the distribution panel if it is not made anywhere else upstream. It is a good practice to separate non-critical loads with critical loads. Modular cubicles can be a source of power quality problems. The electric wiring scheme is often not designed for sensitive equipment. The wiring may be undersized and the connections may be degraded by temporary loads such as space heaters, fans, and other loads that can overload the electrical systems or cause noise problems.

POWER STRIPS Power strips provide a convenient way to quickly increase the number of power outlets. One wall outlet can be expanded into an additional six outlets. High quality power strips can be effective against noise and surge problems. But, many power strips are not well made. They can have undersized wiring or poor connections. In some units, wires have been connected to the wrong prong of the outlet (hot and neutral reversed). Many devices have spike or surge suppressors that

56


may fail and provide a short circuit between neutral and ground. Plugging one power strip into another for more outlets also causes potential problems. One facility that had five or six power strips attached to a single 15 or 20 amp circuit had an AC current that was high enough to burn the first power strip in the chain.

NEUTRAL-GROUND VOLTAGES Neutral-ground voltages can be kept low by keeping circuit runs short and loads low. The neutral-ground voltage can be measured with a digital voltmeter and a special adapter. The amount of voltage measured is a function of the circuit’s neutral current times the neutral impedance plus any ground current times the ground wire impedance. Besides measuring neutral-ground voltage with a voltmeter, it can be monitored with an oscilloscope. This will indicate if high frequency noise is present. Some equipment manufacturers recommend that their equipment not be used when a certain neutral-ground voltage is exceeded. Isolation transformers have a neutral-ground bond providing a zero voltage neutral-ground reading. If neutral-ground voltage measured at the distribution panel is very high, then there could be a problem such as a missing neutral-ground bond for a service transformer.

GROUND CURRENTS Making ground current measurements at various points in an electrical system can be used to indicate the state of the system. There should be a small amount of ground current since almost all electrical equipment has some leaking to ground. The measured ground current should be only 1 to 2% of the phase reading. A zero reading can indicate an open ground which is a safety hazard if a wiring or equipment fault occurs. If large ground currents are measured, then these need to be investigated. Isolating the source of ground currents may be difficult. Check all ground wires in the system. An electrical one-line diagram and physical drawing of the electrical system wiring will be helpful. Make measurements for each link on the drawings. Measure all bonding conductors, connections to the grounding system and even conduits.


57

Ground currents will return to the source through the lowest impedance path. These paths include ground wires, conduits, equipment chassis, grounding electrodes or building steel. The ground current can be measured with a clamp-on ammeter without removing wires. Clamp the meter around the phase conductors and neutral. Ideally, the current combined in the phase conductors and neutral should sum to zero. If the meter does not read zero, then ground current is flowing. Another method of finding ground currents is to place the ammeter around the ground conductor in the distribution panel and turning off circuit breakers one at a time. If the current reading drops when a circuit breaker is turned off, that circuit is at fault.

POWER CONDITIONING Power conditioning equipment includes uninterruptible power supplies (UPS), static transfer switches, shielded isolation transformers, power distribution units, magnetic synthesizers, transient voltage surge suppressors and filters for noise and harmonics. These must all be properly installed, grounded and wired. A power conditioner will not cure a poorly constructed electrical system. An electrical system contains a chain of transformers, generators, switches, circuit breakers, distribution panels, power conditioners, cables and other equipment. The chain depends on its weakest link. UPSs, static transfer switches, generators and automatic transfer switches have become more electronic. Many devices are microprocessor controlled with much sensitive circuitry. They need proper wiring and grounding, like most electronic devices. Some equipment may use internal static switches that switch neutrals, so special attention may be needed for grounding requirements. Signal cables may be sensitive to noise.

GROUNDING SCHEMES One facility decided it was best to totally isolate the data center grounding system from the rest of the building. A ground ring was placed outside and all ground feeds coming into the data center were disconnected. The data center floor, power conditioning equipment and

58


computer equipment were tied to the isolated ground ring. This is in violation of the National Electrical Code and a fault in a piece of equipment could cause dangerous voltages to develop on the chassis of the equipment. Later, a lightning strike would hit the ground close to the ground ring. Since the ground ring was connected to the data center and not earth ground, the energy flowed into the data center and destroyed most of the data center’s equipment. In another instance, a high neutral-ground voltage was measured at a power receptacle. The equipment manufacturer recommended less than half a volt. Another neutral to ground bond was added in the upstream distribution panel. This made two neutral to ground paths in the same part of the electrical system, forming a ground loop and violating the National Electrical Code. Neutral currents were allowed to flow through both neutral and ground conductors. The excess ground current created enough noise voltage to create problems in the entire computer network system.

POWER AUDITS Interruptions of service, network problems and other unexplained hardware failures may indicate that there is a power quality problem. A facility power audit is one way to determine the state of an electrical system. It can be a valuable aid on where to begin the troubleshooting process. Power quality audits begin with a review of the facility electrical systems one-line diagram. An inspection of the facility main electrical service follows. It should be properly bonded and grounded. An inspection of the grounding electrode system should confirm that it has not been damaged or altered. Measure the current through the main service neutral-ground bonding conductor and the current flowing into the ground electrode system. This is an important indication of grounding problems. Ground impedance testing can also be conducted, a good ground should have an impedance of less than 0.1 ohms. Start at the main electrical service entrance and inspect the switchgear, wiring routes, cables and distribution panels from source to load. Take sample readings throughout the facility. These include voltage, current, and harmonics. A power quality analyzer can be used if a problem is suspected.


59

Review the application and operation of any power conditioning equipment. Inspect the connections to load equipment. There may be better ways of connecting the loads. A power problem may have more than one solution. At one facility, a computer network system was installed in an old warehouse. The branch circuits had no ground wire because the branch circuit conduit was being used as the grounding conductor. The computer network was experiencing problems. The electrical system met building code requirements, but it was thought to be inadequate for the network application. One recommendation was to rebuild the electrical system. Another suggestion was to power each network workstation with a small UPS. This solution was inexpensive and tried, but did not resolve the problem. Finally, it was determined that the network cable runs were too long, combined with the poor electrical environment which resulted in poor operation of network. The solution was to replace the network wiring with fiber-optic cables. References Buff, Katrina, Editor, Energy and High Performance Facility Sourcebook, Editor, Proceedings of the 26th World Energy Engineering Congress, The Association of Energy Engineers, Inc., The Fairmont Press: Lilburn, GA, 2003. Minkel, J.R., “Heating the Grid: Several Near-term Solutions can Keep the Juice Flowing,” Scientific American, Vol. 289, November 2003, pp. 18-20. Pansini, A.J., Power Systems Stability Handbook, The Fairmont Press, Inc.,: Lilburn, GA, 1992. Russell, Eric, “Virtual Utilities: the Shape of Things to Come,” European Power News, Vol. 26 No. 5, May 2001, pp. 7-9.


Standby Power Systems

61

Chapter 3

Standby Power Systems An important measure of an organization’s strength is its ability to respond successfully to emergencies. The risk posed by emergencies can be reduced by careful planning using adequate guidelines, yet flexible enough to be adaptable to sudden changes and varying demands. A sound electrical system is needed with emergency and standby power supplies to keep critical processes and equipment going during the emergency. Ideally, there should be a seamless transition between normal and standby power. In real life, however, this may be difficult and too costly, so compromises must be accepted.

STANDBY POWER REQUIREMENTS The first part in determining standby power requirements is to determine the duration of power interruption that can be tolerated. This sets the criticality of the electrical load. If a load cannot be interrupted for more than half a cycle (1/120 of a second for a 60-Hz system), it is called a critical load. If a load cannot be interrupted for more than 10 seconds, it is classed as an essential load. If a load can be interrupted for the duration of the normal power failure, it is called a non-essential load. The type of load determines the standby power requirement. The source of energy storage determines the type of standby and emergency system.

STORAGE DEVICES Types of energy storage available include thermal, mechanical and chemical. In thermal storage, wind-generated energy is converted into heat using an electric-resistance heating. This heat is stored as hot water, a warm bed of rock or gravel, or molten heat-storage salt. Mechanical 61

62


energy storage involves lifting an object such as water so gravity can later return it. Energy storage can also use a flywheel since a spinning disk stores energy. The more weight, and the faster it spins, the more energy a wheel can store. Advanced flywheels are not very heavy, but they spin at 30,000 rpm or more. The energy stored in a flywheel is directly proportional to weight but it increases with the square of the rpm. So, you get quadruple the energy for double the rpm. A large amount of energy can be stored in a fast-spinning wheel. At such high speeds, air friction is considerable, so some flywheels are installed inside a vacuum chamber. The bearings must be very precise, carefully designed and built. Chemical energy storage can be done using electrical power to split a compound, like water, into its constituent parts, hydrogen and oxygen. These constituents are stored separately and later recombined to produce electrical power as needed in a fuel cell. These cells are becoming more available and while still expensive, offer a way of using the energy of chemical bonds to provide energy storage. Another approach to chemical energy storage is the more traditional battery storage device. The metal plates inside these batteries act as receptors for the metal atoms from the electrolyte (acid in lead-acid batteries) as the batteries are being charged. When the battery discharges, these metals return to solution in the electrolyte, releasing electrons and generating direct current at the battery poles. The various types of different batteries are named for the type of plates or electrolyte used. The most common are the lead-acid and the nickel-cadmium batteries. Lead-acid batteries are used in cars, golf carts and other common applications. Nickel-cadmium, or nicad batteries are used when higher costs are not a penalty for lower weight and improved tolerances to overcharging. Airlines generally use nicads. While you must be careful not to overcharge a lead-acid battery, or discharge it too quickly, nicad batteries can stand up to these abuses. But, if you discharge a nicad only half-way enough, eventually half-way will be as far as the discharge will go. Batteries are effective in providing an emergency power source for fire, alarms, emergency communication, exit signs, protective relays and emergency lighting. These loads generally have individual batteries with a battery charger, that is connected to AC power. The National Electrical Code (NEC) states that the system should be capable of maintaining the load for 90 minutes without dropping below 87.5% of normal


63

voltage. These systems are simple and reliable and robust. Maintenance requirements are minimal, but since all batteries have a finite life, they must be tested and replaced.

BATTERY RIDE-THROUGH Lead-acid battery based ride-through systems are common. Storage batteries generate electricity by chemical action. Electric storage was used by the Italian count Alessandro Volta. His voltaic pile in 1800 used a cell with round plates of copper and zinc as electrodes. Cardboard soaked in salt water was the electrolyte. Current flowed but this primary cell could not store power for any length of time. The lead-acid battery was invented over 100 years ago. It consists of a number of cells connected in series. A fully charged cell will measure close to 2.2 volts. The cells are enclosed in individual compartments in a rubberoid or high-impact plastic case. The compartments are sealed from each other and except for zero maintenance types are open to the atmosphere. The lower walls of the individual compartments extend below the plates to form a sediment trap. Filler plugs are located on the cover and can be combined with wells or other visual indicators to monitor the electrolyte level. The cells consist of a series of lead plates connected by internal straps. The plates are divided into positive and negative groups and separated by means of plastic or fiberglass sheeting. Some very large batteries, which are still built almost entirely by hand, continue to use fir or cedar separators. A few batteries intended for transport service have a woven fiberglass padding between the separators and positive plates. The padding helps support the lead filling and reduces damage caused by vibration and shock. Both sets of plates are made of lead. The positive plates consist of a lead grid works that has been filled with lead oxide paste. The grid is stiffened with a trace of antimony. Negative plates are cast in sponge lead. The plates and separators are immersed in a solution of sulfuric acid and distilled water. The standard proportion is 32% acid by weight. The level of the electrolyte drops in use because of evaporation and hydrogen loss. Sealed batteries have vapor condensation traps molded into the roof of the cells. Nonsealed batteries must be periodically replenished with distilled water. When a cell is fully charged, the negative

64


plates consist of pure sponge lead. The electrolyte consists of water and sulfuric acid. During discharge both sponge lead and lead dioxide become lead sulfate. The percentage of water in the electrolyte increases as the SO4 radical splits off from the sulfuric acid to combine with the plates. During the charge cycle the reaction reverses. Lead sulfate is transformed back into lead and acid. A small quantity of lead sulfate remains in crystalline form and resists breakdown. After many charge-discharge cycles, this residual sulfate reduces the battery’s output capability. The battery is then said to be sulfated. Sulfation becomes more certain below a 70% charge. In addition to the possibility of damage to the plates, a low charge brings the freezing point of the electrolyte up to near 32°F. Temperature changes have a major effect on the power density. Batteries perform best at room temperatures. At 10°F the battery has only half of its rated power. Both the energy density (Whr/lb) and power density (W/lb) are not high. The average life is about 360 complete charge-discharge cycles for non-deep cycle batteries. It is not possible to discharge a lead-acid battery completely, even those that have be stored for years still have some charge. During charging, about 3/4 of the input can be retrieved giving an efficiency of 75%. Sodium-sulphur, lithium-halide, lithium-chlorine and zinc-air batteries can provide better performance, but are not as common due to high costs or complex charging procedures. Most chemical batteries must be replaced every 6 to 7 years. The replacement and disposal cost must be factored into the total energy cost. Battery systems are heavy relative to the amount of power they can deliver and those applications requiring significant power storage may weigh several tons. Over 99% of all power disturbance are a few cycles to a few seconds long. Many times an UPS will be used for the few seconds that it takes for a standby engine generator to start and become operational. Storage batteries are typically used to provide the necessary backup power to the load during this short time period. These battery systems need constant monitoring, maintenance and service. Batteries also require a large environmentally controlled storage area and special disposal procedures, which increases the operational costs through regulatory compliance. Each discharge and charge cycle also decreases the useful cell life.


65

When lead-acid batteries are regularly deep cycled to 80% or more of full discharge, their life is usually limited to less than 600 charges/ discharges. When only shallow discharges are used, leaving 2/3 or more of the battery’s full capacity, the number of charge/discharge cycles can go above 10,000. For most UPS ride-through applications, the discharges are relatively infrequent (less than 100 per year) and a short duration does not require complete discharge. Chemical-battery-based, ride-through systems typically use sealed, maintenance-free automotive batteries with thicker lead plates and higher electrolyte levels. Smaller systems can provide 250-kW for 10 seconds while larger 10-MW, 10-second are available. A 1-MW system will use 240 lead-acid cells weighing 46,000 pounds. These are capable of delivering 1,800 amps for five minutes. Full output is achieved within two milliseconds of a utility power interruption.

BATTERY TYPES Performance is limited by the lead-acid battery packs which are generally the most affordable option. More exotic batteries like nickel metal hybrid (NiMH) packs have also appeared. The common 12-volt lead-acid battery has six cells, each containing positive and negative lead plates in an electrolyte solution of sulfuric acid and water. This proven technology is not expensive to manufacture and it’s relatively long-lasting. But, the energy density of lead-acid batteries, the amount of power they can deliver on a charge, is poor when compared to NiMH and other newer technologies. The United States Advanced Battery Consortium (USABC) is a Department of Energy program launched in 1991. Since 1992, USABC has invested more than $90 million in nickel metal hybrid batteries. These batteries are much cheaper to make than earlier nickel battery types, and have an energy density almost double that of lead-acid. NiMH batteries can accept three times as many charge cycles as leadacid, and work better in cold weather. NiMH batteries have proven effective in laptop computers, cellular phones, and video cameras. NiMH batteries can power an electric vehicle for over 100 miles, and are still several times more expensive than lead-acid. NiMH Batteries from Energy Conversion Devices were installed in GM’s EVI and S10 electric pickup truck, doubling the range of each. Chrysler has also

66


used NiMH batteries, made by SAFT of France its Electric Powered Interurban Commuter (EPIC) vans, adding 30 miles to their range. Other battery technologies include sodium-sulfur which was used in early Ford EVs, and zinc-air. Zinc appeared in GM’s failed Electrovette EV in the late 1970s. Zinc-air batteries have been promoted by a number of companies, including Israel’s Electric Fuel, Ltd. Zinc is inexpensive and these batteries have six times the energy density of lead-acid. A car with zinc-air batteries could deliver a 400 mile range, but the German postal service demonstrated that these batteries cannot be conventionally recharged. Other battery types are more promising, including lithium-ion, which is used in a variety of consumer products. Lithium batteries could offer high energy density, long cycle life, and the ability to work in different temperatures. However, like the sodium-sulfur batteries in the Ford Ecostar, lithium-ion presents a fire hazard since lithium itself is reactive. Plastic lithium batteries could prove to be very versatile. Bellcore is working on a lithium battery that would be thin and bendable like a credit card for laptop computers and cell phones. Each cell is only a millimeter thick. The plastic batteries are lightweight and have been tested for automotive applications. Canadian utility Hydro-Quebec has been working with 3M on a lithium-polymer unit which could be the first dry electric vehicle battery. Like the Bellcore product, this dry battery uses a sheet of polymer plastic in place of a liquid electrolyte. Also working on this technology is a team at Johns Hopkins University. This is also a plastic battery that can be formed into thin, bendable sheets. These batteries also contain no dangerous heavy metals and are easily recycled.

BATTERY CONSTRUCTION AND OPERATION A battery is made up of one or more cells. Each cell contains alternating negative and positive plates. Between these are the plate separators which are insulators. The negative plates are connected together, as are all the positive plates. Each plate has a grid-like frame, and on the grid is the plate’s active material. The grid provides the physical structure for the plate. The active material is a substance that produces the electron flow. When a lead-acid battery is fully charged, the active material in the


67

negative plates is mostly sponge lead. In the positive plates, it is lead dioxide. The plates are in contact with a solution of sulfuric acid which acts as the electrolyte. As the battery discharges, the acid from the electrolyte combines with the active material in the battery plates, forming a solution of lead sulphate and water. The water dilutes the acid solution. As the battery is charged, water is removed from the acid solution, increasing the strength of the electrolyte. A portion of the plate material is used to formed the lead sulphate. The chemical reaction is as follows: PbO2 + Pb + 2H2 SO4 < 2PbSO4 + 2H2O A charged lead-acid cell will produce a voltage of close to 2.2 volts. This is the voltage that results from the lead dioxide and lead in the sulfuric acid. Cells made of other metals and electrolytes have different voltages. The battery capacity depends on the size of the cell. The more lead dioxide and lead paste available, the greater the storage capacity. A larger, thicker plate will produce the same voltage as a smaller, thinner plate, but it will provide more electrons resulting in a greater current flow. The plates are made porous so that the acid can filter or diffuse through the plate.

CHARGING When a battery is charged, the charging current will first reconvert to active material the lead sulphates most accessible to the electrolyte. This takes place relatively quickly. Then the rate of charge is limited as acid filters out of the active material and water filters in. During charging and especially in the final stage of charging, some of the water in the electrolyte breaks down into its component parts of hydrogen and oxygen. These bubble out of the electrolyte lowering its level. This water must be replaced by filling the battery cells with distilled water through the vent caps. In many newer batteries this water loss is prevented by a catalyst that forces the hydrogen and oxygen to recombine into water. These newer batteries are sealed. Some batteries use a catalyst with vent caps. The battery plates in lead wet cells have small quantities of antimony added to the lead. This strengthens the grid and helps to lock the active material in the grids to reduce shedding. But antimony also has the effect of promoting small internal galvanic currents in the battery.

68


This slowly discharges the battery. The process is known as self-discharge. Antimony also increases gassing during charging. Some batteries are allowed to build up a certain amount of internal pressure. Under this pressure, small amounts of hydrogen and oxygen are produced during charging which recombine into water. These batteries are sometimes called recombinant. Excessive charging will cause excessive amounts of hydrogen and oxygen to be produced, so recombinant batteries use pressure relief valves to vent excess gases. GEL-CELLS In a gel-cell, the electrolyte is in the form of a gel with the consistency of soft wax. During manufacture this gel is pasted on the battery plates and separators. The active material in the battery reacts with the gel, but there is not the same fluid movement in the electrolyte as in a wet cell. The battery plates in a gel cell are relatively thin to allow diffusion around them. The gel cannot be replaced during service and gelled batteries are built as sealed no-maintenance units. It is important to prevent gassing during charging, since this causes the electrolyte to dry out and the battery will fail. Several methods are used to prevent gassing. The charge voltage is controlled to prevent overcharging. The antimony used to reinforce conventional battery plates is replaced with calcium. The resulting plate grid is not as strong, but it is not so prone to gassing or self-discharge. Gel batteries are sometimes SVRs (sealed valve regulated) batteries. Some maintenance-free batteries are wet batteries with excess electrolyte contained in partially sealed cases. The excess electrolyte is slowly used up. A true no-maintenance battery is an SVR or recombinant. A starved or absorbed electrolyte battery is an SVR with even less electrolyte. These are sometimes called AGM (absorbed glass mat) batteries.

BATTERY FAILURE One of the main causes of battery failure is shedding of the active plate material. When a battery is discharged and charged, the chemical


69

process in the plates, reconverting sponge lead and lead dioxide to lead sulphate, tend to weaken the bond between the active material and the plate grids. Each time a battery is discharged some active material is loosened and shed from the grid. This is a normal process of aging and will eventually result in failure of the battery. The material builds up in the base of the battery until it reaches the level of the plates and causes a short circuit or the loss of the active material from the plates. This causes the battery to no longer have enough capacity. Thin-plate, low-density batteries are more susceptible to damage than thick-plate, high-density units. Shedding is accelerated by deep discharges, high rates of discharge and charge, and by gassing during an overcharge, which washes material out of the plates. Gel-cells shedding can occur and the material tends to be trapped in the gel rather than falling to the base of the battery. All wet-cell batteries tend to gas as they approach full charge. Explosive gases are given off and the corrosive vapors are vented from the filler caps. The battery will use water, requiring regular refills. Gelcells minimize these problems, but they have drawbacks of their own. Newer batteries are built with envelope plate separators. These are sealed on the sides and bottom, so any shed material remains within the envelope. This reduces plate shorting, but not the shedding.

SULPHATION The lead sulphate formed in the battery plates during discharges is initially soft but if the battery is left in a discharged state, the sulphate hardens into crystals that are more difficult to reconvert into active material. This reduces battery capacity and is known as sulphation. Sulphation can occur from leaving a battery discharged or from undercharging. This means a percentage of the battery’s active material is always left uncharged. Failing to charge the inner areas of thick plates or plates with dense active material will result in sulphation. Idle batteries will slowly self-discharge and sulphate over time. Gel-cells have a slower rate of sulphation. Short-term high loads pull the charge off the battery plate surfaces. Long-term lower loads drain a battery steadily over a number of hours, allowing it time to stabilize internally, which drains the charge from less

70


accessible plate areas as well as the surface areas. When it is time to recharge, these more inaccessible inner plate areas must also be recharged. This requires enough time for the electrolyte to diffuse. If charging times are limited, the battery will not be full charged. Some of the lead sulphates formed in the inner plate areas when the battery was discharged will remain. These can slowly crystallize and charging will not reconvert them to active plate material. The battery then becomes sulphated. Increasing the plate thickness and density increases the odds of damage from sulphation. A problem with wet-cell batteries is the antimony used to reinforce the plate grids. It causes small discharge currents in the battery. These currents will slowly discharge an unused battery. If the battery is not regularly recharged, the lead sulphates forming in the plates will harden, causing a loss of capacity. To minimize damage from sulphation, a wet-cell, deep-cycle battery should be periodically (at least monthly) returned to a full charge. If it has been heavily discharged during the month, a controlled overcharge called equalization or conditioning should be used to soften up hardened sulphates. This is done by charging the battery at 3% to 5% of its rated amp-hour capacity (3 to 5 amps for a 100-Ah battery). The battery voltage should be between 15.0 and 16.2 volts for a 12-volt battery. During this charge the battery must be isolated from all loads. Equalization generally requires several hours.

OVERCHARGING Overcharging can be damaging since it leads to gassing. This results in water loss and if the lost water is not replaced, the plates will dry out. Gel-cells have less electrolyte than wet-cells and there is no way to replace the lost electrolyte. During overcharges, galvanic activity in the battery attacks the positive plate grids, causing them to deteriorate. The grids in thin-plate batteries fail sooner than those in thick-plate batteries. A battery has an internal resistance that increases as the battery is fully charged. The more current driven through it, the warmer it gets. This can cause the battery to heat up internally until the plates distort, shorting out adjoining plates. Automotive batteries are designed for engine starting. These bat-


71

teries have many thin plates with low-density material. This maximizes the plate surface area and minimizes the diffusion time of acid through the plates. These thin plates and the low-density active material cannot handle repeated deep discharges and recharges (cycling). In each cycle some of the active material falls out of the plate grids. This is not a big concern in automotive cranking since the batteries are normally discharged only a few percent and little shedding occurs. But, if repeated deep discharges occur automotive batteries can quickly fail. A poorly constructed automotive battery can fail in as few as a dozen complete discharge/recharge cycles. Better-quality ones may survive more than 30 to 40 deep cycles. Deep-cycle batteries have much thicker plates, stronger grids, denser active material and heavier plate separators. There is still some shedding of active material from the plates with every discharge cycle, but not as much as from thin-plate batteries. Some high-quality automotive batteries are built like deep-cycle batteries.

LIFE CYCLES The life cycles of a battery are the number of times a battery can be pulled down to a certain level of discharge, and then recharged, before it fails. Some manufacturers estimate life cycles using a 50% discharge and recharge cycle. Others use an 80% or even 100% discharge/recharge cycle. The greater the depth of discharge in each cycle, the fewer life cycles the battery has and the shorter the life expectancy of the battery. If one battery has the same number of life cycles at an 80% discharge as another has at a 50% level of discharge, the former is preferred, it will have more life cycles at a 50% discharge than the latter. Another factor is the manufacturer’s definition of failure. This may be defined when the shedding of active plate material reduces the battery’s capacity to 80% of its original value or when it has fallen to as little as 60% of original capacity, which will increase the number of life cycles. If a battery is to be used for loads similar to engine cranking (motor loads) and immediately recharged, this directly parallels automotive use and an automotive cranking-type battery is adequate. For other applications, the batteries will usually be cycled at some

72


time, especially batteries that are to be used for emergency power as opposed to engine-cranking service. These should be good-quality, deep-cycle batteries. Prevailer gel-cells are known to stand up to heavy cycling in marine diesel use. Prevailers are manufactured under license in the USA and their cost is comparable with top-quality wet-cells. Their success has spawned clones such as Dynasty and Lifeline batteries. Premium deep-cycle batteries include the Surrette (Tilton, NH) and Rolls (Salem, MA) wet-cells and the Prevailer gel-cells (Lyon Station, PA). In the UK, Lucas/Yuasa makes premium traction batteries, while Prevailers are sold by FWO Bauch under the Sportline DryFit label. Wet-cell deep-cycle batteries acquire their long cycling life from heavy, antimony-reinforced plate grids. These are thick plates with highdensity active material and multiple plate separators. If properly cared for, these batteries may be cycled thousands of times. These batteries are several times more expensive than an automotive battery and they also suffer with a loss of performance in other respects. The thick plates and dense active material slow the rate of acid diffusion through the battery and thus the rate at which a charge can be withdrawn or replaced. When these batteries are under a high load, such as motor or cranking loads, the battery voltage tends to drop off. This voltage drop may be critical with DC to AC inverter use depending on the inverter. Some loads will suffer a serious loss of performance as battery voltage declines.

BATTERY CAPACITY RATINGS A battery is in a constant state of chemical activity and is affected by temperature changes, aging, humidity, and current demands. The traditional measure of a battery’s ability to do work is its ampere-hour (Ah) capacity. The battery is discharged at a constant rate for 20 hours so that the potential of each cell drops to 1.75-V. A battery that will deliver 6-A over the 20 hour period is rated at 120-Ah (6-A X 20 hr). A 120-A battery will not deliver 120-A for 1 hr. The amp-hour rating defines the capacity in the number of amphours available from a battery at 80°F (26.7°C), at a relatively slow rate of discharge. In the USA, the discharge period is normally 20 hours while in the UK, it is 10 hours.


73

This means that a battery rated at 100-Ah can deliver 5 amps for 20 hours in the USA, called the C20 rate, or 10 amps for 10 hours in the UK, called the C10 rate. Shorter time periods are also used, 5 hours is used on deep-cycle (traction) batteries, called the C5 rate. Here, a 100-Ah battery would be able to deliver 20 amps for 5 hours. A battery has a lower Ah capacity at higher rates of discharge. If a USA-rated 100-Ah battery (C20 rate) is discharged at a rate of 10 amps, (C10 rate) it will deliver about 85-Ah before its voltage drops below the threshold level. A UK-rated 100-Ah battery (C10 rate) can deliver a full 100-Ah at this rate. When comparing battery ratings, the same rating period should be used. Battery capacity ratings also include reserve capacity and cold cranking amps. Reserve capacity (or minutes) is an automotive industry rating. The reserve capacity indicates how long, in minutes, a battery at a temperature of 80°F (26.7°C) will support a specified load before its voltage drops below 1.75 volts per cell (10.5 volts for a 12-volt battery). The normal rating load is 25 amps, but other amperages are used. Battery comparisons should use the same load. Cold cranking amps (CCA) rate the engine cranking ability. As the temperature drops, it takes more energy to turn an engine, while at the same time the battery has less available power. The standard USA coldcranking rating is at 0°F (-17.8°C) for a time of 30 seconds. It indicates the maximum discharge rate in amps the battery can deliver to the starter motor without dropping below 1.2 volts per cell (7.2 volts for a 12-volt battery). There is also a marine cold cranking amps, which uses a higher temperature of 32°F (0°C). This is more than the CCA rating. In the UK, the British Standard Rate uses a temperature of 0°F (-17.8°C) with a time period of 180 seconds. The final terminal voltage is 1.0 volt per cell. The International Electrotechnical Commission rating is at the same temperature with a time period of 60 seconds and a final terminal voltage of 1.4 volts per cell. Zero cranking power is a hybrid measurement expressed in volts and minutes. The battery is chilled to 0°F; depending on battery size, a 150- or 300-A load is applied. After 5 seconds the voltage is read for the first part of the rating. Discharge continues until the terminal voltage drops to 5-V. The time in minutes between full charge and effective exhaustion is the second digit in the rating.

74


HYDROMETER TESTING As the battery discharges, some of the sulfuric acid in the electrolyte decomposes into water. The strength of the electroltye in the cells is an index of the state of charge. The measurement of specific gravity is done with a hydrometer. It consists of a rubber bulb, a barrel, and a float with a graduated scale. The graduations are in terms of specific gravity. Water is assigned a specific gravity of 1. Pure sulfuric acid is 1.83 times heavier than water and thus has a specific gravity of 1.83. The height of the float above the liquid level is a function of fluid density, or specific gravity. The battery is said to be fully charged when the specific gravity is between 1.250 and 1.280. Water should be added several operating days before the test to ensure good mixing. Use a hydrometer reserved for battery testing. Do not use one that has been used as an antifreeze tester. Trace quantities of ethylene glycol will shorten the battery’s life. American hydrometers are calibrated to be accurate at 80°F. For each 10° above 80°F, add 4 points (0.004) to the reading and for each 10°F below the standard, subject 4 points. The standard temperature for European and Japanese hydrometers is 20°C, or 68°F. For each 10°C increase add 7 points (0.007) and subtract a like amount for each 10°C decrease. Some hydrometers have a built-in thermometer and correction scale. All cells should read within 50 points (0.050) of each other. Greater variation is a sign of abnormality. The state of charge is only indirectly related to the actual output of the battery. Chemically the battery might have full potential, but unless this potential passes through the straps and terminals, it is of little use. A more reliable test is to load the battery with a resistive load while monitoring the terminal voltage. The battery should be brought up to full charge before the test. The current draw should be adjusted to three times the ampere-hour rating. A 120-Ah battery would be discharged at a rate of 360A for 15 seconds. The terminal voltage should not drop below 9.5-V. KINETIC RIDE-THROUGH Kinetic energy storage can be used as a ride-through technology. It has been used for many years in motor-generator sets for load isolation and conditioning.


75

Some power protection systems combine slow-speed conventional flywheels with other energy storage and generation technologies. These include synchronous generators, diesel or natural gas engines with flywheels that rotate at less than 6,000 rpm. Direct circuit (DC) system flywheel energy storage technology can be used as a substitute for batteries for providing backup power to an interruptible power supply (UPS) system. The initial cost may be higher, but flywheels offer a much longer life, reduced maintenance, a smaller footprint, and better reliability compared to batteries. The combination of these characteristics generally results in a lower life-cycle cost for a flywheel compared to batteries. Flywheels designed for UPS application typically provide power for about 15 seconds. Using a flywheel instead of batteries requires a generator that can come up to full power in about 10 seconds. Flywheels can also be used to extend battery life. A majority of power events last 5 seconds or less. A flywheel can be added to a battery backup system and controlled so that the flywheel provides power for short-duration events while the battery is used for longer outages. Flywheels are highly tolerant to frequent cycling while batteries are not and batteries can provide power for a longer period. Batteries for UPS applications are typically sized for about 15 minutes of full load power. Many UPS systems are integrated with diesel-fired generators that can come up to full power within 10 seconds. In kinetic energy storage, energy is stored by causing a disk or rotor to spin on its axis. The stored energy is proportional to the flywheel’s mass moment of inertia and the square of its rotational speed. Motor generators use flywheels to isolate electric loads from electricity supply disturbances. These flywheels limit disturbances to less than 1 second of backup power and only utilize a few percent of the flywheel’s stored energy. A more effective use of flywheel technology requires a means of disconnecting the kinetic energy stored in the rotating mass from the energy demands of the load. The addition of a variable speed drive and inverter components provided a solution, but it was initially cumbersome and expensive. Advances in electronic power conversion and control technology, coupled with resourceful integration of the components, have been key to the technology’s development. The superior energy storage density of flywheels compared to batteries has always been recognized. Much of the flywheel develop-

76


ment in recent times has been in space vehicles and satellites, where mass and volume constraints are severe. Flywheel development in these applications has focused on increasing the rotor speed to maximize the energy density. These speeds require the use of magnetic bearings, which are now used in moderate rpm flywheels as well. Flywheel systems must have low frictional losses. Magnetic bearings are one solution since they allow a flywheel’s rotating shaft to float, with no surface-to-surface contact. Hybrid magnetic bearings use permanent magnets and provide stability with much smaller and more energy efficient electronic components. Alternatives to magnetic bearings include ceramic bearings that provide nearly frictionless rotation, in spite of the surface-to-surface contact. DC flywheel energy storage systems are an alternative or supplement to lead-acid batteries. Batteries have the advantage of providing backup power for a period measured in minutes rather than seconds, but this advantage has limited value if reliable backup generators are available. Batteries will usually have lower first costs, but their significantly shorter life and greater maintenance requirements compared to flywheels generally gives flywheels lower life-cycle costs. Standby losses for flywheels range from about 0.1% to 1.0% of rated power. This includes power to overcome frictional losses and auxiliary equipment. The life of a flywheel is typically about 20 years, while most batteries in UPS applications will only last 3 to 5 years. Batteries must also be kept at a narrow operating temperature while flywheels are tolerant of outdoor ambient temperature conditions. Frequent cycling has little impact on flywheel life, while frequent cycling significantly reduces battery life. Flywheel maintenance is generally less frequent and less complicated than for a battery. Flywheel reliability is 5 to 10 times greater than a single battery string or about equal to two battery strings operating in parallel. Flywheels avoid battery safety issues associated with chemical release. They are also more compact, using only about 10 to 20% of the space required to provide the same power output from a battery. With a much higher power density than batteries, typically by a factor of 5 to 10, flywheels are more attractive where floor space is expensive. Batteries usually have a lower first cost than flywheels, but suffer from a shorter equipment life and higher annual operation and maintenance expenses. Flywheels are attractive in operating environments that


77

are detrimental to battery life such as frequent cycling due to main power supply problems. Flywheels can be classed as low speed or high speed. The low range is in thousands of revolutions per minute (rpm) measured while the high range is tens of thousands of rpm. Doubling the rpm quadruples the stored energy, so increasing rpm considerably increases the energy density of a flywheel. Low-speed flywheels are usually made from steel while highspeed flywheels are usually made from carbon or carbon and fiberglass composites that will withstand the higher stresses associated with higher rpm. Higher rpm also creates greater problems with friction losses from bearings and air drag. High-speed flywheels typically use magnetic bearings and vacuum enclosures to reduce or eliminate these sources of friction. Magnetic bearings allow the flywheel to levitate, fundamentally eliminating frictional losses associated with conventional bearings. Some low-speed flywheel use only conventional mechanical bearings but most flywheels use a combination of the two bearing types. Vacuums are also used in some low-speed flywheels.

INSTALLATION Most flywheels are attached to a concrete slab and connected to the DC bus of the UPS system with a DC disconnect switch to allow servicing. A 120-V AC service is required for most flywheel systems auxiliary equipment, such as vacuum pumps. Some flywheels require a higher AC voltage for recharging. A DC flywheel system must have an output voltage that matches the UPS system DC bus voltage. DC flywheel energy storage can be applied anywhere a battery is currently used to provide backup power for a UPS system. The flywheel can be used as a substitute or supplement for the battery. UPS batteries are sized to provide backup power for periods from about 5 minutes up to around 1 hour, but more commonly about 15 minutes. A period of 15 minutes is generally adequate to allow an orderly shutdown of equipment. The backup period for flywheels is commonly about 15 seconds. This is enough time to allow the flywheel to handle the majority of power disruptions that last for 5 seconds or less and still have time to cover slightly longer outages until a backup generator can come up to

78


full power. Flywheels should not be used alone for backup power, without a battery and/or a fuel-fired generator.

PRODUCTS Several companies offer products where the flywheel is an integral part of the UPS system rather than being a direct substitute for a battery. Low-speed units may use unenclosed steel rotors with conventional bearings while high-speed units have composite material rotors operating in a vacuum with magnetic bearings. Intermediate products with elements of these two endpoints also exist. Further product development and enhancement is likely coupled with the ever increasing needs for more reliable, higher quality power. The focus has been on stand-alone DC flywheel energy storage systems that can substitute or supplement batteries in a UPS system. Some manufacturers offer flywheels as an integral part of a UPS system while others have developed DC flywheel energy storage systems. See Table 3-1. Material advances have benefited flywheel systems. In the last 30 years, the tensile strength of graphite composite materials has increased some five-fold, while the cost per pound has dropped over 90%. The development of Kevlar was a breakthrough in flywheel system development. Recently, T1000 graphite composites with tensile strengths of up to a million pounds have allowed flywheels to approach 100,000 rpm, significantly increasing their energy storage capacity. Magnetic materials have also seen marked improvements in recent years with the development of rare earth magnets such as samarium cobalt-17 in the mid-1970s and neodynium-iron-cobalt (NdFeB) in 1983. The latter is used extensively by the automotive industry in alternators. Materials costs for NdFeB are quite low, since neodynium is the third most common rare earth element, and cobalt is used only in small amounts.

KINETIC SYSTEMS A flywheel-coupled motor-generator is used in International Computer Power’s Dynamic Energy Storage System (DESS). This is a kinetic


79

Table 3-1. Flywheel Companies ———————————————————————————————— Manufacturer Active Power Acumentrics AFS Trinity Power Beacon Power Caterpillar Designed Power Solutions Flywheel Energy Systems GE Digital Energy Holec Indigo Energy Optimal Energy Systems Pentadyne Power Piller Powerware Preciso Power Regenerative Power & Motion Reliable Power SatCon Power Systems Statordyne Urenco Power Technologies

Products UPS, DC energy storage UPS, DC energy storage DC energy storage UPS, DC energy storage UPS UPS with generator, DC energy storage DC energy storage UPS with generator UPS with generator UPS, DC energy storage UPS, DC energy storage DC energy storage DC energy storage, UPS, UPS with generator UPS, DC energy storage UPS DC energy storage UPS UPS with generator UPS with generator DC energy storage

———————————————————————————————— storage system for system ride-through. These units can be used to replace UPS battery backup systems in standby engine generator applications. Kinetic-storage can also be used with an existing battery storage system to prevent unnecessary battery system discharging during shortduration power anomalies. The Statordyne system uses hydraulic storage with back-up generation. The system uses a synchronous motor/generator that provides power factor correction by running as a synchronous condenser while the utility is supplying power. It switches to a generator when the utility power is interrupted. Energy is stored in a mechanical flywheel that keeps the generator’s speed at about 1,777 rpm for the first 100 to 200 milliseconds. Then a hydraulic motor is engaged to keep the generator’s shaft spinning at 1,800 rpm until a diesel or natural gas generator starts and takes over. These systems range from 100 to 800-kW, and cost about $1,000/kW depending on the size and features.

80


The Holec system also uses flywheel-inertia storage with back-up generation. This system integrates low-speed flywheel technology with a synchronous generator to protect facilities from power outages 2 to 3 seconds in duration, with a diesel or natural gas engine taking over for protection from longer power outages. The Holec system is installed in about 300 facilities worldwide. The system uses a 3,600-rpm induction motor, with a rotating stator that is mechanically attached to the rotor of an 1,800-rpm motor/generator. Under normal conditions, both motors operate at full speed and the rotor of the induction motor turns at 3,600 + 1,800 rpm for an overall rotating speed of 5,400 rpm. It acts like an energy storage flywheel. When an electric power interruption occurs, the induction motor is reconfigured to operate as an eddy current clutch. It transfers energy from itself to keep the generator’s shaft at full speed. These systems are available in sizes ranging from 100 to 2,200 kVA and cost about $700 per kVA for a 1-MW system.

MAINTENANCE DC flywheel maintenance depends on the flywheel design. Although valve regulated lead acid (VRLA) batteries do not require monitoring and maintenance of electrolyte fluid levels, manufacturers generally recommend quarterly inspections to check tightness of connections, remove corrosion, measure voltages, and check for cracks and swells in the battery cases. Periodic replacement of individual batteries may be required, while replacement of the entire battery system can be expected about every 4 years. Routine maintenance for flywheels involves changing cabinet air filters and checking vacuum pump oil level every few months. The vacuum pump oil should be changed once a year. Magnetic bearings require no maintenance, while replacement of mechanical bearings is expected every 3 to 10 years, depending on the flywheel design. The vacuum pump may need replacing every 5 to 10 years. Bearing replacement for lower rpm flywheels with mechanical bearings costs $5/kW to $15kW depending on flywheel design. Replacing the vacuum pump costs about $5/kW. While most flywheels have a design life of about 20 years, they will probably last longer with regular maintenance. Flywheel backup costs vary from $100/kW to $300/kW. The lower end of the range represents larger, lower rpm units while smaller, higher


81

rpm models have higher per kW costs. Installation is about $20kW to $40kW. Standby power consumption is about $5kW/year for lower rpm flywheels with mechanical bearings. It is only about 10% of this for higher rpm flywheels using only magnetic bearings. VRLA battery purchase costs, measured in $/kWm (dollars per kilowatt-minute) are about $17/kWm for 5 minutes of backup power, and drop to about $13kWm, $10kWm, and $8kWm for 10, 20, and 30 minutes of backup power. Annual battery maintenance costs are about $3.50/kWm for 5 minutes of backup power, dropping to about $2.25/kWm, $1.50kWm, and $1.25kWm for 10, 20, and 30 minutes of backup power. Standby power consumption, or float loss, is negligible for batteries compared to flywheels, but the differences in footprint and floor-space requirements are much more. Flywheel footprints range from 0.04 ft2/kW to 0.12 ft2/kW depending on the size and type. Battery footprints per kW depend on the backup time. For 5 minutes of backup power, VRLA battery footprints start at about 0.15 ft2/kW. This doubles for 20 minutes of backup power and depends on the backup time. Electromechanical flywheels avoid the environmental and safety issues associated with battery requirements. This includes the need for eye wash stations, spill containment, hydrogen detection, and elevated room ventilation rates. Batteries must also be kept at normal indoor air temperature or suffer from degradation in expected life. This adds to cooling system operating costs. INSTALLATIONS Several hundred units have been installed with about a dozen at Federal facilities. Most of these were at military installations, but the State Department and Veterans Affairs also have installations. At Fort McPherson in Atlanta, Georgia, the U.S. Army has equipment that must be running on a 24/7 basis. Grid power was backed up by a UPS system with wet-cell batteries and diesel-fired generators. While the diesel generators would come online within 10 seconds, the existing battery system was becoming less reliable and more expensive to maintain. Annual battery maintenance was almost $30,000. The UPS system consisted of four 500-kVA units operating in parallel, each with its own wet-cell battery string. The UPS and battery system was about 15 years old, which is well beyond the expected battery life.

82


One of the battery strings had already failed and represented a hazardous condition for maintenance personnel. The batteries needed to be replaced and the options were either VRLA or wet-cell batteries, or flywheels. Reductions in power demand reduced the requirements to two 500-kVA units operating in parallel, with either unit able to meet the building’s critical power demand. It was decided to install two flywheels, one for each 500-kVA UPS unit instead of replacement batteries. A wet-cell battery system would cost approximately the same as a flywheel, but would probably have a shorter life and is expected to have higher annual maintenance costs. A VRLA battery system would cost about half as much as a flywheel, but would have to be replaced several times over the life of the flywheel and would incur higher annual maintenance costs. Using a flywheel would free the 2400 square foot room reserved for the battery strings. Room ventilation and cooling costs could also be reduced. Environmental and safety issues associated with the batteries were eliminated. The diesel generators would come online within 10 seconds, which minimizes the value of the longer backup period provided by batteries.

LIFE-CYCLE COSTS Since flywheels will cost more than batteries, but require less maintenance and last much longer, they will generally be less expensive on a life-cycle cost basis. At one facility a 250-kW UPS system is backed up by a generator that can come up to power in 10 seconds. Backup power could be provided by either a battery or a flywheel. The comparison in Table 3-2 is based on a low-rpm flywheel with a life of 20 years and a VRLA battery with a life of 4 years. Battery life will vary depending on the operating conditions, but a typical lifetime of 4 years is used. The battery is assumed to provide power for 10 minutes. A discount rate set by the National Institute for Standards and Technology for federal energy projects is used. The resulting present value of life-cycle costs is $248,129 for the battery option and only $105,572 for the flywheel option. The savings is $142,557, about 60%. The short battery life compared to flywheels results in much greater lifecycle cost for the battery option.


83

Table 3-2. Batteries Versus Flywheels ————————————————————————————————

Batteries Cost ($13/kWm × 250-kW × 10 minutes) $32,500 Installation ($30/kW × 250-kW) $7,500 Total $40,000 Replacement every 4 years $40,000 Annual maintenance ($2.25/kWm × 250-kW × 10 minutes) $5,625 Floor-space/year (0.22 ft2/kW × 250-kW × $10/ft2) $550 Standby power/year (250-kW × 8760 hrs × 0.01% × $0.063/kWh) $14

Flywheels Cost Installation Total Bearing replacement every 5 years Vacuum pump replacement every 7 years Annual maintenance Floor-space Standby power

($200/kW × 250-kW) ($30/kW × 250-kW)

$50,000 $7,500 $57,500

($10/kW × 250-kW)

$2,500

($5/kW × 250-kW) ($5/kW × 250-kW) (0.08 ft2/kW × 250-kW × $10) (250-kW × 8760 hours x 1% × $0.063 kWh)

$1,250 $1,250 $200 $1,380

———————————————————————————————— SUPERCONDUCTER STORAGE Electrical energy may be stored in a coil of superconducting wire submerged in liquid helium. The technique can be used for both ridethrough and energy storage systems. Since the late 1960s, over 20 development projects have begun, with some outstanding results. Storing large amounts of energy this way is feasible and the ability to deliver high bursts of power make it attractive for facility ride-through systems. This technology stores electrical energy without intermediate conversions to mechanical or chemical energy. A superconducting magnetic coil is immersed in liquid helium at 4.2°K (-455.5°F), causing its resistance to DC current to fall to zero. High electrical currents can be sent to the coil and the current will circulate without any losses, until it is diverted from the coil to the facility elec-

84


trical system. The very cold temperatures used require continuous cooling. The energy used by the cooling systems is about 25-kW for a moderately sized unit that stores about 0.28-kWh. Superconductivity, Inc., has been shipping commercial systems since 1992 with over a dozen ride-through systems currently installed. Systems are available for about $1,000 per kW. The company is now part of American Superconductor, Inc. located in Westborough, MA.

UNINTERRUPTIBLE POWER SUPPLIES (UPS) Any strategy for safeguarding against power interruptions or poor power quality should include an uninterruptible power supplies (UPS). These are one of the best power-quality defense tools that one has. An UPS can monitor and regulate the utility power entering a facility. An UPS is used for critical loads when continuous power is required and can protect the load from system power disturbances such as harmonics, transients, voltage surges and sags. The quality of service and frequency of utility power failure is important. This depends on the utility service company, the geographical isolation or location and the time of the year. It is useful to obtain data for the average duration as well as the frequency of the outages. The reliability of the utility service and the duration of the average power outage will determine the type and size of the standby power. If the quality of utility power is not always satisfactory, an UPS can improve this situation as well. A useful step is to classify the loads into functional categories depending if the load is supporting staff, equipment or building functions. Loads associated with human safety such as life support systems in a hospital, air-traffic, police and fire control systems require an UPS. Systems for data processing or critical industrial processes may need to operate under any conditions. An UPS can help ensure that critical data and telecommunications systems function properly and consistently. Building support systems include lighting, HVAC, elevators, communication systems, security systems and fire alarm. The last two systems will require battery backup or UPS. While the others will typically use batteries or emergency generators for backup. Batteries are often used for smaller loads such as emergency lighting, alarms, control circuits, telephones, fire protection systems and security alarms. For larger


85

loads such as HVAC, elevators and other essential devices, a standby generator is required. In case of power anomalies, such as surges, an UPS may have transient voltage surge-suppression circuitry to prevent equipment damage. In the event of brownouts and blackouts, the UPS takes over to keep systems running for a period of time. When the UPS’s battery begins to be drained, before regaining utility power, the UPS’s software can save open files and shut down critical systems. UPS systems may be rotary or static. A rotary system uses a motorgenerator set to isolate the critical load from normal power. During an interruption, the motor-generator set will continue to provide power to the load for 100 milliseconds or more with its kinetic energy. Flywheels will extend this to many seconds. A bypass circuit allows the system to operate with normal power if the UPS malfunctions. Transfer to the bypass circuit may be manual or automatic. To protect against data loss due to interruptions or fluctuations in commercial power, many critical use areas use a static UPS on all computers and attached equipment. An UPS is a specially-designed power supply, with batteries that can supply power to the computer for short periods. If the commercial power fails, the computer runs off the UPS batteries until it can be shut down properly. A static UPS uses a rectifier and inverter circuit along with batteries. The rectifier circuit converts AC power to DC which charges the batteries and supplies the inverter. The inverter converts DC power back to AC. During an interruption, the batteries continue to supply the inverter until they are discharged or normal power is restored. Normally, the batteries are large enough to last a minimum of 20 minutes. To enhance the protection offered by a UPS, many software versions include a built-in UPS monitoring feature. This feature provides an interface between the UPS unit and the operating system. Through this interface, the UPS can signal its status to the operating system and thereby coordinate an orderly shutdown. The UPS can provide more complete power protection but not all UPS systems are alike. Early UPS units were designed simply to provide backup power in the event of a blackout, leaving the equipment exposed to other types of power irregularities. Newer UPS systems are designed to protect against a full range of power problems. UPS systems can be grouped into offline, online, and hybrid. The newer types act as an intelligent power supply. A standby or offline UPS

86


operates by switching from commercial power to battery power when the commercial power drops below a certain voltage level. The inverter is not powered, except during power outages. Since the load receives power directly from the utility line as long as it is available, the load remains exposed to power disturbances such as spikes, noise, surges, and brownouts. Because of this, these systems should be used with a line conditioner or voltage regulator. Another disadvantage is the time it takes for the unit to switch over after it senses that the voltage has dropped. Some sensitive equipment may not be able to ride through this transition period without being affected. A switching time of four milliseconds or less is needed for electronic loads such as computer or control hardware. The major advantage of an offline UPS is its low cost.

UPS OPERATION An online UPS acts as alternate source of power. It continuously converts commercial AC power to DC charging power to keep a battery charged. Its inverter uses the DC power from the battery to create new, clean AC power. Critical loads run off the power generated by the online UPS at all times. Since the online UPS uses the commercial line voltage to keep its batteries charged, but never to supply power directly to the load, there is complete isolation of the load from the commercial power line and any power fluctuations. Also, there is no switching of the load between commercial and battery power. A hybrid UPS may use an offline UPS with electronic or ferroresonant conditioning/filtering to smooth the transition from utility to inverter power. The quality of output power from a hybrid UPS depends on its filtering and conditioning capabilities. Many of these systems are advertised as line interactive, no break, load sharing and bidirectional. Sometimes hybrids are even labeled as online products although they do not function the same as online UPS units. Hybrid and offline UPS units do not regenerate power continuously so the load either receives raw or partially filtered utility power. Sensitive equipment may suffer from potential damage from this raw or partially filtered utility power. Many UPSs are specifically designed to protect computer systems. They add software capabilities that work with the computer or network


87

operating system. The interface to the operating system is more than a status indicator of power or battery charge. Built into the software are UPS monitoring functions to provide the network or workstation with an orderly, automatic, and unattended shutdown. A true regenerative, online, sinewave UPS provides the most complete protection while maintaining a solid output during any power interruption. These type of systems are the most expensive and may not be able to handle all the loads.

SPECIFYING AN UPS UPS units come in a variety of sizes and price levels. Units can range from a $100 to many thousands of dollars, depending on whether you need to control one PC/workstation or an entire data center. The most basic UPS can protect a single PC or a smaller workstation. It provides 200-650 VA and has an average running time of 6 to 12 minutes. Larger systems, that are capable of running an entire data center, can operate for a half hour or more and can have a rating from 1,000 to 5,000 VA. There are also different types of UPS technologies to consider. These include online, line-interactive and standby (offline) units. Each of the UPS technologies keeps its output between +10% and -20% for 120 volt outputs and most have a UL 1449 listing for urge-suppression protection. The most common UPSs for computer equipment including servers and networks are standby, line interactive, standby online, hybrid, standby-ferro, double conversion online, and delta conversion online. The standby UPS system is used for personal computers. In the standby UPS system, when the primary AC source fails, the transfer switch moves from the AC line to a battery/inverter system for power. Standby systems will only start during a power failure. These are battery/inverter power units characterized by high efficiency, small size and relatively low cost. Standby UPS units switch from utility to battery power when the utility power falls below a certain voltage, the unit then begins drawing power from the UPS’s internal battery. The technology has advanced in the past few years and most standby UPSs switch over in 2 to 10 milliseconds. This is usually fast enough to prevent any damage. Online UPSs use batteries as a power buffer to absorb spikes, surges and sags on a full-time basis. They convert the line AC to low-

88


voltage DC to charge the battery and then covert it back to AC to operate critical systems. The online UPS operates constantly, so the power to critical systems is uninterrupted, even in power failure. While this is very useful, it can be costly.

LINE INTERACTIVE Line-interactive UPSs constantly monitor the quality of your utility power. If required they make the necessary adjustments to raise or reduce the line voltage. A line-interactive UPS acts as a power conditioner for sensitive systems. The initial and daily operating costs are less than online systems, but more expensive than standby units. The line interactive UPS is commonly used for small business computer equipment including Web processing and departmental servers. It’s battery-to-AC power converter (inverter) is always connected to the output of the UPS. When the AC power is functioning normally, the battery is charging as the inverter is operating in reverse. When the input power fails, the transfer switch functions and power flows from the battery to the UPS output. Unlike the standby UPS, the inverter in the line interactive UPS is always connected to the output. Since the inverter is always connected to the output, additional filtering is used to provide lower switching transients. In a line interactive UPS, the inverter also provides some line regulation, to correct brownout conditions. The switch to battery operation is not needed when the UPS is used at sites subject to brownouts. The problem of a single point failure in the UPS can be eliminated with this system. If the inverter fails, power will still flow from the AC input to the output since this design provides two independent power paths. It is an efficient, reliable system that provides power protection. High efficiency, low cost, and high reliability along with the ability to correct low or high line voltage conditions makes the line interactive UPS a prime choice in the 0.5-5 kVa power range. STANDBY ONLINE HYBRID The most used common UPS system for under 10-kVa is the standby online hybrid. Although it is called an online system, it is not


89

a true online system. The power path from the battery to the output is online, (the inverter half), but the DC-DC converter half, is operated in the standby mode. The standby converter from the battery is switched on when there is a power failure. The switch to standby power is almost instantaneous, so there is virtually no transfer time.

STANDBY FERRO-RESONANT The Standby Ferro-Resonant UPS, also called a Standby Ferro UPS, operates in the 3-15 kVa load range. The transformer in this unit has three windings for power connections. The primary power path is from AC input, through a transfer switch, through the transformer, and to the output. In the event of a power failure, the transfer switch opens and the inverter takes over the output load. The inverter, which is normally in the standby mode, becomes active when the input power fails and the transfer switch opens. The transformer provides some regulation and output waveform shaping. Standby-Ferro UPS systems are sometimes called online units, however they are not true online units. They have a transfer switch, but the inverter operates in the standby mode and, during a power failure, they go into a transfer mode. Standby Ferro systems are reliable and provide good line filtering, but they can also be inefficient.

DOUBLE CONVERSION ONLINE This is the most common UPS for loads above 10-kVa. It is the main protection for servers, networking equipment and data centers. The primary power path of the Double Conversion Online UPS is the inverter. In the Standby UPS, the main power path is the AC line power. In a double conversion online UPS, incoming AC power is converted to DC power and the DC power is converted to AC power to the load. During normal operation, the batteries on the DC bus are charged.

90


When there is an AC power failure, the batteries run the inverter and isolate the load. Since both the battery charger and the inverter convert the entire load, more heat results which reduces the efficiency in this type of UPS. The extra heat in the power components reduces the reliability over other types and the extra energy consumed results in less efficiency. This contributes to the increased life-cycle costs of this design. The input power required by the large battery charger is often non-linear and can cause problems in power quality.

DELTA CONVERSION ONLINE The delta conversion online UPS is available for loads of 5 kVA or more. In this type of UPS, the inverter supplies the load voltage as in the double conversion UPS. But instead of the rectifier charging the batteries as in the double conversion UPS, the delta conversion uses bi-directional converters that are connected to a battery. These convert AC power to DC power and back to AC with minimal energy loss. The converters are connected in series between the power source and the load. They also compensate for differences between the required output voltage and the utility voltage. Harmonic distortion is reduced and energy efficiency is increased. The delta conversion online UPS has the same output characteristics as the double conversion online UPS. But, the delta conversion reduces energy losses and costs. The input power quality of the delta conversion UPS is also better especially in the larger kVa sizes. Many UPS systems offer some protection against spikes, surges, and sags, but they disregard brownouts. Offline and other non-regenerative UPS units do not perform well during sustained low voltage conditions. Switching times for these UPS units tend to increase as the utility voltage decreases. A unit with a 5-millisecond transfer time at 120 VAC may exceed twenty milliseconds at 100 VAC. A brief period of low voltage precedes most blackouts and this places the load at greater risk. Offline and hybrid UPS units may also sense a brownout as a blackout and prematurely switch to battery. During a sustained brownout, an offline UPS can discharge the battery and lose power even though the utility power is still on.


91

OUTPUT WAVEFORMS UPS units are available with either sinewave or squarewave outputs. These are often modified or approximated in some way. Sinewave is usually considered best since it is the same waveform provided by the utility companies and is the waveform that most equipment is designed for. Sinewave is better since some hardware may be affected by both linear or RMS currents and nonlinear or peak currents. A squarewave output only approximates a sinusoidal waveform and puts stress on RMS-sensitive system elements, while starving peaksensitive elements. This can cause excessive heating and hardware failures. The excess energy in squarewaves in the form of harmonics can affect electronic circuitry and cause data errors. Computer grade sinewaves try to use enough squarewave increments to closely approximate a sinewave. This usually provides the needed RMS voltage and limits the peak voltages to equipment design values. But, these apparent sinewaves may not provide the precise timing needed by many computer monitors. The timing mechanism is referenced to the zero crossing of the utility sinewave. It regulates the scanning of the monitor. The imprecise screen scans from the zero crossing problem result in the screen appearing to swim or undulate. The waveform from the UPS must also be a pure alternating current with no DC component. Even a small percentage of direct current in the output can saturate magnetic loads such as fans and transformers and make them inoperative.

RATINGS The UPS must be able to start up all connected loads at a critical time. When the loads are energized again after a power anomaly, the initial inrush current is much higher than the normal operating power requirement. If additional equipment connected to the UPS is powered on when the demand for UPS power is already high, an overload condition may occur. The additional devices may not receive power, or other equipment attached to the UPS will be shut down. Online UPS devices may require an inrush capacity as high as 1000%. Online UPS units with inadequate inrush capacity cannot be used to power equipment up to the rated output of the UPS.

92


UPS units must have a power rating sufficient to support all of the loads that will require power from the unit. Power ratings are usually specified in terms of volt-amps (VA) or watts. Extended operation on the UPS batteries in the absence of commercial power will shorten the life of the batteries. Most UPS systems have a maximum time before the unit switches back to commercial power. The increasing need for reliable power quality in today’s industrial and commercial facilities has resulted in the availability of a wide variety of power quality migration equipment, among them static and rotary UPS systems. The level of power protection provided by these systems depends on the maintenance of electrochemical storage batteries.

RELIABILITY A standby power source must be reliable in order to function properly. All of these systems are made up of many parts and the reliability of the individual parts make up the overall system reliability. A robust system is characterized by simplicity, so reducing the number of parts improves reliability. Some of the larger electronic chips will have a major impact on system performance. In a standby generator, the transfer switch has a critical role since it must operate properly during normal and emergency conditions. If the transfer switch fails, it could transfer to generator power in error, causing a short power interruption. If it fails to transfer during a power failure, there will be no backup power. If the anticipated utility outages are longer than the capacity of UPS batteries, then an auxiliary generator must be used to charge the batteries. Many power conditioning problems are interactions with load devices. A solid-state UPS may not synchronize with the load it is carrying and passes the load off to the bypass line. A standby generator cannot be used to power even an unloaded UPS, because the current demands of the UPS rectifier are so distorted that the generator must be several times the size of the UPS to provide the correct energy. These problems can often be explained by harmonic distortion demands. Most loads are designed to operate with 60-Hz electrical power, but under some conditions the incoming service, transformers, busways, panel boards, branch circuits, and load distribution devices may demand several frequencies due to harmonics.


93

An UPS may be installed in order to keep all loads running even when the power company has an anomaly in their service. The UPS supplies 60-Hz power, which is supported by batteries for the time needed. However, suppose the loads also need 180-Hz, 300-Hz, 420-Hz, and other frequencies. Newer loads, for both single phase and three phase equipment, may make demands upon the power source for a mixture of frequencies for the current they take to operate. The higher frequencies above 60-Hz are a part of the current spectrum which characterizes these non-linear devices. These extra frequencies cause heating in the wiring and electrical distribution equipment, lowering the efficiency of the system in handling energy. In a three phase load, these extra currents may reduce the power factor by 10 to 15%. This creates an unwanted peak energy demand for the system. If enough of these devices, such as variable speed drives (VSDs) are installed on a system, the resultant harmonic currents could be large enough to cause voltage distortion in excess of the amount which could be handled by the power supply. In single phase loads, such as those caused by personal computers or electronic lighting ballasts, the loss of capacity starts on the common neutral wire of three phase, four wire distribution systems. The overloading of the neutral occurs due to the high content of 3rd harmonic, 180-Hz, which does not cancel when each of the phase contributions arrives on the common neutral. A 750-kVA distribution transformer, 480 volts to 208/120 volts, may only carry 100 amperes on each of the secondary phase wires when 1/2 loaded. But, harmonic currents can increase the neutral current into the transformer to over 200 amperes. The current in the neutral should cancel, except for the unbalanced portion at 60-Hz. New drive equipment often lowers the power factor. But, adding power factor correction capacitors to improve the power factor and avoid penalty billing may cause the circuit breaker to trip from the inrush current to turn on the capacitors. High frequency currents from an elevator controller caused such a voltage distortion that 120 volt control circuits in energy management devices were damaged. The manufacturer of the elevator controller needed to add filters for the distorted currents.

94


MAINTENANCE AND TESTING Proper maintenance and testing procedures are needed along with good equipment records. An effective maintenance program will allow the system to operate when it is needed. A critical part of a preventive maintenance program is periodic testing of the total system. This requires that power be shut down and that means some scheduled downtime for the facility. In most cases this can be minimized with minimal impact by testing during off-hours or holidays. Maintenance requirements depend on the equipment. Batteries may need to be checked for water level, clean connections and specific gravity. Generators will need fluid changes and checks. Periodic starting of the engine is needed along with generator capacity, transfer switch and control circuit testing. The maintenance procedures for UPS units should be done in accordance with the manufacturer’s recommendations. The cost difference between the technologies is large. Online units can cost three times more than line interactive units and line-interactive units can cost three times more than standby systems. An important consideration is the cost of operation and maintenance. Fewer systems means less maintenance but not necessarily less operating cost. An online system becomes part of the power generation system since it is always operating. It consumes power and generates heat. Online and line-interactive types will minimize battery inventory and maintenance. It is often a wise choice to use a combination of online or line-interactive and standby units. The working life of the unit’s battery should be at least 3 to 5 years. It is good maintenance practice to discharge and recharge the battery every six months to prolong its working life. An important battery feature is recharge time. In the event of battery activation, recharge time is critical. Many units that are completely drained have recharge times of between 2 and 12 hours, depending on the battery size and UPS type. Testing the UPS on a regular basis is one way to guarantee that when power is needed, it will reach critical systems. It is important that the UPS monitors the battery’s condition with periodic tests and have a built-in mechanism to warn both audibly and visually, when battery power is low. The UPS’s firmware or software programming is also important. Select a unit that constantly monitors the unit’s condition, particularly in


95

cases of overload. Units with software can be programmed to determine what action the UPS will take for different types of power anomaly events. The unit’s software should include a feature to program test times and log these results. Since it is important to know when a power fluctuation or failure occurs, select a unit that has interface connections that will alert you through network messaging, such as e-mail, or telephone paging, when a power anomaly occurs. A cost/benefit analysis should be used to determine the need for a standby power system. Protection of life, property and the business future may be involved. Building code and equipment requirements may also be involved. These include sections of ANSI/NFPA, ANSI/ IEEE, ANSI/UL or ANSI/NEMA documents. Standby power decisions can be based on economic considerations, but factors such as the type of power interruptions can be tolerated, initial costs, operating and maintenance costs will provide an optimum solution. References Buff, Katrina, Editor, Energy and High Performance Facility Sourcebook, Editor, Proceedings of the 26th World Energy Engineering Congress, The Association of Energy Engineers, Inc., The Fairmont Press: Lilburn, GA, 2003. Gustin, Joseph F., Cyber Terrorism: A Guide for Facility Managers, Lilburn, GA: The Fairmont Press, Inc., 2004. Ondrejack, Richard, “Fade to Black,” Communication News, July 1999, Vol. 36, p. 36. Wells, Editor Joyce, Solutions for Energy Security and Facility Management Challenges, Proceedings of the 25th World Energy Engineering Congress, Lilburn, GA: The Fairmont Press, Inc., 2003.


Emergency Generators

97

Chapter 4

Emergency Generators Electric generators are a common source of emergency power. The engine may be powered by bottled or natural gas, gasoline or diesel. The load is coupled to normal power and the emergency generator through a transfer switch. Under normal conditions, the transfer switch connects the load with the utility power source. If there is loss of utility power, the generator will be started and the transfer switch will connect the load to the generator. This process takes about 10 seconds. When the utility power is restored, the load is transferred to normal AC power after 15 minutes. The generator can provide power to the load for as long as there is an adequate supply of fuel. Since during a power interruption there is a 10-second delay before the generator can power the load, the load must be able to survive this 10-second ride-through period. Natural gas generators are efficient and easy to maintain. Pollution is low and the flue gas is not a serious problem. But natural gas generators are not considered true emergency systems, since if there is an interruption from the utility gas supply, the unit will not function. Agencies such as the Joint Commission for Accreditation of Hospitals do not consider a natural gas unit as a true emergency generator. For critical loads, propane, gasoline or diesel powered generators are available. Smaller units generally use propane or gasoline while larger units use diesel with its lower operating costs, lower maintenance requirements and safety. This is because diesel fuel has a high flash point and low volatility. One method of providing standby power is serving a load with a double-ended or triple-ended power station. This technique is useful if both feeders are powered by different and independent sources. This arrangement provides redundancy for feeder cables, transformers and circuit breakers. The transfer to the alternate feeder may be manual or automatic. 97

98


There will be a power interruption of about 10 seconds during the transfer. AC GENERATORS AC power requires a stable frequency of 60-Hz (USA) or 50-Hz (UK). The frequency can be established by running the generator at a virtually constant speed regardless of the load placed on it. Modern solid-state electronic circuitry can be used to produce a stable output frequency from a fluctuating generator frequency. This allows stable AC power to be produced from a variable-speed power source. The different approaches to powering generators include a separate engine, regulated and governed to a constant speed and coupled directly to a generator. This is the traditional stand-alone generator set or genset. A variable-speed clutch driven off the engine can mechanically compensate for changes in engine speed, providing a constant speed of rotation to the generator. Solid-state variable-speed technology (VST) takes the fluctuating output of an engine-driven alternator and feeds it through an inverter to produce a stable source of AC power. In an alternator, a direct current (DC) flows through a set of field windings on the rotor, creating a magnetic field that produces an output in the stator windings. On start-up the residual magnetism in the rotor induces an AC voltage. Diodes built into the rotor rectify this to DC, which is used to power the field windings. An armature type of AC generator has several coils wound around the rotor or armature. Two or more electromagnets are mounted inside the generator case. The armature rotates inside these magnets, producing an alternating current in the armature coils. This AC output is transferred to slip rings on the end of the armature shaft, where it travels through spring-loaded brushes and then to the generator’s output terminals. This type of generator may have two, three, or four slip rings and brushes, depending on the configuration and power output. The smaller AC generators have two slip rings and brushes. Larger generators have four slip rings and brushes. The simplest generators have two field windings (a two-pole generator) to produce one magnetic field (north and south), but most have two sets of windings (a four-pole generator).


99

ALTERNATORS Instead of using fixed (permanent) magnets in the rotor, an alternator rotor has a soft-iron core wrapped with wire to form the field winding. When direct current flows through the coil, the iron core is magnetized. The stronger the current, the greater the magnetism. The alternator output is controlled (regulated) by varying the field current in the field winding. Current flows to the field winding through two slip rings rotor shaft. Contact is made by a spring-loaded carbon brush in a brush holder fixed to the alternator housing. A compound rotor has multiple interlocking fingers and produces multiple north and south magnetic poles when the field winding is energized. As these north and south poles spin inside the coils of the stator, alternating current is generated in the coils. The number of coils varies among alternators, when there are three coils spaced properly the resulting power output will be three phase. The current from each phase is alternating and to be of use in charging a battery it must be rectified to direct current. This is done with silicon diodes. A diode acts as an electronic switch device that allows the current to flow in only one direction. A single-phase alternator can be rectified with four diodes. Each end of the stator winding is connected to a pair of diodes. One pair passes only positive waveforms and is connected to the battery’s positive terminal. The second pair passes only negative waveforms and provides a return path from the battery’s negative terminal to the winding. The complete circuit is known as a bridge rectifier. In spite of a constantly reversing current flow in the stator, current flows in only one direction to the battery. In a three-phase alternator, three positive diodes are tied together to form the alternator’s positive terminal, which is connected to the positive terminal of the battery. Three more diodes form the alternator’s negative terminal.

GENERATOR MAINTENANCE Most generators have bearings that are sealed for life. Some have grease fittings, which need grease once or twice a year. The brushes on armature-type generators carry the full output current of the generator.

100


If they are allowed to wear down or stick in their brush holders and make an imperfect contact with their slip rings, arcing can occur with damage to the slip rings. The brushes need to be checked regularly, some manufacturers recommend every 200 hours. They should be replaced once they are worn to half their original length. They must move freely in their brush holders and have enough spring tension. Brushes should always be replaced back in the same holders.

DIESEL BACKUP Rudolf Diesel was involved in the new technology of refrigeration with several patents for a method of producing clear ice. Diesel spent several years in Paris, working on an ammonia engine, but was defeated by the corrosive nature of this gas at pressure and high temperatures. The theoretical basis of this work was a paper published by N.L.S. Carnot in 1824. Carnot addressed the problem of determining how much work could be accomplished by a heat engine with a repeatable cycle. Carnot’s engine used a boiler or other heat exchanger. As the air in a chamber is heated, it expands driving a piston. The temperature of the air and the pressure is higher during expansion than during compression. Since the pressure is greater during expansion, the power produced by the expansion is greater than that consumed by the compression. This results in a power output that can be used for driving other machinery.

DIESEL ENGINES Diesel engines are simple in principle, and require little routine maintenance, although this maintenance is essential to their longevity. In a diesel engine, a piston compresses air in a cylinder. The compression is measured as a compression ratio, where the cylinder volume with the piston at the bottom of the cylinder is compared with the volume when the piston is at the top of its stroke. The more air that is compressed, the hotter it becomes. At compression ratios between 16:1 and 23:1, the air temperature rises to over 1,000°F (580°C), which is well above the diesel fuel’s ignition temperature of 750°F (400°C). When the piston is near the top of its stroke, diesel fuel is injected


101

into the cylinder under pressure. As the mixture ignites, it raises the temperature and pressure even higher, which drives the piston down the cylinder. This is called the power stroke. Diesels are often called compression-ignition (CI) engines since they do not have an ignition system. The diesel fuel is ignited by the high temperature from compressing air. In a diesel engine, the air and fuel remain separated until the edge of combustion. The fuel is injected late in the compression stroke, when the cylinder pressure is high. Injection pressures range from about 1600 psi to more than 25,000 psi for modern engines. The amount of fuel supplied depends on load and speed requirements. Air enters through the intake manifold, similar to a spark ignition engine. But, the air supply is unthrottled, so that the diesel engine draws the same volume of air per revolution, regardless of its speed. At idle, the open manifold supplies a large surplus of air, which passes through the engine without taking part in combustion. Typically, the idle speed air consumption is about 100 pounds per pound of fuel consumed. At high speeds or heavy loads this ratio drops to about 20:1. With no throttle valve, a diesel breathes easily at low speeds, while consuming little fuel. A spark ignition engine requires a fuel-rich mixture at idle to generate enough power to overcome the throttle restriction. The absence of an air restriction and an ignition system can sometimes cause a runaway condition where lube oil enters the combustion chamber where it is burned as fuel. Oil getting by worn piston rings can cause a runaway engine that accelerates itself until the air intake is blocked.

IGNITION AND COMBUSTION Spark ignition (SI) engines are fired by an electrical spark timed to occur just before the piston reaches the top of the stroke. Since the complete charge of fuel and air are available, combustion proceeds quickly in a kind of controlled explosion. The cylinder pressure rises rapidly within a few crankshaft degrees of piston motion. Thus, the volume of the cylinder above the piston undergoes little change from the moment of ignition and the point of peak pressure. This makes SI engines almost constant volume engines. Compared to spark ignition, compression ignition takes some time

102


for the fuel spray to vaporize and the spray to reach ignition temperature. The fuel continues to be injected during this time. Once it is ignited, the accumulated fuel burns quickly with corresponding increases in cylinder temperature and pressure. The injector will continue to deliver fuel through this period of rapid combustion and into the controlled combustion that follows. Then, injection stops and combustion enters what is known as the afterburn period. Diesel fuel quality is expressed by the cetane number. The more easily the fuel ignites, the higher the cetane number. The rough combustion associated with ignition lag is sometimes called diesel detonation. However, diesels do not detonate like SI engines. Some vibration or clatter can occur during start-up but if the condition persists, there may be injector or fuel quality problems. In normal operation, with ignition delay under control, cylinder pressures and temperatures rise more slowly to higher average levels than SI engines. Cylinder pressures tend to remain more constant through the expansion (or power) stroke. The pressure rise is relatively smooth and diesel engines are sometimes called constant pressure devices to distinguish them from constant volume SI engines.

DIESEL CYCLES With these exceptions, CI and SI engines operate on similar cycles, consisting of an intake, compression, expansion, and exhaust. Four-cycle engines of either type require one up or down stroke of the piston for each of the four tasks. Two-cycle engines compress this into two strokes of the piston, or one crankshaft revolution. Among these are the Detroit Diesels which use a mechanically driven supercharger mounted in the air inlet to compress the incoming air. When the piston nears the bottom of the power stroke, the exhaust valves are opened and the exhaust gases exit the cylinder. The descending piston uncovers a set of ports in the cylinder wall and the pressurized air flows in expelling the remaining exhaust gases and refilling the cylinder with fresh air. The piston reaches the bottom of its stroke and is on its way back up the cylinder. The exhaust valves close and the ascending piston blocks off the inlet ports in the cylinder wall. The cylinder is full of clean air and a new compression stroke begins.


103

In Detroit Diesel two-cycle engines, the blower purges the cylinder without raising its pressure much above atmospheric. The exhaust valve remains open until the ports are closed to eliminate a supercharge effect. In a four-cycle diesel engine, there are four piston strokes in a one complete cycle. After the compression and ignition strokes, on the next upward stroke the piston expels the burned gases through the exhaust valve. On its next downward stroke it takes clean air into the cylinder through the inlet valve. In the four cycle engine, air enters through the open intake valve and fills the cylinders as the piston falls on the intake stroke. The intake valve closes as the piston rounds bottom dead center. Injection begins near top dead center on the compression stroke. The fuel ignites and drives the piston down on the expansion stroke. The exhaust valve opens and the piston rises on the exhaust stroke, purging the cylinder of the spent gases. As the piston again reaches top dead center, the four strokes of the cycle are complete, taking two crankshaft revolutions. Diesel engines must be sturdier than SI engines. They are assembled with higher grade components such as anodized pistons, forged crankshafts and complex oil systems. Diesel manufacturers rely heavily on modular construction. Cylinder liners, pistons, rods, valves and injectors are common in engine families.

FUEL EFFICIENCY In theory, a diesel engine requires 10,000 volumes of air per volume of fuel oil, but in practice, air volume is increased by 1/4 to reduce smoke. Thus, the high-pressure injector has only 1/12,500 the capacity of the cylinder. The need to inject a fine fuel spray with enough force to penetrate cylinder compression requires injection pressures of 25,000 psi in modern, emissions-rated systems. Older engines may operate at 2,500 psi. Fuel metering parts are manufactured with extreme precision and lapped to tolerances expressed in wavelengths of light. Pump plungers are non-interchangeable and plungers that have been exposed to direct sunlight cannot be fitted to their barrels without first equalizing the temperatures. High compression ratios or large ratios of expansion, give diesel engines greater thermal efficiency. Under the best conditions, a well

104


designed SI engine utilizes about 30% of the heat from the fuel. The rest is lost as heat through the exhaust, cooling system, lubricating oil and surrounding air. Thermal efficiencies for diesel engines can reach 40% or more. Gas turbines have higher efficiencies, but only at a constant speed. This thermal efficiency, plus good volumetric efficiency and the capability to recycle some exhaust heat by turbocharging improve fuel economy. Diesels can provide a fuel consumption of 0.35 pound per brake horsepower hour operating near its torque peak. A gasoline engine may burn 0.5 pound of fuel per hp-hour under the same conditions. The weight differential between diesel fuel (7.6 pound/US gallon for No. 2-D) and gasoline (about 6.1 pound/US gallon) gives the diesel an even greater advantage when consumption is measured in gallons per hour or mile. Diesel passenger cars and light trucks can deliver 30% to 50% better mileage than the same vehicles with gasoline engines although there is some trade-off in acceleration and top speed. The high-octane fuels used for SI engines have low cetane numbers. Aviation gasoline is a poor diesel fuel, usable only if the compression ratio is raised to generate the necessary ignition temperature. However ether or amyl nitrate, which would deteriorate in SI engines, are excellent starting fuels for diesels. Fuel sold in this country must have a cetane number of at least 40. Heavier crudes tend to have higher numbers. Lighter, more aromatic crudes can be boosted to this level during refining. Viscosity, or the pourability of the fuel, also affects performance. Lower, more gasoline-like, viscosity fuels tend to atomize better and have been shown to generate less exhaust smoke. Low viscosity fuels tend to leak past the pump plungers, so that less is available for injection. At room temperature, the viscosities of light diesel fuels sold in this country vary from about two centistrokes to more than four. Fuel quality can also affect engine and component life. The abrasive ash left after combustion can collect on injector nozzles where it disrupts spray patterns. This ash can also increase upper cylinder and ring wear. U.S. fuels contain 0.01% ash by weight. Water, when present in sufficient quantities, promotes bacterial growth. RED AND BLUE DIESEL Since 1994, diesel fuel sold in this country comes in red, blue, and clear varieties. These are three versions of what was formerly known as


105

No. 2 Diesel. The 1990 amendments to the Clean Air Act called for a reformulated diesel fuel, with significantly less sulfur than the 0.5% then in effect. Sulfur and its compounds are pollutants that cannot be effectively addressed by vehicle emissions systems. The new fuel, which contains 0.05% sulfur by weight, was phased in during 1993 for use in commercial trucks and diesel passenger cars operating on the public roads. It was decided to continue to provide high-sulfur fuel for agricultural and construction equipment, railroad, commercial vessels, and diesel-fired heating systems. Users of this fuel do not pay the federal highway tax. The Environmental Protection Agency (EPA) made low-sulfur fuel colorless and the high-sulfur fuel tinted blue. The blue dye (1.2 dialkylamino-anthraquinone) gave the sulfurous fuel a green tint. Vehicles operated on public highways by government agencies and the American Red Cross (but not other charitable organizations) are exempted from the highway tax. But, the fuel for these vehicles falls under the low-sulfur rule. The IRS wanted this untaxed fuel to be stained red. Diesel fuel in California for over-the-road vehicles conforms to federal sulfur regulations, but contains no more than 10% by volume of aromatics. Federal fuel can contain as much as 35%. Aromatics contribute to particulate solids that blacken diesel exhaust.

STARTING Most diesel engines are fitted with electric starter motors. Some engines used to power construction machinery may use a gasoline engine rather than an electric motor as a starter. In cold weather the heat generated by compression tends to dissipate through the cylinder and head metal. Cold clearances may also cause some of the compressed air to escape past the piston rings. Other problems include the effect of cold on lube and fuel oil viscosity. The spray pattern coarsens and heavy oil friction between the moving parts increases. Many diesels have a cold-start mode that provides extra fuel to the nozzles and makes combustion more likely. Lube oil and water immersion heaters can also be mounted permanently on the engine. In cold

106


weather areas additional batteries can be wired in parallel. If chilled beyond the cloud point, diesel fuel enters the gelling stage. Flow through the system is restricted, filter efficiency suffers and starting is affected. Racor is one manufacturer of fuel heaters. These combine electric resistance elements with a filter or use a resistance wire in a flexible fuel line. Most indirect injection engines use glow plugs as a starting aid. These engines would be extremely difficult to start without some method of heating the air in the prechamber. A low-resistance filament (0.25 to 1.5 ohms, cold) draws a heavy current to generate 1500°F at the plug tip. Early types used exposed filaments which sometimes broke off. Improved versions contain the filament inside of a ceramic cover. Glow-plug systems are energized by a switch and may use a timer. The systems used in modern automobiles automatically initiate glowplug operation during cranking and switch them off when the engine starts. A solid-state module with an internal clock is used. A pulsed system opens the glow-plug power circuit for successively longer intervals as the engine heats and the timer counts down. In any version, the glow-plug resistance varies with temperature and the plugs function as heat sensors. During cold starts, a relay may direct full battery voltage to the glow plugs, as the engine heats the relay opens. Starting fluid can be used in the absence of intake air heaters. Aerosol cans are available for injection directly into the air intake. Some engines are fitted with starting devices for a capsule of fluid. A needle pierces the can releasing the fluid. COOLING Most diesel engines are liquid-cooled with a radiator, circulation pump and a thermostat for quicker warm-ups. Stationary and large vehicular engines may include oil and transmission coolers. With the thermostat closed, water flow is limited to a small port in the thermostat and the engine heats quickly. At a predetermined temperature the thermostat opens and circulation is unobstructed. GOVERNORS The majority of older diesels employ centrifugal flyball governors, which are part of the injector pump. The flyball governor invented by


107

James Watt for his steam engines, but instead of opening and closing a steam valve, the diesel governor rotates the injection pump plungers to deliver more or less fuel per stroke. Pneumatic governors respond to air velocity entering the manifold which is a function of piston speed. Coarse regulation is used in most installations and can result in speed change peaks of 10% or so. Fine regulation cuts this in half, for maximums of 2.5% over or under the desired rpm. Governors may require adjustment (high- and low-speed limits), cleaning, and replacement. Bearings and pivots can wear, coarsening the regulation. Diaphragms can fail from leaks or age hardening. The CAV governor is a typical mechanical governor with a manual override and a lever pivoting on a axle. The assembly pivots on command from the throttle. When the engine is running at about half-speed, a change in speed will be reflected by the flyweights. If the engine accelerates, the weights will press outward against their springs and move the lever, reducing fuel delivery. Under load the engine slows, and the weights move inward, causing the throttle to open. Pneumatic, or flap valve, governors operate on the venturi principle. When a moving fluid encounters a restriction, its velocity increases. The vacuum the venturi draws is a function of air velocity through it. Air velocity in a diesel engine is dependent on piston speed, so the venturi-induced vacuum can be used to monitor rpm. The vacuum is conveyed, by a tube, from the venturi to a diaphragmed chamber at the governor. The diaphragm movement is transferred to the fuel valve by a link. The application of digital electronics moved governors from simple speed-control devices to comprehensive engine management systems.

ELECTRONIC GOVERNORS Electronic governors are replacing mechanical governors on power generation, rail traction, marine propulsion, and other industrial engines. In large engines, mechanical governor functions include stepped speed control, reduced fuel delivery at high turboboost pressures, and automatic shutdown in the event of loss of oil pressure. The first step away from pure mechanical governors was control functions using ganged relays. Relay logic for on/off control has been

108


replaced with solid-state programmable logic controllers. The early electronic governors were analog devices that worked like mechanical governors. Instead of a force generated by a flyweight against a spring, analog governors sense speed as a voltage signal, which depends on engine rpm. This signal is compared to a reference voltage and the difference control the fuel flow. These analog governors were replaced with digital computers. Rather than using an analog voltage, digital governors count pulses that pass a sensing head. The computer then calculates engine speed and compares this to the programmed speed. A difference between the two causes the computer to generate a command signal to a fuel control actuator. The computerized engine control systems used on small and midsized engines are more like electronic governors with some self-diagnostic capabilities. A switch on the electronic control module (ECM) or unit (ECU) allows the operator to select the speed-control program. A minmax mode is used for power generation. The ECM may use sensors to measure manifold air pressure (MAP) or manifold air temperature (MAT). The ECM restrains fuel delivery until the MAP sensor reports that the engine has come up to the speed. Bearing noise means major and expensive repairs. By the time bearings reach this point, the journals have been pounded and bearing fragments have circulated through the engine. The technology exists to detect bearing failures early, while the damage is still minor. Crankpin bearing clearance checks help but the best protection is spectroscopic analysis of the lube oil.

TURBOCHARGERS A turbocharger is an exhaust-powered supercharger. The exhaust stream impinges against the turbine wheel and provides the energy to turn the compressor. This turbo boost is usually limited to 10 or 12 psi, but it is enough to increase engine output by 30% to 40%. Turbocharging represents the least expensive way to enhance performance. This energy would otherwise be wasted as exhaust heat and noise. Turbochargers develop maximum boost at high engine speeds and loads.


109

AIR SYSTEM Diesel engines run with open manifolds and are restricted only by the pressure drop across the air filter. The amount of air intake is about 200 cubic feet per pound of fuel for four cycle engines and significantly more for two cycle engines. Most of the air cycles through the engine, without taking part in combustion. But it must be filtered to remove abrasive dust. Most lethal particles are between 10 and 20 microns in diameter. Larger particles (or particulates) tend to pass out the exhaust without doing much damage. Small particulates are less aggressive, although they can still cause damage. The air cleaner cannot stop all particulates and some of the smaller particulates get through, which causes engines to wear rapidly in dusty environments. Air filters may be wire mesh, fabric, foam, or fiber. The filter can be dry, like the pleated-paper filters, or oil-wetted to improve particulate adhesion. The oil bath air cleaner combines oil-wetted filtration with inertial separation. There are two stages of separation. Air enters the top of the unit through the precleaner, or cyclone. Internal vanes cause the air stream to rotate, which tends to separate out the larger and heavier particulates. The air then goes through a central tube to the bottom of the filter, where it reverses direction and some of the remaining particulates drop into the oil. Polyurethane foam filters require reoiling in normal operation and tend to go dry in storage as the oil migrates to low spots. Unless oiled, these filters are ineffective. Pleated-paper filters work better when slightly dirty, but cannot really be cleaned. Solvents cause the fibers to swell. Overfilling an oil bath air cleaner can cause the engine to run away on the oil drawn from the reservoir. Some air cleaners include a pressure sensor that alerts the operator when the filter needs service. A manometer can also be installed on the intake manifold, downstream of the filter. The pressure drop across a new filter should be about 2 inches of water at normal engine speeds.

MAINTENANCE Most diesels will run trouble-free for thousands of hours with regular filter and oil changes. There may be several grease points that

110


need periodic attention. Belts to auxiliary equipment must be checked and kept tight (less than 1/2 inch or 13 mm of deflection under moderate finger pressure in the center of the longest belt run). The cooling system liquids will need periodic replacing. Starter batteries should be checked for water levels if they are not sealed. Most manufacturers have specific schedules for overhaul procedures. Problems such as difficult starting, changes in oil pressure or water temperature, smoky exhaust, vibration, or noise should be resolved quickly since delays may cause expensive repairs. The efficient operation of a diesel engine depends on maintaining compression. Small amounts of dust passing through a ruptured air filter or a leaking airinlet manifold can lead to rapid piston-ring wear and scoring of cylinders. Small particles can become embedded in the surfaces of pistons and bearings, which accelerates wear. If a filter is not changed and becomes plugged, it will restrict airflow to the engine. This limits the oxygen reaching the cylinders, and combustion, especially at high loads, suffers. The engine loses power and the exhaust will have black smoke from improperly burned fuel. Pistons, valves, turbochargers, and exhaust passages will carbon up, reducing efficiency and leading to other problems. The engine may overheat and even seize up. Air filters should be kept clean. The interval for changing filters depends on operating conditions. Most small diesels have replaceable paper-element-type filters. Some use the oil-bath type which forces the air to make a change of direction over a reservoir of oil. Particles of dirt are trapped in the oil. The air then flows through a fine mesh, which depends on the oil mist drawn from the reservoir to keep it lubricated and effective. As the reservoir fills with dirt, the oil becomes more viscous, less oil mist is drawn up and the filter becomes less efficient. Periodically the oil must be refreshed in the reservoir and the pan cleaned with diesel fuel or kerosene. The screen should also be flushed with diesel fuel or kerosene and blown dry. Refill the reservoir with oil, but do not overfill it, excess oil can get into the engine.

FUEL SYSTEM Conventional diesel fuel is a middle distillate, slightly heavier than kerosene or jet fuel. The composition varies with the source crude, refin-


111

ing processes used, additive mix, and regulatory climate. Fuel sold in this country must conform to standards set by the Environmental Protection Agency or to the more rigorous version of these standards accepted in California. High-cetane numbers provide good ignition and the fuel ignites at a low temperature and burns quickly for minimum ignition lag. Engines start easily on high-cetane fuels and, once started, tend to run smoothly. Fuel-injectors contain precise parts that can be affected by dirt or traces of water. It is important to keep the fuel clean. According to CAV, one of the largest manufacturers of fuel-injection equipment, 90% of diesel engine problems result from contaminated fuel. Minuscule particles of dirt can lead to the seizure of injection-pump plungers or the scoring of cylinders and plungers resulting in plugged or worn injector nozzles. Water in the fuel leads to a loss of lubrication of injection equipment and can result in seizures. In the combustion chamber, the result is misfiring and lowered performance. Water droplets in an injector can turn into steam in the high temperatures of a cylinder under compression. This creates an explosive force, which can blow the tip off the injector. Water in the fuel system will also cause rust to form on many parts. Fuel system malfunctions can produce multiple effects. An air leak in the system can limit prime to the injection pump and prevent the engine from starting. Less serious leaks introduce bubbles into the fuel stream and cause hard starting, refusal to idle, and loss of power. Leaks can also cause engine speed to rise and fall, or hunt. A restricted fuel filter can affect idle quality, throttle engine output, and cause the engine speed to surge. Inappropriate or contaminated fuel can make the engine refuse to start, misfire and exhibit loss of power. A fuel sample can be taken upstream of the filter. Allow the fuel to settle for a few minutes in a glass container. Cloudiness is an indication of water. Organic contamination will appear as jell-like particles floating on the surface. Placing a few drops of fuel between two pieces of glass will make it easier to see impurities. Several other tests for checking fuel quality can be made. Fuel that burns cleanly in an oil lamp, emitting little smoke, will also burn cleanly in the engine. Water can be detected by wetting a strip of paper with fuel, setting it alight and listening for the crackle as any water bursts into steam. Mixing a small quantity of fuel with sulfuric

112


acid will release the carbon and resins, which appear as black spots. The fewer of these spots, the better. Litmus paper can be used to show the presence of acids. If fuel quality is unknown, it would be helpful to determine the cetane value. Bacteria can grow in apparently clean diesel fuel, creating a film that can plug filters, pumps, and injectors. The microbes live in the fuel/ water interface, using both liquids to survive. There are favorable growth conditions in the dark, quiet nonturbulent environment found in fuel tanks. Two types of biocide are available to kill these bacteria. One is water soluble, the other is diesel soluble which is preferred. Some of the diesel-fuel treatments contain alcohol to absorb water, but this attacks O-rings and other nonmetallic parts in some fuel systems. Biocides can help to amend problems after they develop, but preventive measures should be used to avoid contaminated fuel. A length of clear plastic tubing, plugged off with a finger can be used to bring up a sample of fuel from all levels of the barrel, allowing an examination for contamination. Also, filter all fuel, using a funnel with a fine mesh or one of the multistage filter funnels. Regular samples from the bottom of the fuel tank can be used to check for contamination. At the first sign of contamination, drain the tank or pump out the fuel until no trace of contamination remains. A dirty batch of fuel should be discarded. When refueling fill the fuel tank to the top. This eliminates any air space and cuts down on condensation in the tank.

FUEL FILTERS Fuel filters function as a defense against contaminated fuel. Their function is to deal with minor contamination that escapes preventive measures to keep the fuel tank clean. A diesel engine may have a primary and a secondary fuel filter. A primary filter is the main defense against water and larger particles in the fuel supply, but it does not guard against microscopic particles of dirt and water. These are filtered out by the secondary filter. A primary filter is generally a sedimenter filter designed to separate water from fuel. Sedimenter filters consist of a bowl and deflector plate. The incoming fuel hits the deflector plate, then flows around and under it to the filter outlet. Water droplets and large particles of dirt


113

settle in the bowl. Better-quality filters then pass the fuel through a relatively coarse filter element of 10 to 30 microns. A micron is one millionth of a meter, or about 0.00004 inch. Some filters have a sensing device that sounds an alarm if water reaches a certain level. Others have a float that shuts off the flow of fuel to the engine if the water reaches a certain level. Some engines have two or more primary filters. This allows either filter to be closed off and changed without shutting down the engine. A secondary filter is designed to remove very small particles of dirt and water droplets. It cannot handle major contamination because its fine mesh would plug up. Secondary filters are usually of the spin-on type with a specially impregnated paper element to trap dirt. Water droplets are too large to pass through the paper and they will adhere to it. As more water is trapped, the droplets settle to the bottom of the filter, from where they must be periodically drained. The filter mesh is generally in the range of 7 to 12 microns. In some filters it may be as small as 2 microns.

LUBRICATION Lubricating oil in a diesel engine works much harder than in a gasoline engine, due to the higher temperatures and pressures used. This is especially true with modern lightweight, turbocharged diesels. Diesel engine oil must also contend with more acid and soot formation. Diesel fuels contain traces of sulphur. When a diesel engine is operated for long periods at light loads, it runs cool, which causes moisture to condense in the engine. This moisture combines with sulphur to form sulphuric acid, which attacks the engine surfaces. A lowload and cool running also generates more carbon (soot) than normal. This carbon can cause piston rings to stick and coats valve surfaces and stems, leading to a loss of compression and numerous other problems. Diesel engine oils are specially formulated to hold soot in suspension and handle the acids and other harmful effects of combustion. Using the correct oil in a diesel engine is important. Many oils are designed for gasoline engines and are not suitable for use in a diesel. The American Petroleum Institute (API) uses the letter C for compression ignition to indicate oils for diesel engines, and the letter S for spark ignition to indicate oils for gasoline engines. The C or S is followed by

114


another letter which indicates the additives in the oil. An oil rated CC, CD, CE, or CF-4 is suitable for diesel engines. Detroit Diesels use CD-II. As the oil is used to lubricate the engine, the additives and detergents are gradually used up. The oil must be replaced at regular intervals, more frequently than in gasoline engines. High-sulphur fuel is used in many Third World countries and much of the Caribbean and extended periods of low-load operation will increase the soot content, so oil-change intervals should be shortened to about 50 hours. Each time the oil is changed, a new filter should be installed to clear the engine of its contaminants. Without regular oil changes, the acids formed start to attack engine surface and the carbon overpowers the detergents in the oil. This can result in sludge in the crankcase and oil cooler. This sludge will begin to plug narrow oil passages and areas through which the oil moves slowly, causing a loss of oil pressure and lubrication. Major mechanical breakdowns can be avoided by regular changes of oil and filters. One major bearing manufacturer estimates that almost 60% of bearing failures are due to dirty oil or a lack of oil.

CENTRIFUGES AND BYPASS FILTERS A typical full-flow engine oil filter has a mesh of about 30 microns. This is relatively large since a finer mesh would plug up faster and need changing more often. Particles of dirt smaller than 30 microns can pass through the filter and circulate with the oil. Studies have shown the microscopic particles that pass through the filter and are the most destructive on engine wear are in the 10- to 20-micron range. Two methods are used to trap these particles: centrifuges and bypass filters. A centrifuge is a bowl mounted on bearings with small nozzles on the base of the bowl. The oil in these nozzles is under pressure from the engine oil pump, causing the assembly to spin. The centrifugal force generated by the spinning bowl causes the particles of dirt to be thrown out onto the centrifuge’s outer housing. This dirt accumulates as a dense, rubbery mat and periodically the outer housing is removed and cleaned. Centrifuges can remove particles down to 1 or 2 microns in size. They are usually found on engines of 100 to 200 hp or more. Bypass filters are used on all sizes of engines. They use a fine mesh filter element which, depending on the element, can filter particles down to 1


115

micron in size. A restriction built into the filter controls the flow rate at a level that will not cause a drop in the engine oil pressure. One type of bypass filter by TF Purifiner has a heating element to vaporize water or fuel in the oil. A centrifuge or bypass filter results in cleaner oil with a significant reduction in engine wear. Engine life in many applications is doubled. Many larger engines have one or the other as standard equipment, but few smaller engines have either.

OIL ANALYSIS Oil analysis has become routine for larger engines. The technology was developed for diesel locomotives and perfected by the Navy during the fifties. It can predict engine failure with the trace materials found in the oil. Spectroscopic analysis involves vaporizing an oil sample in an electric arc. Each element has its own frequency signature as it gives off light. Normally some 16 elements are tracked in concentrations of as little as one part per million. This information, along with engine history and user profile, gives a good representation of engine condition. The cost of the service is nominal and the transaction can be handled by mail.

COLD STARTS Many emergency diesel generators are installed outdoors in metal enclosures. The enclosures are often ventilated rain shelters which enclose the engine, generator and controls. Since diesel engines rely on the compression of air in the cylinders, the compression must create enough heat to cause combustion of the injected fuel. When the engine is cold, combustion may not occur or it may occur unevenly across the cylinders. Failed starting or rough starting causes premature wear on the engine. After starting, the engine should be warmed up before the electric load is applied. In the case of an outage, the engine may be loaded quickly as it reaches operating speed. Loading the engine before it reaches its normal temperature places stress on the bearings and other moving parts.

116


ENGINE HEATERS Most diesel generators have electric heaters to ensure that the engines will start in cold temperatures. These heaters are usually electric resistance heaters. The heaters are connected to the cooling water jacket that surrounds the cylinders of the engine. The heaters are usually set to maintain a 130-170°F water temperature. Water returning from the engine is about 10° cooler than the water leaving the heater. The heaters are generally sized at about 1-kW of heater capacity for each 100-kW of engine capacity. The heaters also reduce condensation and corrosion in the engine. Alternative methods involve the use of heat pumps for water heating, solar water heating, improved enclosure insulation and solar electricity to supply the resistance heaters. Air to water heat pumps for jacket water heating can reduce loads during warm days and night, but they require electric resistance heating when temperatures drop below about 38°. Solar electricity from photovoltaics can power the existing heaters, but may not be cost effective for the power levels needed. Solar water heating is feasible, but the high supply and return temperatures needed impact the efficiency of the solar water heating system. Evacuated tube solar collector systems can provide the high temperatures needed with high efficiency, they are not generally capable of disconnecting from the load under full sun conditions when the generator is running.

SOLAR AIR HEATING One approach to reduce the heat loss from an electrically heated engine is to use solar air heating for the enclosure. The solar air heating system provides a thermal blanket of air around the engine which reduces conduction and radiation of heat from the engine. The air is solar heated up to 110°F. The jacket water returning to the heater is at a higher temperature and requires less energy for reheating by the electric heater. Engine manufacturer warranties are affected since there is no direct connection between the engine and the solar air heating system. The maximum air temperature within the enclosure must be maintained at a temperature below the upper ambient operating temperature of the engine.


117

Heating of the enclosure also puts warm air in the area of the engine air intake. During the start of the engine, this warm air helps to ensure high compression cycle temperatures, quick combustion and a rapid start. The volume of warm solar heated air available in an enclosure during the start cycle is sufficient for about 15 to 30 seconds of operation. The mass of the generator equipment, which can be 5 tons for a 350-kW diesel generator alone and the structure, act as a thermal mass that maintains elevated temperatures in the enclosure after solar heating stops. Heating of the oil in the sump maintains the lubricating oil in a less viscous state. This results in more rapid lubrication of the bearing surfaces during start-up. When insignificant lubrication exists on the bearing surfaces, during the initial revolutions of the engine, high wear on the engine takes place. Reducing the viscosity of the oil by heating accelerates the flow of oil to critical wearing surface.

APPLICATION AND INSTALLATION Solar air heating generators have been used at the U.S. Geological Survey Headquarters. Solar thermal tile air heating systems were designed to reduce the electricity use of the jacket water heaters on these engines. One Cummins natural gas fired generator is enclosed in a well insulated enclosure. The walls of the enclosure are weather tight aluminum panels insulated with 1.5-inch-thick fiberglass contained behind perforated aluminum sound dampening panels. Temperature in the enclosure varied between 71 and 56°F, as the outside ambient temperatures varied from 62 to 23°F, during October, with no solar heating. The enclosure temperature, generally varied within 7 degrees of the daily high temperature. Other generators on the site were in fully ventilated enclosures. These enclosures have no insulation and open louvers. The temperature in these enclosures is usually within a few degrees of the ambient outside air temperature. Measurements of the electric energy of the electric engine heaters taken between July and December indicated that the heaters for the 750kW generator in the well insulated enclosure required a total of 1.4-kW in air temperatures near 70°F. This increased to 1.9-kW at 45°F air temperatures. Each of the 800-kW generators required 2.4-kW in 70° ambi-

118


ent air in the fully ventilated enclosures. The temperature range for the average high and low daily temperature throughout the year in Washington, D.C., is 26 to 89°F. Based on the average annual temperature in Washington, D.C. of 57°, it was determined that there was a potential reduction in heater demand from 2.9-kW to 1.45-kW. Annual electrical energy use for the 750-kW generator in the tight enclosure was 13,962 kWhr (48 million Btu). Each of the 800-kW generators in the fully ventilated enclosures were expected to use 25,147 kWhr (86 million Btu) throughout the year. The typical electric energy use for hot water heating in a single family home is 3,017-kW per year. At an average cost of $.046/kWhr, the demand to heat each generator with electric resistance heating, costs $642 per year for the 750-kW generator and $1157 each of the two 800-kW generators. The 750-kW generator in the insulated enclosure of 11' × 23' × 10' high has a solar tile system of 40' × 6'. This system is 160 feet from the generator enclosure. No dedicated thermal storage was used but the mass of generator and enclosure provide some thermal storage. The system delivers solar heated air when the sun is up and shuts off when collector temperatures drop below the enclosure temperatures. Another system heats a 250-kW generator in a smaller, fully ventilated enclosure. This enclosure is 4.5' × 10' × 5' high. The solar tile field is 30' × 8'. This system is 35 feet from the generator enclosure. In these systems, air is moved by a 240 watt, 10" tubeaxial fan in the solar tile support structure, through a 6" insulated duct to the enclosures. The duct is buried underground in a plastic liner. For the 40' × 6' system, a 120 foot portion of the duct is run above grade but below a ground cover. This routing was used to prevent trenching damage to tree roots. Both systems use switches to disconnect the system fan when the generator is running. This prevents excess heat around the operating generator. The radiator fan on the engine with the engine combustion air move almost 15 times more ambient air though the enclosure when the engine is operating, than the solar fan delivers for standby heating. The delivered air temperatures do not exceed 110°F since some diesel fuels can degrade above this temperature. Automatic controls will shut off the fan if temperatures in the space exceed 130°F. The 40' × 6' and 30' × 8' solar heating systems use a diamond slate solar thermal tile. The 12" × 12" glazing tiles are installed over corru-


119

gated metal absorbers to form a weather-tight collector surface. The tiles are fastened like slate roofing. The tiles are supported on a structure of self-framing walls and corrugated steel roof deck. The systems are tilted to 45° and face due south. As the air flows through the channels between the corrugated absorbers and the tiles, it is heated to about 70° above ambient temperatures at an airflow rate of 1 cfm per square foot of tile surface. This air temperature can be increased to about 100°F above ambient by varying the airflow rate. A differential controller senses when the collector temperature drops below the enclosure temperature and shuts off the fan. There was a 10-12°F temperature drop through the 160 foot duct with a 55-70°F temperature difference between solar heated air at the fan and ambient mid-day air. In October of 2001, an airflow of about 200 CFM in the 40' × 6' system resulted in a temperature rise at the fan of 60-75° above ambient temperatures. The ambient temperatures reached 85°. At 2 p.m. the collector fan temperature was 145°F with the delivered air temperature into the enclosure at 122°F. The thermally activated outside air damper had not yet been installed to limit the delivered air temperature to 110°F. On one November day with the outside air temperature at 56°F, the delivered air temperature to enclosure was 98°F.

GAS TURBINES Small gas turbines or micro gas turbines are being used more and more as backup or auxiliary power sources. The concepts of a gas turbine and a steam turbine appeared at about the same time. A 1791 patent for the steam turbine described other fluids or gases as potential energy sources. John Barber’s idea for gas turbine was a unit in which the gas was produced from heated coal, mixed with air, compressed and then burnt. This provided a high speed jet on the radial blades of a turbine wheel. Earlier developments in this area included Branca’s impulse steam turbine in 1629, Leonardo da Vinci’s smoke mill in 1550 and even Hero of Alexandria’s reaction steam turbine in 130 BC. These early versions of the gas turbine were really turboexpanders, since the source of compressed air or gas is a by-product of a separate process. These concepts turned into practical working equipment in the

120


late 19th Century by Charles de Laval and others. These units employed an impulse type of turbine wheel with expanding nozzles. The Industrial Revolution needed the power of steam turbines and the technology expanded to gas turbines, gas generator compressors and power-extraction turbines. The axial flow compressors in today’s gas turbines resemble a reaction steam turbine with the flow direction reversed.

POWER TURBINES In 1905, a gas turbine and compressor unit at the Marcus Hook Refinery of the Sun Oil Company near Philadelphia, PA, provided 4,400 kilowatts for hot pressurized gas and 900 kilowatts for electricity. The first electricity generating turbine for a power station appeared at Neuchatel, Switzerland, in 1939. This was a 4,000-kilowatt turbine with an axial flow compressor that delivered excess air at 50 pounds per square inch to a single combustion chamber for driving a multi-stage reaction turbine. Excess air was used to cool the exterior of the combustor and to heat air for the turbine. An early utility gas turbine in the U.S. was installed at the Huey Station of the Oklahoma Gas & Electric Company in Oklahoma City. This 3,500-kilowatt unit was installed in 1949. It was a simple-cycle gas turbine with a fifteen stage axial compressor, six straight flow-through combustors placed circumferentially around the unit, and a two-stage turbine. During World War I, the reciprocating gasoline engine was refined for the small, light aircraft of the time. Gas turbines were big and bulky, with too large a weight-to-horsepower output ratio for aircraft power plants. But, the turbo-charger became an addition to the aircraft piston engine. The exhaust-driven turbo-charger was developed in 1921, which led to the use of turbo-charged piston engine aircraft in World War II. In 1937, the British Thomson-Houston Company built and tested Frank Whittle’s jet engine. It had a double entry centrifugal compressor and a single stage axial turbine. A turbojet engine with a compound axial-centrifugal compressor similar to Whittle’s design and a radial turbine was built by the German aircraft manufacturer Heinkel. In 1939 a turbojet aircraft powered by this engine made the first flight of a jet powered aircraft. During the war years various changes were made in the design of


121

these engines. Radial and axial turbines were used along with straight through and reverse flow combustion chambers, and axial compressors. The compressor pressure ratio advanced from 2.5:1 in 1900, to 5:1 in 1940, 15:1 in 1960, and is now approaching 40:1. Since World War II, improvements made in aircraft gas turbine-jet engines have been transferred to stationary gas turbines. After the Korean War, Pratt & Whitney Aircraft provided the cross-over from the aircraft gas turbine to the stationary gas turbine. In 1959, Copper Bessemer installed the world’s first aircraft industrial gas turbine, in a compressor drive. This unit provided 10,500 brake horsepower (BHP) for a pipeline compressor. Airborne application units are referred to as jets, turbojets, turbofans, and turboprops. Land and sea-based applications units are referred to as mechanical drive gas turbines. Jet engines function as gas generators where the hot gases are expanded either through a turbine to generate shaft power or through a nozzle to create thrust. Some gas generators expand the hot gases only through a nozzle to produce thrust. These units are identified as jet engines or turbojets. The turbojet is the simplest form of gas turbine since the hot gases generated in the combustion process escape through an exhaust nozzle to produce thrust. Jet propulsion is the most common use of the turbojet, but it has been adapted to drying applications, supersonic wind tunnels, and as the energy source in a gas laser. Gas turbines that expand some of the hot gas through a nozzle to create thrust and the rest of the gas through a turbine to drive a fan are called turbofans. The turbofan combines the thrust provided by expanding the hot gases through a nozzle (as in the turbojet) with the thrust provided by the fan. The fan acts as a ducted propeller. In recent turbofan designs the turbofan approaches the turboprop in that all the gas energy is converted to shaft power to drive the ducted fan. When most of the hot gases through the turbine drive the compressor and the attached propeller with no thrust from the gas exiting the exhaust nozzle, the unit is called a turboprop. Turboprops have much in common with land and sea-based gas turbines. The engines used in aircraft applications may be either turbojets, turbofans, or turboprops, but they are commonly called jet engines. Since turboprops use the gas turbine to generate shaft power, they can be used wherever there is a need for large amounts of horsepower. At the 1967 Indianapolis 500 Race a Pratt & Whitney turboprop powered

122


car led the race for 171 laps, but suffered a gearbox failure on the 197th lap. The car had an air inlet area of 21.9 square inches. Later, race officials would restrict the air inlet area to 12.99 square inches which effectively eliminated gas turbines from racing. Some engines evolved from aircraft engines, but most land based gas turbines were derived from the steam turbine. Like steam turbines, these gas turbines have large, heavy, horizontally split cases and operate at lower speeds and higher mass flows than aircraft units of equivalent horsepower. Gas turbines in the small and intermediate size horsepower range incorporate features of aircraft and heavy industrial gas turbines. By the mid-1960s, the U.S. Navy was installing gas turbines as a ship’s propulsion power plant. The first combat ship to use a gas turbine was the USS Achville, a patrol gunboat commissioned in 1964. Larger ships like the Arleigh Burke Class destroyers use four aircraft derived gas turbines as the main propulsion units with 100,000 shaft horsepower. At the end of 1990s, the U.S. Navy had over 140 gas turbine propelled and 27 navies of the world had over 330 ships with about 800 gas turbines. By 1993, about 25,000 megawatts of electric power were generated by gas turbines. Many of these facilities utilize cogeneration to recover waste heat from the turbine exhaust. Gas turbines have also been used to power automobiles, trains, and tanks. The Abrams tank has a gas turbine engine that moves the 63 ton unit at over 40 miles per hour on level ground.

TURBINE CONFIGURATIONS Gas turbines can have different configurations with single or dual shafts and hot or cold end drives. A gas turbine is made up of a gas generator and a power-extraction-turbine. The gas generator consists of a compressor, combustor, and a compressor-turbine to drive the compressor. The power-extraction-turbine drives the external load. The compressor provides the high pressure, high volume air needed. It is heated in the combustor and expanded through the turbine section. Both axial and centrifugal compressors are used. Centrifugal compressors are used in smaller units, while medium and high horsepower applications use axial compressors. The energy that is developed in the combustor, by burning fuel


123

under pressure, is the gas horsepower (GHP). In turbojets, the gas horsepower that is not used to drive the compressor is converted to thrust. In turboprops, mechanical drive, and generator drive gas turbines the gas horsepower is used by the power extraction turbine to drive the external load. The gas horsepower may be expanded through the remaining turbine stages in a single shaft machine, or through a free power turbine, as done on a split shaft machine. A single spool-split output shaft gas turbine is also called a splitshaft mechanical drive gas turbine. The front shaft drives the compressor and the rear shaft drives the output load. The rear shaft comes off a free power turbine. The compressor/turbine component shaft is not physically connected to the power output shaft, but it is coupled aerodynamically. The aerodynamic coupling is also known as a liquid coupling. It provides easier, cooler starts on the turbine components and allows the gas turbine to reach self-sustaining operation before it drives the load. The gas turbine can operate at a low idle speed without the driven equipment rotating. The output shaft may be an extension on the turbine output (hot end drive) or it may be an extension of the compressor shaft (cold end drive). At the turbine end, the exhaust gas temperatures may reach 1,000°F (538°C). There are also exhaust gas turbulence as well as maintenance accessibility problems. At the cold end, accessibility is improved, the temperature is ambient, but the inlet duct must be turbulent free for proper turbine operation. The shaft and generation configuration must not impede a uniform, vortex free, flow through the input duct. Inlet turbulence can cause surges which can destroy the turbine blades. The dual spool split shaft gas turbine uses high and low pressure compressors and turbines. In this type of unit, there are three shafts and each operates at different speeds. Dual spool units are used for compressor, pump and generators drives in the higher horsepower ranges. Liquid coupling is used in compressor and pump drives as well as electric generator drives. This arrangement also allows the power turbine to operate at the same speed as the driven equipment. In generator drive applications the power turbines may operate at either 3,000 or 3,600 rpm to match 50 cycle or 60 cycle generators. Centrifugal compressor and pump application speeds are usually in the 4,000 to 6,000 rpm range. Matching the speeds of the drive and driven equipment eliminates the need for a gearbox. Gas turbines are used in the Alyeska pipeline to pump about 2

124


million barrels of crude oil per day some 800 miles to Valdez, Alaska. Aero-derivative single spool-split output shaft gas turbines are used to drive the large centrifugal pumps. Aero-derivative gas turbines are also used in the 900 mile Saudia Arabian East-West pipeline.

LUBRICATION Lubrication reduces friction between the rotating and stationary bearing surfaces and removes excessive heat from those surfaces. The bearings may be hydrodynamic or anti-friction types. The lubrication in a hydrodynamic bearing converts sliding friction into fluid friction. Anti-friction bearings work on rolling friction. The shaft load is supported by the rolling elements and bearing races in a metal-to-metal contact. Most heavy frame gas turbines use hydrodynamic bearings with mineral oil while the aircraft derivative gas turbines use anti-friction bearings with a synthetic oil. Mineral oils are distilled from petroleum crude oils and are less expensive than synthetic lubricants. Synthetic lubricants do not occur naturally but are made by reacting organic chemicals, such as alcohol or ethylene with other elements. Synthetic lubricants are used in high temperature applications of less than 350°F (175°C) or where fire-resistant qualities are required. The bearings in gas turbines are lubricated by a pressure circulating system. This consists of a reservoir, pump, regulator, filter, and cooler. Oil in the reservoir is pumped under pressure through a filter and oil cooler to the bearings and then returned to the reservoir for reuse. In cold climates a heater in the reservoir warms the oil prior to startup. The reservoir also serves as a deaerator. As the lubrication oil moves through the bearings, it can entrap air in the oil. This results in oil foaming. The foam must be removed before the oil is returned to the pump or the air bubbles may result in pump cavitation. To deaerate the oil the reservoir surface area must be large enough, so screens and baffles may also be built into the reservoir. Filters remove the wear particles from the oil. While 5 or 10 micron filters can be used for running conditions, 1 to 3 micron filters are used for break-in periods or after overhauls. Redundant oil filters are used along with a three-way-transfer valve. If the primary filter clogs, the transfer valve is switched over to the clean filter. Wear particles from the pump and gas turbine bearings


125

accumulate in the filter element along with temperature related oil degradation and oil additives. This may create a sludge that accumulates in the filter. As the filter clogs, the differential pressure across the filter increases. An oil pressure differential gauge is used for local readout, and a differential pressure transducer for remote readout and alarm. The differential pressure alarm setting is typically 5 psig. Regulators provide a constant pressure level in the lube system. These regulators allow the operation of a secondary pump for preventive maintenance. Lube oil coolers remove heat from the oil before it is re-introduced into the gas turbine. The amount of cooling required depends on the friction heat generated in the bearings, heat transfer from the gas turbine to the oil by convection and conduction and heat transfer from the hot gas path through seal leakage. The oil is cooled to 120°F-140°F (50°C60°C). To maximize the heat transfer, fins are installed on the outside of tubes and turbulators are placed inside each cooling tube. The turbulators transfer heat from the hot oil to the inner wall of the cooling tubes and the fins help dissipate this heat. Cooling media may be either air or a water/glycol mix. Air/oil coolers are used in desert regions, while tube and shell coolers are found in Arctic and most coastal regions. Air/oil coolers use ambient air as the cooling media. Cooling fans are usually electric motor driven, often with two speed motors. This allows high and low cooling flows. To closely match the cooling flow to the required heat load, changeable pitch fan blades may be used. As the heat load changes, the blades can be adjusted to meet the new heat flow requirements. Air/oil coolers may also include top louvers to protect the cooling coils from hail. These louvers are not effective for temperature control. Gas turbines have always been tolerant of a wide range of fuels from liquids to gases, to high and low Btu heating values and are now functioning satisfactorily on gasified coal and wood.

MICRO TURBINE POWER GENERATION Micro gas turbine based generator systems are becoming more popular for providing electric power and heat in cogeneration applications. High-speed turbo-generator sets are very compact in size and competitive for cogeneration. The electric generator is directly coupled

126


to a high-speed turbine. One of the newer power sources is the 20 to 60 kilowatt, regenerated, gas turbine power package. This package, in combination with a battery pack, can also deliver low emission power in automobiles. Operating at high speed, micro and mini turbines run in the range of 30,000 RPM to 120,000 RPM. The higher power units use lower speeds. A 40-kW micro turbine-generator may operate at 120,000 RPM while a 500-kW mini-turbine may operate at 30,000 RPM. Most micropower units use permanent magnet generators. In some designs, induction generators are used. Induction generators can provide lower costs, higher cycle efficiency and safety of operation.

TURBO GENERATOR SYSTEMS In most of these electric power systems the generator and the turbine are directly coupled. A constant speed of operation is used, but loads may cause an operating speed range between +0 to -5%. These integrated power systems are located close to the user such as in a factory building, hospital, store or office complex. Vehicle mounted applications are also used, but in either case the length of the feeders is short. Compatibility with utility power systems is usually required. The electrical power output is typically 3-phase AC with singe or multiple voltage lines. DC output may be required for some industrial applications. In most AC power systems, 50/60-Hz frequency is required, but a 400-Hz frequency may be required for aircraft/military/aerospace applications. In situations with stand-alone capability, isolation from the utility is required. In most emergency power applications, power transfer from utility to the turbo-generator and back is necessary. The generator must also provide electric start capability during the initial start-up of the turbine. The generator set must have a cooling system that is compatible with the system environment. Typically, either air, lubricant oil, or a water glycol mixture is used.

GENERATOR TECHNOLOGIES Three different generator technologies are suitable for high-speed operation. These are permanent magnet (PM), induction, switched reluc-


127

tance (SR), synchronous reluctance and homopolar. Permanent magnet technology is used by most micropower systems. The generator has two electromagnetic components. A rotating magnetic field is provided with permanent magnets. The stationary armature is made with electrical windings in a slotted iron core. The permanent magnets are made from high-energy rare earth materials such as Neodymium Iron Boron or Samarium Cobalt. A highstrength metallic or composite containment ring holds the magnets on the shaft. The stationary iron core is made of laminated electrical grade steel. The electrical windings are made from high purity copper conductors insulated from one another and from the iron core. The armature assembly is impregnated with high temperature resin or epoxy. The voltage output from the generator is unregulated, multiple phase AC. This voltage varies as a function of the speed and load. The output is connected to a solid state power conditioning circuit. This circuit uses buck/boost transformer switching to regulate the output. Induction generators are based on electric motor technology. Induction motors are the most common types of electric motors. Older induction generators were made using capacitors for excitation. These fixed capacitors could not be adjusted as the load or speed changed from nominal values. Induction generators were used in large power systems by utility companies. Excitation was provided from the infinite power bus as demanded by the load and speed conditions. Since the availability of high power switching devices, induction generators can be provided with an adjustable excitation and can be used in isolation. Induction generators use a rotating magnetic field constructed with high conductivity, high strength bars in a slotted iron core to form a squirrel cage rotor. A stationary armature is used similar to those used in PM generators. The voltage output from the generator is regulated, multiple phase AC. Control of the voltage is accomplished with a closed loop operation where the excitation current is adjusted to generate a constant output voltage regardless of the variations of speed and load current. The excitation current is supplied to the stationary armature winding from the short circuited squirrel cage which provides a secondary winding in the rotor. A switched reluctance (SR) generator uses the concept of magnetically charged opposite pole attraction. An unequal number of salient

128


poles are used on the stator and the rotor. Both the stator and rotor are constructed using laminated electrical grade steel. If the number of poles on the stator is 6, then the number of poles on the rotor will be 4. Other pole combinations such as 8/6 or 10/8 are also used. There is no electrical winding on the rotor. The armature coils for the stator poles are concentric and are isolated from one another. When the coils on opposite poles are excited, the corresponding stator poles are magnetized. The rotor poles that are closest to the stator poles are magnetized to opposite polarity by induction and are attracted to the stator poles. If the prime mover drives the rotor in the opposite direction, voltage is generated in the stator coil to produce electric power. The output from the generator is DC with a high ripple content. The voltage output may be filtered and regulated by adjusting the duration of the excitation current. Commutation of the current through the stator coil is accomplished by the controller.

INDUCTION GENERATORS Induction generators have several advantages in micro and mini turbine based power systems. The use of electromagnets instead of permanent magnets results in lower costs. The rare earth permanent magnets used in PM generators are much more expensive than the electrical steel used in electromagnets. They also must be contained using additional supporting rings. PM generators also require special machining operations and special handling of the magnets is required. The PM generator produces an unregulated voltage. Depending upon the changes in load and speed, the voltage variation can be wide. This is especially true for generators exceeding about 75-kW. An induction generator produces AC voltage that is almost sinusoidal. When an internal failure occurs in a PM generator, the failed winding will continue to draw energy until the generator is stopped. In highspeed generators, this may take some time during which further damage to electrical and mechanical components can occur. This can also be a safety hazard. An induction generator is shut down by the de-excitation that occurs within a few milliseconds. When an induction motor is operated at a speed higher than the synchronous speed, the shaft torque drops as the motor goes into the


129

generate mode. Electric power is generated from the mechanical input power from the prime mover. The generated power is a function of the slip, which is the speed in excess of the synchronous speed. In the generate mode, the slip is controlled according to the load requirements and the induction generator delivers the necessary power. The synchronous speed is a function of the electrical frequency applied to the generator terminals, but the operating shaft speed is determined by the prime mover. The electrical frequency must be controlled as changes in the load and the prime mover speed occur. The excitation current is provided to the generator stator windings for induction into the rotor. The magnitude of the excitation current determines the voltage at the bus. The excitation current is regulated to keep a constant bus voltage. The controller for the induction generator must adjust the electrical frequency to produce the slip corresponding to the load requirement and also adjust the excitation current to provide the required bus voltage. The electrical frequency must be changed from 100% at no load to about 95% at full load if the prime mover speed is stable at 100%. The controller for an induction generator consists of a power module, sensing circuits and a control module. The power transistors are IGBTs or MOSFETs in a conventional multiphase circuit configuration, the number of phases is the same as the phases in the generator winding. The sensing of currents and voltages is done at the load as well as in the power section of the controller. The speed of the shaft is measured and a PID control algorithm is used with switching commands for the power transistors generated by the control. This provides the necessary frequency and amplitude of the excitation currents for the induction generator windings. The control also includes over-current, overvoltage and over-temperature protection. The control of induction generator slip requires precise measurement of speed while an SR generator requires precise measurement of the rotor position. In the SR generator, the operating frequency can be 6kHz, at 60,000 RPM. In an induction generator, the operating frequency is 1-kHz to 2-kHz at 60,000 RPM depending upon whether a 2 pole or 4 pole generator is used. In an SR generator, the higher rates of change of currents and voltages result in higher stress levels for the power electronic devices. The induction generator has a sinusoidal output that can be more easily

130


conditioned. The controller for an induction generator is smaller and generally costs less than a controller for PM or SR generators.

TURBINE ADVANCES Gas turbine advances in recent years has been driven by metallurgical improvements that allow high temperatures in the combustor and turbine components. Other factors include both aerodynamic and thermodynamic breakthroughs and the use of computer technology in the design and simulation of turbine airfoils and combustor and turbine blades. There have been improvements in compressor, combustor and turbine design. The properties of creep and rupture strength improved from the late 1940s through the early 1970s. In 1950 390°F (200°C) was achieved in operating temperature. This resulted from age-hardening and precipitation strengthening which utilized aluminum and titanium in the nickel matrix to increase strength. Since 1960, sophisticated cooling techniques have been used for turbine blades and nozzles. Since 1970, turbine inlet temperatures have increased to 500°F (260°C) and in some units as high as 2,640°F (1,450°C). The increase in turbine inlet temperature was made possible with new air cooling techniques and the use of complex ceramic core bodies for hollow, cooled cast parts. Turbine blades and nozzles are formed by investment casting. A critical factor is the solidification of the liquid metal after it is poured into the mold. Nonuniform grain sizes, shapes, and transition areas can cause premature cracking of turbine parts. The equiaxed process provides more uniformity of the grain structure. The strength is improved if grain boundaries are aligned in the direction normal to the applied force. This elongated or columnar grain formation in a preferred direction is called directional solidification and was introduced by Pratt & Whitney Aircraft in 1965.

TURBINE CONTROL Newer turbines are under the control of a highly responsive unit using computer control technology. Computers start, stop, and govern the operation of gas turbines. They also provide diagnostics and predict future failures. Gas turbines are highly responsive, high speed units. In


131

an aircraft, a gas turbine can accelerate from idle to maximum take-off power in less than 60 seconds. In industrial gas turbines, the acceleration rate is limited by the mass of the driven equipment. Without the proper control system, the compressor may go into surge in less than 50 milliseconds and the turbine can exceed safe temperatures in less than a quarter of a second. A power turbine can go into overspeed in less than two seconds. The gas generator turbine and power-extraction turbine control takes place by varying the gas generator speed, which is done by varying the fuel flow. The following may be monitored: fuel flow, compressor inlet and discharge pressures, shaft speed, compressor inlet temperatures, turbine inlet and exhaust temperatures. At a constant gas generator speed, as the ambient temperature decreases, the turbine inlet temperature will decrease slightly and the gas horsepower will increase considerably. The increase in gas horsepower results from the increase in compressor pressure ratio and aerodynamic loading. This means the control must protect the gas turbine on cold days from overloading the compressor airfoils and overpressurizing the compressor cases. For maximum power on hot days it is necessary to control the turbine inlet temperature to constant values, and allow the gas generator speed to vary. The ambient inlet temperature, compressor discharge pressure and gas generator speed are the three main variables that affect the amount of power that the engine will produce. Sensing the ambient inlet temperature also helps to insure that the engine internal pressure are not exceeded, and sensing the turbine inlet temperature insures that maximum allowable turbine temperatures are not exceeded. Sensing the gas generator speed allows the control to accelerate through critical speed points. Gas turbines are typically flexible shaft machines and have a low critical speed. The control may be hydromechanical (pneumatic or hydraulic), electrical (wired relay logic), and computer (programmable logic controller or microprocessor). Typical hydromechanical controls include cams, servos, speed (fly-ball) governors, sleeve and pilot valves, metering valves and temperature sensing bellows. Electrical controls include electrical amplifiers, relays, switches, solenoids, timers, tachometers, converters, and thermocouples. Computer controls may simulate many functions such as amplifiers, relays, switching and timers. These functions are programmable. Modifying the program may be done by the

132


user or operator in the field. Analog signals such as temperature, pressure, vibration, and speed are converted to digital signals before they are processed. The computer may output signals to components such as the fuel valve, variable geometry actuator, bleed valve and anti-icing valve. Until the late 1970s, turbine control systems operated only in real time with no ability to store or retrieve data. Hydromechanical controls had to be calibrated frequently, weekly in some cases, and were subject to contamination and deterioration due to wear. Multiple outputs such as fuel flow control and compressor bleed-air flow-control required independent, control loops. Coordinating the output of multiple loops, using cascade control, was a difficult task and often resulted in compromises between accuracy and response time. Many tasks had to be done manually. Station valves, prelube pumps and cooling water pumps were manually switched on before starting the gas generator. Protection devices were limited and the margin between temperature control setpoints and safe operating turbine temperatures had to be made large since hydromechanical controls cannot react quick enough to limit high turbine temperatures, or to shut down the gas generator, before damage may occur. By the early 1970s, electric controls consisted of a station control, a process control, and a turbine control. All control functions such as start, stop, load, unload, speed, and temperature were generated, biased, and computed electrically. Output amplifiers were used to drive servo valves, using high pressure hydraulics, to operate hydraulic actuators. These actuators may also contain position sensors to provide electronic feedback. The advent of programmable logic controllers and microprocessors in the late 1970s eliminated these independent control loops, and allowed multi-function control. Control system functions include sequencing, routine operations and protection. Sequencing steps are used to start, load, unload, and stop the unit. The typical cycle used in a normal start is as follows: • • • • • • •

Starter on Fuel on Engine lights up - Exhaust gas rise Engine reaches self-accelerating speed Ignition off Starter cuts off Engine reaches idle RPM


133

When the start sequence is complete, the gas turbine has reached self-sustaining speed and the control mode goes into the routine operation control. This mode is used until the process starts to load the unit. Before initiating this loading, the sequence controller will position the inlet and discharge valves and circuit breakers. In electric generator drives this is when the synchronizer is used to synchronize the unit to the electric grid. When these steps are complete, control goes back to the routine operation mode. At this time, the speed control governor, acceleration scheduler, temperature limit controller, and pressure limit controller become active. Changing the fuel flow results in higher or lower combustion temperatures. When the fuel flow is increased, combustor heat and pressure increase and heat energy to the turbine is increased. Part of this energy is used by the compressor-turbine to increase speed which causes the compressor to increase airflow and pressure. The remaining heat energy is used by the power extraction turbine to produce more shaft horsepower. This cycle continues until the desired shaft horsepower or some parameter limit such as temperature or speed is reached. If the fuel flow is increased too quickly, excessive combustor heat is generated and the turbine inlet temperature may be exceeded or this increase in speed may drive the compressor into surge. When the control reduces the fuel flow, combustion heat and pressure drops, the heat energy available to the compressor-turbine drops and the compressor-turbine slows down. Lowering the compressor speed along with the airflow and pressure continue until the desired shaft horsepower is reached. If the fuel flow is decreased too rapidly, then the compressor may not be able to reduce airflow and pressure fast enough. This can result in a flame-out or compressor surge, since speed decreases move the compressor operating point closer to the surge line. Flame-out creates thermal stresses that become critical with each shutdown and re-start. High turbine inlet temperature will shorten the life of the turbine blades and nozzles, and compressor surge can severely damage the compressor blades and stators and possibly the rest of the gas turbine. The control must also protect against surge during rapid power changes, start-up, and periods of operation when the compressor inlet temperature is low or drops rapidly. A gas turbine is more susceptible to surge at low compressor inlet temperatures. Normally, changes in ambient temperature are slow compared to

134


the response time of the gas turbine control system. The temperature range from 28°F (-2.0°C) to 42°F (6.0°C) with high humidity is a major concern. Operation in this range can result in ice formation in the plenum upstream of the compressor. Anti-icing schemes increase the sensible heat by introducing hot air into the inlet. Anti-icing is another control function that must address the effect temperature changes can have on compressor surge. An acceleration schedule is used to load the unit as quickly as possible. As the load goes from the idle-no-load position to the full-load position, the fuel valve is opened and as the load approaches the setpoint, the speed governor begins to override the acceleration schedule output until the fuel valve reaches its final running position. During this time the temperature limit controller and the pressure limit controller monitor temperatures and pressures so that the preset levels are not exceeded. The temperature limit control for turbine inlet temperature uses the average of several thermocouples taking temperature measurements in the same plane. When the temperature or pressure reaches its setpoint, the limit control will override the governor controller and maintain operation at a constant temperature or pressure. The control allows the operating point to move along a set of points that define the operating line for the load conditions. Protection control continuously checks the speed, temperature, and vibration for levels that may be harmful to the operation of the unit. Usually two levels are set for each parameter, an alarm level and a shutdown level. When the alarm level is reached, the system provides a warning that there is a problem. If the transition from alarm to shutdown condition takes place too rapidly and operational response is not possible, the unit is automatically shut down. Overspeed is one parameter that does not include an alarm signal. Turbine inlet temperature is the most frequently activated limiting factor. One level is set for base load operation and a higher level is set for peaking operation.

STARTING Auxiliary turbine equipment includes the starting system, ignition system, lubrication system, air inlet cooling system, water or steam injection system for NOx control or power augmentation and the ammonia


135

injection system for NOx control. These systems may be direct driven and connected directly to the output shaft of the gas turbine. In most units, one of the lubrication pumps is direct connected. Indirect drives include electric, steam, or hydraulic motors. Indirect drives provide redundant systems, increasing the reliability. Electric systems are powered by a directly driven electric generator. Starting systems may drive the gas generator directly or through a gearbox. These may be electric, hydraulic, or pneumatic air or gas. The starter must rotate the gas generator until it reaches its self-sustaining speed and drive the gas generator compressor to purge the gas generator and the exhaust duct of volatile gases before initiating the ignition cycle. The starting sequence engages the starter, purges the inlet and exhaust ducts, energizes the ignitors and switches the fuel on. The starting system must accelerate the gas generator from rest to a speed just beyond the self-sustaining speed of the gas generator. The starter must develop enough torque to overcome the drag torque of the gas generator’s compressor and turbine and the attached load including accessories and bearing resistance. Another function of the starting systems is to rotate the gas generator, after shutdown, beginning the cooldown. Purge and cooldown functions have resulted in the use of twospeed starters. The lower speed is used for purge and cooling while the higher speed is used to start the unit. The starter may be directly connected to the compressor shaft or indirectly connected with an accessory gearbox or impingement air may be directed into the compressor-turbine. Starters for gas generators include alternating current and direct current motors, pneumatic motors, hydraulic motors, diesel motors and small gas turbines. If alternating current (AC) power is available, threephase induction type motors are preferred. The induction motor is directly connected to the compressor shaft or the starter pad of an accessory gearbox. Once the gas generator reaches self-sustaining speed, the motor is de-energized and usually disengaged with a clutch mechanism. If AC power is not available, called a black start, direct current (DC) motors are used. The source of power is a battery bank. One technique is to convert the DC motor electrically into a electric generator to charge the battery system. This is useful where the battery packs are also used to provide power for other systems. Battery-powered DC motor

136


starters are mostly used in small, self-contained gas turbines under 500 brake horsepower (BHP). Electric motors require explosion proof housings and connectors and must be rated for the area classification in which they are installed. Typically this is Class 1, Division 2, Group D. Pneumatic starter motors may be the impulse-turbine or vane pump type. These motors use air or gas as the driving force, and are coupled to the turbine accessory drive gear with an overriding clutch. The overriding clutch mechanism disengages when the drive torque reverses and the gas turbine self-accelerates faster than the starter. Then, the air supply is shutoff. Air or gas must be available at approximately 100 psig and in sufficient quantity to sustain starter operation until the gas generator exceeds self-sustaining speeds. If a continuous source of air or gas is not available, banks of high and low pressure receivers and a small positive displacement compressor can provide enough air for a limited number of start attempts. The starting system should be capable of three successive start attempts before the air supply system must be recharged. In gas pipeline applications, the pneumatic starter may use pipeline gas as the source of power. Hydraulic systems are often used with aircraft derivative gas turbines. Hydraulic pumps are used to drive hydraulic impulse turbine, Pelton Wheel starters or hydraulic motors for starting. Small gas turbines may be used to provide the power to drive either pneumatic or hydraulic starters. In aircraft, a combustion starter, which is essentially a small gas turbine, is used to start the gas turbine in remote locations. They are not used in industrial applications. Large heavy frame, 25,000 SHP and above, gas turbines require high torque starting systems. Most of these units are single shaft machines and the starting torque must be great enough to overcome the mass of the gas turbine and the driven load. Diesel motors are preferred for these large gas turbines. Since diesel motors cannot operate at gas turbine speeds, a speed increaser gearbox is used to boost the diesel motor starter speed to gas turbine speeds. Diesel starters are usually connected to the compressor shaft. Besides the speed increaser gearbox, a clutch mechanism is needed to disengage the starter from the gas turbine. Diesel motors can run on the same fuel as the gas turbine, eliminating the need for separate fuel supplies. Impingement starting involves jets of compressed air piped to the compressor turbine to rotate the gas generator. The pneumatic power source for impingement starting is similar to air starters.


137

IGNITION The ignition system is not energized until the gas generator reaches cranking speed and remains at this speed long enough to purge volatile gases from the engine and exhaust duct. When the igniters are energized, fuel can then be admitted into the combustor. These two functions are often done simultaneously and called pressurization. Two igniters are usually used, one on each side of the engine. During the start cycle each igniter discharges about 2 times per second and provides an energy pulse of 4 to 30 joules. One joule is the unit of work or energy transferred in one second by an electric current of one ampere in a resistance of one ohm. One joule/second equals one watt. Once the gas generator starts the igniter is no longer needed and any further exposure to the hot gases of combustion shortens its life. Some igniters are spring loaded and retract out of the gas path as the combustion pressure increases. Ignition systems include inductive and capacitive AC and DC, with high and low tension systems. The capacitive systems generate the hottest spark. Since the energy stored in the capacitor is proportional to the square of the voltage, it is more economical to use a high voltage to charge the capacitor. A radio frequency interference filter is used to prevent ignition energy from affecting local radio signals. The potential at the spark plug is about 25,000 volts. The ignition harness to each igniter plug is shielded and the ignition exciter is hermetically sealed. An AC transformer, or transistorized chopper circuit transformer, boosts the voltage to about 2,000 volts in the low tension system. A rectifier allows the flow of current into the storage capacitor but prevents most of the return flow. This voltage is boosted up to charge a smaller high tension capacitor. The low voltage charge in the storage capacitor is not enough to jump the gap across the spark plug electrodes. The initial path is provided by a higher voltage discharge from the high tension capacitor. It discharges first to bridge the gap across the electrodes of the spark plug and reduces the resistance for the low tension discharge. The low tension capacitor then discharges providing a long, hot spark. Inductive ignition systems use the rapid variation in magnetic flux in an inductive coil to generate enough energy for the spark. This system produces a high voltage spark, but the energy is relatively low and is only suitable for easily ignitable fuels.

138


Ignitor plugs use an annular-gap or constrained-gap. The annulargap plug projects slightly into the combustor, while the constrained-gap plug is positioned in the plane of the combustor liner and operates in a cooler environment. References Calder, Nigel, Boatowner’s Mechanical and Electrical Manual, 2nd edition, Camden, ME: McGraw-Hill Companies, 1996. Dempsey, Paul, Troubleshooting & Repairing Diesel Engines, 3rd edition, New York, NY: Tab Books, Division of McGraw-Hill, 1995.

Alternate Power Sources

139

Chapter 5

Alternate Power Sources BLACKOUTS AND POWER When blackouts struck much of the northeastern U.S., Ontario and Rome in the summer of 2002, users on both continents were reminded of just how fragile electricity supplies can be. The massive disruptions stranded commuters, defrosted freezers, shut down businesses and refocused attention on power. Most of our power comes from oil-and gas-fired generators and nuclear plants. These sources are tied into creaking infrastructures. They also pollute the environment and many feel they pose unacceptable health risks. The 21st century brings numerous challenges and opportunities that will affect the nation’s energy, economic, and environmental security: global economic and population growth. There will be a greater demand for power quality and reliability. Environmental sensitivities will also be among the major driving forces for improving the nation’s systems for generating and delivering energy. The prices of natural gas and oil continue to increase and have shown significant volatility. Electricity restructuring in California caused serious increases in electricity prices. Stress on the transmission and distribution system causes widespread power outages that affect millions of people and thousands of businesses. Renewable energy is an important element that improves our energy security, preserves the environment and supports our affluence. Renewable energy technologies can provide an important fraction of the nation’s electricity generation requirements and along with other generation sources provide more reliable power. Alternatives include clean energy from renewable resources. Fuel cells, wind turbines and solar panels can provide power free from dependence on local grids. The search for alternative energy is not new, but the current focus is on the goal of making clean and sustainable 139

140


power a mainstream commodity. The preferred energy future is plentiful and reliable sources of clean energy at reasonable prices. Several factors impact these preferences. These include global economic and population growth trends, technology advances, power quality and reliability problems, environmental challenges, and utility restructuring.

PHOTOVOLTAICS The nation’s energy generation and delivery systems will be changing in the coming years. Photovoltaics (PV) is only one solution to these challenges, but this renewable-energy option can be an important contributor to reducing the energy needs of the United States and the world. Photovoltaics has several characteristics that make it an important component of improving our nation’s power needs. PV is a versatile technology that can be used for applications from very small loads to moderate loads. It is a modular technology that allows generating systems to be built incrementally to match growing demands. PV is easy to install, maintain, and use. It is a convenient technology that can be used anywhere there is sunshine and it can be mounted on almost any surface. PV can be integrated into building structures to maximize aesthetics and multifunctional value. These positive features allow PV to address the problem of reliable power, power quality and power backup. The cost of power interruptions is high and customers with essential web servers or critical hospital or industrial needs cannot tolerate power interruptions or poor-quality power. Each year, U.S. businesses spend about $2 billion for uninterruptible power supplies and consumers purchase almost 200,000 small generators of 3 kilowatts or less. The losses incurred by businesses due to power quality and reliability problems are estimated at more than $30 billion each year. Distributed generation sources, such as PV, can improve grid reliability by reducing loads on transmission and distribution systems. Photovoltaic technologies provide solid power reliability with on-site generation and short feed lines. The reliability of photovoltaics is demonstrated by projects like San Francisco’s plan to install photovoltaicpowered traffic stoplights that have backup battery power at 100 key


141

intersections. The City will use PV to prevent traffic problems during rolling blackouts in California. Battery backup ensures that power is available at night or when the sun is not shining. In a grid-connected application, the grid uses this local generated solar electricity while the sun shines. This electricity is deducted from the power provided by the utility which results in lower power costs and provides more output for the grid. By 2020 the domestic photovoltaic industry will provide about 15% (3,200 MW) of U.S. peak electricity generating capacity. The cumulative installed capacity in the United States for PV will be about 15-GW. Conversion efficiencies are expected to be 18% to 20% at a cost of less than 50 cents per watt. The PV industry will be at $15 billion annually. Production facilities will be highly automated, reducing the cost of production by a hundredfold from 2000 values. PV systems will have plug-and-play qualities that allow them to tie into the existing grid structure for electricity production. Photovoltaics is a semiconductor based technology that converts sunlight into electricity. The photovoltaic effect produces direct-current (DC) electricity, while using no moving parts, consuming no fuel, and creating no pollution. Solar-electric power is suited to be a major contributor to an emerging national energy mix. The U.S. electrical grid will increasingly rely on distributed energy resources in a competitive market to improve reliability and moderate distribution and transmission costs and onpeak price levels. A greater value is being placed on power reliability as well as on lower energy cost. Many regions are becoming limited by transmission capacity and local emission controls. Solar-electric power addresses these issues since it is easily sited at the point of use with no environmental impact. Since sunlight is widely available, the United States can build a solar-electric infrastructure that is geographically diverse and less vulnerable to international energy politics and volatile markets based on fossil fuels. The International Energy Agency (IEA) projects that 3000-GW of new capacity will be required globally by 2020, valued at around $3 trillion. IEA also projects that the fastest-growing sources of energy will be supplied by renewables. Much of this new capacity will be installed in developing nations where solar-electric power is already competitive. The United States has been the world leader in photovoltaic research, technology, manufacturing, and sales. But other countries are

142


increasing their efforts to provide important technologies and gain global market share.

PV CHARACTERISTICS A wide range of PV technologies and applications exist for solar electric power. Photovoltaic cells utilize solid-state technology to produce electric energy directly from the sun. The cells are connected together and laminated in a sealed package called a module, which can be mounted on roofs, packaged into roofing or other building products or installed on the ground. The solar electric power is produced as direct current, or DC power. PV modules are usually measured in kW, and come with a nameplate rating in kW STC (Standard Test Conditions). The alternating current (AC) power rating of kW PIC (PVUSA Test Conditions) provide an estimate of the AC power output. Utility grid systems operate on alternating current so grid-connected applications of photovoltaic modules require an inverter to convert the DC output to AC power. The inverter also provides a means to disconnect the PV module during an outage. Solar electric systems can operate independently from the grid by using energy storage in the form of a battery. The different types of solar modules allow mounting on a variety of building and on-ground environments. The PV modules are one component of the system. The other components are the inverter, mounting system hardware and wiring. The inverter converts the direct current produced by the PV modules to alternating current. The inverter may be the weak point in the PV system since it is the most frequent cause for failures and malfunctions. However, inverters are becoming more reliable. Most PV systems are ground or roof mounted. Ground mounted PV systems require significant amounts of space. There is the potential for vandalism or theft and the system may pose a safety hazard if not protected by fencing. Ground mounted systems may be stationary, one axis tracking or two axis tracking systems. Tracking systems provide a greater power output, but have a higher initial cost and may require additional maintenance. Roof mounted systems take advantage of unused roof space, but there is a risk of reducing the roof’s integrity by installing PV systems that penetrate the roof. This can invalidate the roof’s warranty and cause


143

leaking. Structure and foundation costs can be minimized with roofmounted PV systems. The roof orientation, shape and shading can affect the PV system output. The level of PV energy production depends on the weather at the site. Actual operating conditions will vary with time. There are models that will give estimates of the system’s output. The models depend on plane of array irradiance, ambient temperature and wind speed.

PV APPLICATION PV systems are modular and can be adapted to almost any site. Residential PV systems are usually roof-mounted and less than 5kW. These systems generally have a higher cost/watt than larger systems. The smaller inverters used for residential systems are often the most unreliable part of the system. Without some form of power monitoring it is difficult to determine if they are operating correctly. Large commercial and utility-scale PV systems are usually more cost effective than residential systems. Systems over 70kW have dropped in costs since 1996. A recent trend is building integrated photovoltaics (BIPV). When a building is designed for PV systems, performance can be optimized. In a study of PV systems with a generating capacity of 70kW to over 400kW system costs ranged from $5.31/W to $11.82/W. Large PV systems can differ in several ways, single-axis tracking systems require tracker mechanisms and installation, while roof-mounted systems may require cranes and expensive roof modifications. The availability of state tax credits and incentives, site-specific installation and interconnection requirements and labor costs all affect total project cost. Module costs have declined. The early residential systems were large, customized systems. But, by 2000 the majority of the systems installed were 1.5kW or less and many were standardized packages. Many of these used AC modules with ratings of 500 watts or less. Standardized systems reduce costs and installation becomes easier. In 1992 the Solar Electric Power Association (SEPA) was formed. Programs include its six year long Technology Experience to Accelerate Markets in Utility Photovoltaics (TEAM-UP) and Solar Power Solutions (SPS).

144


BARRIERS Despite the advances that the PV industry has made over the years, a number of barriers still exist. The primary obstacle to market penetration is the high initial cost of the systems. Various federal, state, and local programs help to bring down the cost of solar. The Solar Electric Power Association, with funding from the Department of Energy has developed Solar Power Solution (SPS). This program is designed to address the barriers to commercialization and market penetration. While the technology has made significant improvements in efficiency and reliability, local support for PV is necessary to advance commercialization. The interconnection of solar to the electricity grid should be easier. Net metering compensates the owner of a PV system for the excess electricity that the system produces and feeds back into the utility grid. A number of states have developed net-metering policies, but these are not uniform.

SOLAR DEVELOPMENTS A number of important solar power developments are taking place. Nasosys of Palo Alto, CA, is developing tiny photovoltaic cells that can be incorporated into the fabric of roofing materials to provide power to homes and other buildings. Nanosys is combining the science of solar cells with the science of nanotechnology, which manipulates items as small as an atom to do tasks from switching electricity to storing data to sensing the movement of a bridge girder that is beginning to weaken. Nanosys has already embedded microscopic photovoltaic crystals into plastic sheeting. A prefabricated Nanosys roof could generate enough electricity to run a typical home. Electricity generated during the day can be stored in batteries for use at night. A single square meter of the solar cell plastic will cost about $100 and last about 20 years, so a complete roof would cost a few thousand dollars. The tiles should generate electricity at a cost of about 4 cents per kWh, well below the 20 cents to $1 for traditional solar panels. The company has government contracts from the Defense Advanced Research Projects Agency, the National Science Foundation and the Na-


145

tional Institute of Health. By 2020 the domestic photovoltaic industry is projected to provide about 15% or 3,200 MW (3.2 GW) of U.S. peak electricity generating capacity. This will reduce peakload demand, when energy is most constrained and expensive. Peak shaving will reduce the need to build new power plants and transmission lines. These projects typically meet with customer resistance. This moves the load off the grid and handles peak loads at the point of consumer use providing distributed generation. The importance of PV technology to reducing loads and providing backup power is shown in two major growth areas: installed capacity and costs. The growth by 2010 is expected to be 100 times the level of installed systems in 2000. The total installed (annual) peak capacity is expected to be about 7-GW installed worldwide by our domestic PV industry during 2020, of which 3.2-GW will be used in domestic installations. It is estimated that the mix of applications will be: • • •

50% AC distributed generation, 33% DC and AC value applications, and 17% AC grid (wholesale) generation.

This is based on business plans and market trend projections of the PV industry and published independent analysis. Installed volumes will continue to increase, exceeding 25-GW of domestic photovoltaics during 2030. By 2020, the cumulative installed capacity in the United States will be about 15-GW, or about 20% of the 70-GW expected cumulative capacity worldwide. The system price will be $3 to $4 per watt by 2010. Total manufacturing costs (costs for the system components) are expected to be 50 to 60% of the total installed costs.

HYBRID WIND AND SOLAR SYSTEMS These two resources and technologies are complementary. They not only improve the reliability of a stand-alone power system, but are more cost-effective when combined. Many off-the-grid installations start with a few photovoltaic panels, since they are simple to install and the unit costs are not extreme. A small system may use two or more panels,

146


either in series or parallel batteries and inverter. Until recently, adding a wind turbine was difficult because even small turbines were costly. The emergence of inexpensive micro turbines has lowered the cost of building a hybrid wind and solar system. The wind turbine is usually less than 50% of a hybrid system’s total cost. Since wind has a higher power density than solar, even in lowwind areas, the addition of a small amount of wind capacity, such as provided by a micro turbine, can considerably boost the total energy available. Components such as batteries and inverters are critical parts of hybrid systems. Batteries cost $50-$100 per kWh of stored capacity and only about 50% of the energy stored in a battery can be withdrawn without sulfating the plates and reducing its effectiveness. Batteries also have a limited lifetime of several thousand cycles. After about 2,000 cycles, they have a reduced capacity. If a battery discharges 50% of its gross capacity through 2,000 cycles, it will deliver about 1,000 kWh of net electrical energy over its operating lifetime. Thus, battery storage alone costs $0.05 -$0.10 per kilowatt-hour of usable energy in an off-thegrid system. Inverters, such as the Trace SW (Sine Wave) series, can be programmed to start heavy loads when excess power is available, cut the load when battery voltage falls, or start and stop a backup generator as needed.

INTERCONNECTION TECHNOLOGY A major difference between the early 1980s and today is the type of small wind turbines being used for utility interties. In the 1980s most small turbines, such as those built by now defunct Enertech, used induction generators. Today, nearly all small turbines destined for utility intertie use permanent-magnet alternators and inverters.

GENERATORS AND TRANSMISSIONS Wind power is converted to electrical power by a generator or alternator. The generator could be AC or DC but an alternator produces AC. Automotive alternators contain AC-to-DC rectifiers that supply DC


147

current to the automotive electrical system. AC power can be generated directly at the wind generator, or AC can be made from DC using an inverter. Alternating current can be produced by either induction generators or synchronous generators. They run at an rpm that is governed by the load and the line frequency. Typically, they run at 1,750 to 1,800 rpm, with the higher rpm occurring when the motor is fully unloaded. Direct current is generated by either a DC generator or an AC alternator with rectifiers. The traction motors used in golf carts, fork lifts, and electric cars can be operated as generators. These motors have commutating brushes that carry the full output current of the generator. Alternators come in a variety of sizes and types, from small automotive types to large industrial alternators. Automotive alternators are not as efficient (60%) compared to over 80% for industrial and DC traction types. They must also be driven at high rpm, but they are inexpensive. The alternator or generator is usually coupled to the rotor through some form of transmission that speeds up the relatively slow-turning powershaft to the higher rpm required by the generator. Older machines such as the Jacobs units, used a slow-turning generator that could be coupled directly to the rotor. Newer alternators need to spin much faster and a speed-up gear is needed. Speed-up mechanisms have included chains, belts and gearboxes. Chain drives are less expensive than gearboxes, but oiling and tensioning requirements make them less desirable. Some designs enclose the chains and sprockets in sealed housings with oil bath lubrication and a spring-loaded tensioner. Belt drives include toothed belt and V-belt. Belts can be temperamental. If their pulleys are not properly aligned, the belts will slide off. If the torque is high at low rpm, they can jump teeth and self-destruct. Cold weather can also effect them. Most small wind turbines use permanent-magnet alternators. Wound-field alternator are also used as well as induction generators. Some permanent-magnet alternators used by small wind turbines use a case where the magnets are attached which rotates outside the stator, the stationary part of the generator. The blades are bolted directly to the case. Centrifugal force presses the magnets against the inside wall of the case. In the more conventional alternator, the magnets are thrown away from the spinning shaft. Most small wind turbine alternators produce three-phase AC.

148


Some battery-charging models rectify the AC to DC at the generator. Others rectify at the controller.

INVERTERS Electronic inverters convert direct current into alternating current at a voltage and frequency determined by the inverter’s circuitry. Inverters that convert 12 volts DC into 110 volts AC are available in a variety for amperage ranges and voltage and frequency accuracies. A small 4ampere inverter that delivers 110 volts at 4 amps, or 440 watts, will cost about $100. Voltage output can vary from 105 to 115 volts, and frequency from 55 to 65 cycles per second. The output may not be a smooth sine wave of more expensive inverters, but rather a square wave. This square-wave AC can be used for operating motors, lights, and other non electronic devices. The square waves may effect electronic power supplies. More expensive inverters have close voltage regulation with a quartz crystal to control frequency and a sine wave output. For applications where the AC output will power electronic devices, this type of inverter is needed.

TYPES OF INVERTERS A DC/AC inverter basically reverses the action of a battery charger, producing line voltage, (120 or 240 volts AC) from a DC voltage. Two steps are usually required. The first is to convert DC to AC and then to step the AC up to the required output voltage. Inverters may produce a true sine-wave output similar to that produced by a generator or a stepped-square-wave that approximates a modified-sine. There are two ways to produce either waveform, linefrequency switching or circuit generated switching.

LINE FREQUENCY INVERTERS A line frequency stepped-square-wave inverter uses a transformer. The primary is low-voltage winding and the secondary a high-voltage


149

winding. The negative side of the DC input voltage is switched by transistors to each end of the primary winding. The positive side of the battery is connected to the center of the winding. The switches are turned on and off alternately at a rate determined by a crystal oscillator. As the switches are turned on and off, current flows through the coil in one direction and then in the opposite direction producing alternating current, with the frequency regulated by the oscillator. This alternating current is stepped up by the transformer to produce alternating current in the secondary winding at the required output voltage. The rate at which the transistors switch on and off determines the frequency of the AC output. This is set to produce 60 cycles a second in the USA and 50 cycles a second in Europe. The waveform produced by such a switching inverter is a stepped-square-wave instead of a true sine wave. The peak of the stepped-square-wave is lower than that of a true sine wave. The length of the wave is controlled to maintain an RMS output of 120 volts. As the load on the inverter increases, the battery voltage will decrease. To compensate for this falling voltage the wave length is increased so the inverter maintains its RMS voltage. A line frequency sine wave inverter uses control circuits that produce a controlled peak output voltage with a true sine wave. The process is more complex and these inverters are more expensive. High-frequency inverters use a center-tapped transformer with transistors that switch on and off at a frequency of 16KHz or more. The low voltage, high frequency AC is stepped up by a transformer to a voltage above 140 volts for a 120-volt system. This voltage is then rectified to DC and switched by transistors to produce AC at the correct frequency. The higher the frequency of the switching operation, the smaller and lighter, the transformers and the inverter. A 2-kW line frequency, square wave or sine wave unit will weigh about 50 pounds and occupy almost a cubic foot of space. High frequency models can weigh as little as 8 pounds and occupy one third the space. The high frequency inverter will generate much more radio frequency interference and it has more components which means a higher failure rate. High frequency units are not reversible while some line frequency inverters can be operated in reverse to charge batteries. Inverters tend to operate at the highest efficiency at or near their rated power. AC loads are not always on at the same time, so you can

150


often enhance the overall system efficiency by using small inverters at each important load. These local inverters should be selected to operate at their peak efficiency. Where voltage and frequency are not important, but sine wave output is, a motor-generator type of inverter may be used. Here, the DC powers an electric motor that in turn spins an AC generator. AC generators produce sine waves, but changing loads may change the motor speed and cause frequency and voltage dips. Most loads involving motors, draw a surge of current several times their rated amperage for a few seconds during starting. This current surge can damage an inverter that is not designed for it.

SYNCHRONOUS INVERTERS Synchronous inverters also turn DC to AC, but they can be used to drive an existing AC line. The AC output from a synchronous inverter is fed directly into the line, so this inverter must synchronize its output with the AC line. To be fully synchronized, the wave forms of the sine waves must match. Most synchronous inverters sample the utility wave form for internal voltage and frequency regulation. In the mid-1970s Windworks began using a synchronous inverter to connect 1930s-era Jacobs wind turbines to an electric utility. The technology has grown and is much more mature today. These inverters take DC, or rectified variable-voltage, variable frequency AC from permanent-magnet alternators, invert it to AC, and synchronize it with the AC from the electric utility. Synchronous inverters are line-synchronized, or line-commutated. The early models used SCR (silicon controlled rectifier) switches with analog controls. Since they are line-commutated they need the utility’s line to function. Bergey Windpower and Wind Turbine Industries still produce turbines with line-commutated inverters. Most modern inverters are self-commutated. They produce utilitycompatible electricity using internal circuitry with IGBTs (integratedgated, bipolar transistors) and digital controls. The newer self-commutated inverters provide better reliability and power quality compared to the older line-commutated versions. Self-commutated inverters use the same techniques as sine-wave


151

inverters for off-the-grid power systems. These inverters use the DC from a battery storage system and produce an AC sine wave similar to that of the utility. Modern sine wave inverters, such as Trace’s SW series, have the ability to feed excess power to the utility system. Unlike the old line-commutated systems, the new utility interactive systems require batteries to operate. They act as stand-alone power systems connected to the utility through the inverter. In these utility interactive systems, when electrical demand exceeds supply and the batteries are nearly spent, the inverter automatically draws power from the utility until the batteries are recharged. When there is a surplus of generation relative to the load and the batteries are fully charged, the inverter can also feed the excess power back to the utility. If the utility power fails, the inverter and batteries act as an uninterruptible power supply. The inverter automatically switches to the stand-alone battery system and makes the transfer from utility power. Advances in inexpensive inverters for photovoltaic panels also provide benefits to micro turbines. Some solar panels include their own 100-watt inverter, allowing a plug and play capability. In the future micro turbines could be used like contemporary computer peripherals.

INDUCTION GENERATORS Wind turbines using induction generators are the simplest to connect to the grid. The induction or a snychronous generator uses current in the utility’s lines to magnetize its field. The voltage and current the induction generator produces are now synchronized with that of the utility. The induction generator is unable to operate without the utility and when the utility system is down, the induction generator is down too. An induction generator can be used with a capacitor to charge the field, allowing induction generators to be used in stand-alone power systems. Vergnet wind turbines drive induction generators in this way in wind-diesel and battery-charging systems in France and its overseas territories. Induction wind turbines use control circuits to connect and disconnect from the grid, otherwise the wind turbine can motor in low winds, acting like a fan and consume electricity instead of generating it.

152


In an AC power system, the AC voltage cycles continuously from positive to negative and then from negative to positive. The nominal system voltage is an average known as the root mean square, or RMS voltage. The voltage at the peak of the sine-waves, both positive and negative, is higher than the nominal (RMS) voltage of the system. Peak voltage is found by multiplying the RMS voltage by 1.414, or dividing by 0.707. A 120-volt (RMS) circuit has a peak voltage of 169.7 volts and a 240-volt (RMS) circuit has a peak voltage of 339.36 volts. Alternating current of 120 volts surges first to 169.7 volts positive and then to 169.7 volts negative, for an RMS value of 120 volts.

VOLTAGE REGULATION Regulating the generator may be done with speed control (governing) or voltage regulation. Different approaches are used to control the speed of wind generators. Some use a centrifugally activated friction or air brake. Another technique is furling where a tilt-back mechanism progressively turns the unit away from the wind as the wind speed increase. Yet another technique is to design the blades to change their pitch so that they will not race at higher wind speeds. This is effective but can also be noisy. A conventional alternator is regulated by varying the field current to the field coil. Some wind generators have permanent magnets so this technique is not available for controlling output. Several methods are used to regulate a wind generator’s output. A circuit can be used to sense the battery voltage and control the current from the generator. Unless the generator has a governor, reducing the load can allow the generator to speed up. Some wind generators use this method with a furling tail that keeps the maximum speed below damaging levels. Other small generators use an air brake to control the speed. A shunt regulator can divert the generator’s output to another load as the battery comes up to charge. In such a charge diverting regulator, output is normally shunted to a fixed resistor, but could be switched to another load and put to useful work or sold back to the power company. DC generators require a diode in the positive cable to prevent a reverse drain from the batteries when the generator is not running. This diode is a part of many regulators circuits. Alternators already have this diode in the rectification circuit.


153

WIND AND WATER GENERATORS Small wind and water generators are similar units with different propellers (impellers, turbines, vanes). A wind generator uses a propeller, or turbine, to convert wind energy to a rotating force that is used to operate a generator. In this conversion process a doubling of the propeller diameter produces a theoretical fourfold increase in generator output, and a doubling of the wind speed produces a theoretical eightfold increase in output (see Table 5-1). Table 5-1. Wind Power Conversion Factors ———————————————————————————————— VALUE OF K TO GET WIND POWER IN Horsepower Watts Kilowatts ———————————————————————————————— Windspeed in mph Rotor area in foot2

0.006

4.3

0.004

Windspeed in foot second Rotor area in foot2

0.700

5.2

0.52

Windspeed in meters/second Rotor area in (meters2)

0.002

1.4

0.001

———————————————————————————————— At wind speeds of less than 5 knots there is not sufficient energy to produce any output from a wind generator. At 5 knots the more efficient generators will begin to trickle charge a battery. Less efficient designs will begin to generate when the wind speed reaches 7 knots. The propeller may be used to drive an alternator or a DC generator. Both produce alternating current (AC) in their output windings. But, an alternator rectifies this to DC with diodes, while a DC generator rectifies the output using a commutator and brushes. Small alternator types are made by Ampair, LVM and Rutland and DC generator types are made by Neptune, Fourwinds and RedWing. Some models can also be used as water generators. Alternator types need no brushes to generate electricity and will need less maintenance-free. In a DC generator the commutator and brushes require periodic maintenance.

154


The higher-output wind generators can be quite noisy. In strong winds the centrifugal forces developed by larger propellers (from 50 to 60 inches in diameter on up) can cause some units to self-destruct. Improved blade design, materials, and manufacturing tolerances, combined with methods to regulate the top speed of these generators, have eliminated these problems in some, but not all, generators. NET METERING Many states permit net metering for wind turbines. Net metering allows the sale of excess kilowatt-hours back to the utility. Net metering makes interconnecting a small wind turbine with the utility more attractive. Most states limit net-metering interconnections to 10-kW, although some states have a higher limit: Minnesota, 40-kW; Massachusetts, 30kW; New Mexico and North Dakota, 100-kW. There is no limit in Iowa. In most states, net-metering regulations affect only investor-owned or regulated utilities, thus excluding those connected to rural electric cooperatives. Eleven states offer net metering on both investor-owned utilities and rural co-ops. To find out if your state currently permits net metering, contact the American Wind Energy Association. POWER QUALITY AND THE UTILITY Before interconnecting a wind turbine with their lines, the utility may have some concerns about the power factor, voltage flicker, and harmonics produced by a wind machine. The utility may also require payment for reasonable costs arising from the interconnection. There are now several decades of experience with interconnected wind turbines. Wind turbines have operated more than three billion hours on the lines of electric utilities in Europe, the Americas, and Asia without creating electrical problems. MAINTENANCE Shaft movement up or down or from side to side indicates the need for bearing replacement. Fasteners in wind generators are subject to


155

vibration and sometimes work loose. Adding a drop of Loctite threadsealing compound can help. To cure excessive vibration, if the turbine blades can be detached individually, remove the opposite pairs and weight and correct any differences. Replace as matched pairs and check the alignment. The fiber-reinforced plastic blades in some small micro-generators are UV-degradable in sunlight. If the surface becomes ragged and powdery, they can be sanded and painted with a two-part polyurethane paint. The leading edge of some blades will wear down just from the impact of dust and rain. The blades should be kept smooth for maximum efficiency and noise reduction and recoated with epoxy or twopart polyurethane paint. The pivot points on an air brake, furling, or tilt-back mechanism need to be free and lubricated. Generator brushes and brush springs are the points of failure. Alternators have brushes only on the slip rings. Brushes and brush springs should be checked periodically for wear, corrosion, and loss of tension. If brushes or springs are defective, the commutator or slip rings should be checked for burning or pitting.

WIND POWER TRENDS A surge in the U.S. small wind turbine industry occurred in the early 1980s. It was backed by federal energy tax credits, state incentives and high electricity prices. It peaked in 1983 when energy prices fell, federal energy tax credits expired, and state incentives dropped off. By 1986, the main market was stand-alone or off-grid applications for remote homes. While filling this small domestic market, U.S. manufacturers expanded into overseas markets. Since 1999, electricity prices have been rising. There has been increased concern about the security of energy supplies and the centralized generating facilities that produce these sources of energy. Many want some independence from the electric utilities. There is also increasing concerns about global warming. Many state and local governments now provide significant incentive programs that include reduced costs for small wind turbine systems. These incentives reached a total of $3.5 billion in 2001 for programs that include small wind turbines. These factors have resulted in increased interest in small wind

156


turbines connected to the utility grid. In 2001, the annual sales of the U.S. small wind turbine industry were estimated at 13,000 turbines valued at about $20 million. This is about the same level as sales in the early 1980s, but it is only about 2% of the total sales of large wind turbines in the United States. The success of the large wind turbine industry is due to the support from government programs and policies both at home and abroad. Support such as federal and state tax credits were discontinued in the mid-1980’s for small wind systems. This led to a significant decrease of the industry and less momentum in technology and market developments. Combined efforts from government and industry are increasing the contribution of small wind turbines to the generation mix. There is the potential for real contributions to the energy supply. It is projected that small wind turbines could contribute 3% of U.S. electrical consumption by 2020. While foreign companies may dominate the market for other renewable energy technologies, the U.S. small wind turbine industry is the leader in markets at home and abroad. While producing energy, small wind turbines produce no environmental emissions. The market for small wind turbines has been growing 40% per year. The U.S. small wind turbine industry offers a variety of products for different applications and environments. The following range of products are available: • • •

Small generators of 400 watts (W) Mid-size generators of 3-15 kilowatts (kW) Larger generators of up to 100-kW for commercial operations.

Small wind turbines can be effective in most of the rural areas of the United States. About 60% of the United States has enough wind for small turbines to generate electricity.

TECHNOLOGY ADVANCES The newer small turbines have a high reliability with only two or three moving parts and require relatively low maintenance. Continuous advances in the industry and support with the U.S. Department of En-


157

ergy (DOE) has produced small wind turbines with advanced airfoils, super-magnet generators, smart power electronics, very tall towers, and low-noise. These features help to reduce the cost of producing electricity and have aided in the acceptability of wind power. Small wind technology has been improving since the 1970s. Improvements have taken place in operating reliability, noise concerns, and manufacturing and installation costs. Much has been done to incorporate advanced technologies and to enhance manufacturing. It is estimated that high-volume manufacturing alone could reduce costs 15-30%. Modern small wind turbines are not like the generators of the 1920s and 1930s. Today’s small turbines use aerospace technologies to provide sophisticated, yet simple, designs that allow them to operate reliably for a decade or more without maintenance. As small wind turbine technology has matured, the products have become simpler and more robust. The small wind turbine industry uses advanced component technologies and newer design tools such as three-dimensional solid modeling and computational fluid dynamics. Turbines use high-efficiency airfoils, neodynium-iron-boron super-magnet generators, pultruded FRP blades, graphite-filled injection molded plastic blades, special-purpose power electronics and tilt-up tower designs to lower costs and increase efficiency. Among the goals and trends in wind power is a reduction in cost of energy resulting from turbines that operate in low-wind areas along with new technologies for low-cost towers and low-wind rotors. Turbine costs are expected to drop through improvements in the performance and efficiency of small wind turbines. Technology developments are projected to reduce tower and installation costs by using lower-cost foundation or anchoring systems for towers and automated processes for tower fabrication. Turbine reliability should improve with improved test methods including extreme event testing and extensive multi-year data on turbine performance, reliability, operation, and maintenance. There will be increased participation in small wind turbines, an option in domestic government programs. The Federal Energy Management Program will develop more small wind projects at federal facilities and promote small wind turbines for security and military operations.

158


Reduced manufacturing costs from the increased volume of production and improved manufacturing techniques will include advances in equipment and processes for the mass production of small wind turbine systems. Additional standards will be developed to address reliability, durability, longevity, noise, and power performance. Stronger, certified distribution channels and support should include generic installation and maintenance training programs for small wind turbines along with certification standards. A more consumerfriendly performance rating system will include an updated American Wind Energy Association (AWEA) performance standard with the IEC 61400-12 for small wind turbines. U.S. manufacturers of small wind turbines export more than 50% of their production and have a leading share of the world market. The foreign market for grid-connected wind turbines is driven by electricity prices more than double those in the U.S. About 2 billion people in the world do not have access to electricity for domestic, agricultural, or commercial uses. The traditional method of providing electricity by extending the distribution grid is too expensive and not suitable to the low consumption levels in developing nations. The number of homes without electricity has been increasing because the birthrate has outpaced the electrification rate. Small-scale renewable energy systems (wind, micro-hydro, and solar) are often less expensive to install than power line extensions. Small wind turbines are less expensive to operate and produce much less carbon dioxide per kilowatt hour than diesel generators. Small wind systems can be used to provide power for 500W or 50kW loads. Wind electricity can be used for charging batteries for power backup applications. By 2020 small wind turbines should add 3% or 50,000-MW to the U.S. electric supply. The small wind industry should grow to a billion dollar industry, with more than 10,000 employed in manufacturing, sales, installation and support. Rooftop wind turbines can help power electric loads and send power back to the utility grid. Small wind turbines will become another appliance and may be purchased at local home improvement stores. Installed wind turbines should have a 50-year life based on reliability improvements. Small wind turbines have a generating capacity of up to 100 kilowatts (kW), which requires a 60-foot rotor diameter.


159

WIND ENERGY Energy and power are derived from the wind by using the force it exerts on solid objects. Windmill blades move in response to this force, and wind machines can extract a real portion of the energy and power available. The wind energy available in a unit volume (one cubic foot or one cubic meter) of air depends on the air density (D) and the instantaneous windspeed (V). This kinetic energy of the air in motion is expressed by the following relationship: kinetic energy ——————— unit volume

= 1/2 × D × V2

To find the kinetic energy in a particular volume of air, multiply by that volume. The volume of air that passes through a surface such as the area swept by the blades of a horizontal-axis windmill, oriented at right angles to the wind direction is: volume = A × V × t where: t is the elapsed time (in seconds) A is the area (in square feet or square meters) The wind energy that flows through the surface during time t is then: wind energy = 1/2 × D × V3 × A × t The wind power is the amount of energy which flows through the surface per unit time, and is found by dividing the wind energy by the elapsed time t. The wind power is then: wind power = 1/2 × D × V3 × A Note that both the energy and power are proportional to the windspeed cubed. If all the available wind power on the rotor could be utilized, this formula could be used directly to calculate the power. But, the windspeed near the blades is affected by their motion. The maximum power that can be extracted from the wind is only about 60% of the power available.

160


In actual practice, a wind machine extracts even less power than this maximum. The rotor may capture only 70% of maximum power. Bearing friction will take a few percent. Gears and other losses may lose half of the remaining power. The final output is reduced from the power that is actually available in the wind. This factor ranges from 0.10 to 0.50.

WIND ROTORS The power supplied by a wind machine (P), depends on the windspeed (V), the rotor area (A), the air density (D) and the system efficiency (E): P = 1/2 × D × V3 × A × E The rotor area can be expressed as: P A = ———————— E × F × Ca × Ct Here: F is a factor that depends on windspeed (see Table 5-2) Ca and Ct are correction factors for the air density at other attitudes This equation gives the area in square feet when the power P is expressed in watts. If P is in horsepower, multiply A by 0.737. If you are purchasing a factory-built machine and know its system efficiency, this formula can tell you if its frontal area is suited to your power needs. To size a three-bladed propeller machine to produce 2000 watts at winds that peak at 15 mph, start with the system efficiency. Small propeller-type systems have an efficiency of 15 to 30%. (see Table 5-3) Use 25% for an E of 0.25. F = 17.30 at 15 mph Ca = Ct = 1 at sea level, for standard temperature (60°F). A=

2000 ————————— 0.25 × 17.30 × 1 × 1


161

Table 5-2. Windspeed Factor F ———————————————————————————————— V F ———————————————————————————————— 6 1.1 8 2.6 10 5.1 12 8.9 14 14.1 16 21.0 18 30.0 20 41.0 22 54.6 24 71.0 26 90.1 28 112.6 30 138.4 ———————————————————————————————— This gives an area of 462 square feet and a diameter of 34.6 feet. Table 5-3. Wind System Efficiency ———————————————————————————————— Efficiency ———————————————————————————————— Multibladed farm type 10 - 30 Savonius windcharger 10 - 20 Small prop type (to 2kW) 20 - 30 Medium prop type (2 -10kW) 20 - 30 Large prop type (over 10kW) 30 - 45 Darrieus wind generator 15 - 35 ————————————————————————————————

WIND TURBINE SIZES Wind turbines range in size from micro-turbines like the 20-watt Martec 500, with a rotor only 0.5 meters (1.7 feet) in diameter to large units like Vestas’s 1,650 kW machine with a rotor spanning 63 meters (200 feet). (see Table 5-4)

162


Table 5-4. Characteristics of Small Turbines ———————————————————————————————— Manufacturer/ Rotor Area Rated Model Diameter (m2) Power Wind Speed (m)/(feet) (kW) (m/s) ———————————————————————————————— Micro Turbines Marlec/500 0.5/1.7 0.20 0.02 10.0 LVM/Aero4gen 0.86/2.8 0.58 0.6 10.3 Ampair/100 0.9/3.0 0.66 0.05 10.0 Aerocraft/120 1.2/3.9 1.13 0.12 9.0 Southwest 503 1.5/5.0 1.8 0.5 12.5 LVM/Aero8gen 1.6/5.1 1.9 0.22 10.3 Marlec/1800 1.8/5.9 2.5 0.25 9.4 World/H900 2.1/7.0 3.58 0.90 12.5 Bergy/1500 3.1/10.0 7.31 1.5 12.5 World/3000 4.5/14.8 16.0 3.0 11.0 Vergnet/6.5 6.0/20.0 28.3 5.0 10.0 Bergy/PD 7.0/23.0 38.5 10.0 12.1 Wind/29-20 8.8/29.0 61.4 20.0 11.6 ———————————————————————————————— Size classifications depend upon both the diameter of the rotor and the capacity of the generator. Typically, small wind turbines are machines producing a few watts to 10-20-kW. Wind turbines at the upper end of this range are driven by rotors 7-9 meters (23-29 feet) in diameter. Small wind turbines can be divided further into micro wind turbines, (the smallest of small turbines), mini wind turbines, and homesize wind turbines. Micro turbines are those from 0.5-1.25 meters (about 2-4 feet) in diameter. These machines include the 20-watt Marlec as well as the 300-watt Air 303 from Southwest Windpower. Mini wind turbines are slightly larger and span the range between micro turbines and home-size machines. They range in diameter from 1.25-2.75 meters (4-9 feet). Turbines in this group include World Power Technologies H500 at 500 watts and Bergey Windpower’s 850-watt unit. Home-size wind turbines are the largest of the small wind turbines. They include turbines as small as World Power’s Whisper 1000 with a rotor 2.7 meters (9 feet) in diameter to the Bergey Excel that uses a rotor 7 meters (23 feet) in diameter and weighs more than 1,000 pounds (463 kilograms).


163

WIND TURBINE RATINGS The turbine’s rated wind speed is the wind speed at which the wind turbine produces the power indicated on its nameplate. The peak power is usually higher than rated power. At a given speed, the wind turbine begins governing or limiting the power it produces. In the case of most small wind turbines, the rotor begins furling, or turning out of the wind which reduces the power produced. The Bergey 850 starts generating at 8 mph (3.6 m/s) and reaches its rated power at 28 mph (12.5 m/s). The rotor furls, or folds, toward the tail vane at 33 mph (14.7 m/s). At a wind speed of 18 mph (8 m/s), the Bergey 850 produces 327 watts. In a 12 mph wind influence with a Rayleigh distribution, an 18 mph (5.5 m/s) wind speed will occur 294 hours per year. At this speed, the turbine produces 0.33-kW × 294-h, about 100-kWh per year. Across the entire speed range, the Bergey 850 will generate about 1,400-kWh per year at this site. World Power Technologies estimates that its Whisper 500, which uses a 2.1 meter (7 foot) diameter rotor, will produce about 100-kWh per month at a site with a 12 mph average wind speed, or about 1,200-kWh per year. All this energy may not be put to use. Batteries that are fully charged may not be able to take more energy on a very windy day. Some of the energy that is stored in the batteries is lost due to the inefficiency of battery storage. Additional losses occur in an inverter to convert DC to AC. Only about 70% of the energy delivered to the batteries may be used for power.

ROTOR CONTROL Wind turbines work in a demanding environment. The heavier small wind turbines have proven to be more rugged and dependable than lightweight machines. Most wind turbines have some means for controlling the rotor in high winds. The smaller wind turbines do not use the yaw motors and mechanical drives of the bigger upwind turbines. Most small wind turbines use tail vanes to point the rotor into the wind. Micro and mini wind turbines furl, or fold so that the rotor swings toward the tail vane. Some furl the rotor vertically others furl

164


the rotor horizontally toward the tail. Some home-size turbines pitch the blades and others even pitch the blades and furl the rotor. To furl the rotor in high winds, the rotor axis is offset from the furling axis. In high winds, thrust on the rotor overcomes the force keeping the rotor into the wind, and swings the rotor toward the tail vane. The wind speed at which furling occurs a function of the hinge between the tail vane and the body of the turbine. In vertical furling, the high winds tilt the rotor up, where it resembles a helicopter rotor. A spring or shock absorber is used to dampen the rate at which the rotor returns to its normal position.

MICRO TURBINES Micro turbines can generate about 300-kWh per year at sites with average wind speeds of 5.5 m/s (12 mph) which is similar to that found on the Great Plains. When comparing wind turbines, compare rotor diameter first. For power ratings, always consider the wind speed at which the turbine is rated. There is almost three times more power in the wind at 12.5 m/s (28 mph) than at 9 m/s (20 mph). Only at the windiest locations do wind turbines operate for any length of time in wind speeds about 12 m/s (27 mph). Wind turbines installed in wind regimes where most people live spend most of their time operating in winds less than 12 m/s, generating less than rated power. British manufacturer LVM builds a line of micro and mini wind turbines. They have been building micro wind turbines under the Aerogen trade name with high power-density rare-earth magnets since 1985. Most of LVM’s multiblade turbines are used in the marine market, other applications include electric fencing. LVM’s micro turbines are used to mark the channel to Plymouth harbor in southwest England. Marlec Engineering’s Rutland brand micro turbine is a multiblade machine. Land-based versions use a furling tail while marine versions use a fixed tail. Marlec’s Rutland 913 uses a pancake generator design that seals the generator in plastic. This is an axial field generator, in contrast to the more common radial field designs used in automotive alternators or magnet-can alternators. The Ampair 100 is a small multiblade turbine. The shaft-driven


165

alternator has two, six-pole permanent-magnet rotors inside a cast aluminum body. The two six-pole stators are staggered 30 degrees to minimize cogging in low winds. The Southwest Windpower Air 303 has high performance airfoils that use blade flutter to limit the power in high winds. An industrial version uses cooling fins to keep the generator from overheating. The Air 303 uses three different windings. In low winds, a light winding with thin wire is used for power production. At higher wind speeds, the windings with fewer turns of thicker wire are used to deliver more power as it becomes available. Aerocraft is a German manufacturer that offers micro, mini, and home-size turbines from 120 watts to 5-kW. The smaller units use vertical furling to control rotor speed and the bigger models use pitch weights to regulate the rotor.

MINI TURBINES Mini wind turbines include Southwest Windpower’s Windseeker and Proven’s WT600. These turbines have rotors of 1.5-2.6 m (5-8.4 feet) and can produce 1,000-2,000-kWh per year at 5.5 m/s (12 mph) sites. World Power Technologies has a line of wind turbines from 500W to 4.5-kW. Each product comes in two versions; a standard twoblade model and a three-blade high wind version. The heavier rotor stays pointed into the wind longer before it begins to furl, allowing it to capture more energy in high winds than the two-bladed version. Their H500, H600 and H900 models use injection-molded, polycarbonate blades with fiberglass reinforcement. The bigger turbines by World Power use carbon fiber reinforcement in its composite blades. World Power uses an angle governor for control of the turbine in high winds. In the Bergey Windpower units, the blades are part of the generator case which turns around the stator. Bergey Windpower uses a rotor design where the blades are set at a high angle of attack for starting the turbine in low winds. As the rotor speed increases, a weight twists the flexible fiberglass blades. This changes blade pitch towards the optimum running position. As the rotor speed increases, centrifugal force stiffens the blades.

166


Home-size turbines are suitable for homes, farms, ranches and small businesses, and telecommunications. They range from 1-kW to 20kW. These wind turbines can generate from 2,000-kWh to 20,000-kWh per year at 5.5 m/s (12 mph) sites, like those on the Great Plains. The Proven machines use a direct-drive, permanent-magnet alternator. They use a flexible hinge between the blades and the hub that allows the blades to change pitch with rotor speed. Wind Turbine Industries’ turbines are large machines that can be used in utility intertie applications. Wind Turbine Industries uses designs and parts from the now defunct Jacobs Wind Energy Systems. The original Jacobs turbines go back to the 1930s-era. The new Jacobs units are radically different. The 23-foot (7-meter) model drives a 17.5-kW alternator mounted vertically in the tower with a hypoid gearbox. The 1100 rpm, four-pole alternator produces 40 hertz, three-phase AC, which is then rectified and inverted for delivery to the utility. Fiberglass blades are used which are made from a resin transfer process.

WIND POWER PROJECTS As part of an aid project in Indonesia, World Power and NRG Systems installed 50 Whisper 600s with NRG’s 45-foot (14 m) tower. The off-the-grid package was designed to provide 100-kWh per month at sites with a 12 mph (5.5 m/s) average annual wind speed. In North America, utilities in rural areas are connecting small wind turbines to the grid. Now that an extensive network of lines exists, utilities have found it is expensive to maintain them. Utilities on America’s Great Plains are finding that small wind turbines can offset some of the cost of upgrading or maintaining rural lines. At the University of Massachusetts, a wind generator 33 feet in diameter is mounted on a 50-foot steel pole tower. The electrical output of this machine is used to warm water, which is combined with heated water from solar collectors mounted on a south-facing wall. Concrete tanks in the basement house store this hot water. The fiberglass blades are capable of producing 50 horsepower (37 kilowatts) in a 26-mph wind. Shaft power from the blades drives a generator with an electronic load controller that senses the windspeed and rotor rpm and applies the required current to the field windings of the generator. This prevents overloading or underloading of the blades.


167

EUROPEAN WIND GENERATION The small turbine industry in Denmark evolved much the same way as it did in North America. The first turbine using an induction generator was connected to the grid in the mid-1970s. The small turbines of Danish manufacturers have gradually grown larger, becoming the core of the world’s commercial wind power industry. Hundreds of turbines from the 1970s and early 1980s are still operating in Denmark. More than 80% use induction generators and are interconnected with the utility. They represented almost 3 megawatts of capacity, about 5% of that installed in Denmark in 1998. About half of these early turbines are Kuriants, with a 12-meter (40-foot) three-blade rotor downwind of the tower. Most Danish turbines are upwind designs. Kuriants also use guyed towers, a feature rarely seen in Europe. In Denmark a wind system with cogeneration capabilities has been built at the Tvind School on the Jutland peninsula. The Tvind machine is rated at 2 million watts (2 megawatts) in a windspeed of 33 mph. The fiberglass blades on this machine span a diameter of over 175 feet. Each blade of the three-bladed unit weighs about 5 tons. The Tvind machine was designed to supply the school with its electrical and heating needs and to supplement the grid lines using a synchronous inverter. Alternating current from the generator is rectified to direct current and then reconverted to grid-synchronized AC for the school’s electrical system. The synchronous inverter is rated at 500-kW. The power from high winds will be used to heat water. The electricity is fed to coils immersed in a hot-water storage tank. This type of wind furnace is an application of wind power to raise the temperature of water for agricultural and industrial applications. Most of these applications require more heat when it is windy than when it is calm. Early Danish turbines were about the same size as those being installed in the United States: 7-kW, 10-kW, and 15-kW. In Denmark and since 1991 in Germany, wind turbine owners have been paid a fair price for the electricity they produce. Germany’s Electricity Feed Law, the Stromeinspeisungsgesetz, requires utilities to buy wind-generated electricity at 90% of the retail rate. Danish utilities are required to pay 85% of the retail rate. The Danish government also refunds a carbon dioxide tax collected on all electricity generated in the country. The result has been a steady growth of distributed wind turbines in Germany and

168


Denmark with several thousand wind turbines installed. In the future, small wind turbines will be accepted as common appliances similar to the way that heating and air-conditioning systems are purchased and installed today. High market penetration rates require that small wind turbines be designed to work effectively in low wind resource areas. These turbines need relatively larger rotors to capture more wind energy. But they must also be robust because even areas with low to average wind speeds can experience severe weather. These turbines must be extremely quiet, so that they are not heard above the local background noise. They should be able to operate for 10 to 15 years between inspections and/or preventive maintenance and they should offer a life expectation of a 50 or 60 years. Advances in small wind turbines include major improvements in small turbine manufacturing and more efficient installation techniques. The U.S. Department of Energy (DOE) and the National Renewable Energy Laboratory (NREL) have been accelerating the development and adoption of new small wind turbine technology and manufacturing techniques. Large wind turbines are in their seventh or eighth generation of technology development, while small wind turbines are only in their second or third. The industry is striving to reduce the cost of electricity generated by small wind turbines. In 2002, typical 5- to 15-kW residential wind turbines cost about $3,000 per installed kilowatt. These turbines produce about 1,200-kWh per year of electricity per kilowatt of capacity in a wind area with a DOE class 2 wind. By 2020, the installed cost should be 1,200 to $1,800 per kilowatt. Smaller systems are relatively more expensive. The productivity should rise to 1,800-kWh per installed kilowatt. The 30-year life cycle cost of energy will then be in the range of $0.04 to $0.05/kWh. This is lower than almost all residential electric rates in the country today. The attractiveness of small wind turbines to consumers depends on meaningful, cost-effective standards and certification programs for them. There have been some instances of exaggerated claims and the standards and certification programs for large wind turbines are not appropriate for small wind turbines. Appropriate standards for small wind turbines are under development. Related standards, such as electrical grid interconnection standards, are also needed and should reduce the costs of owning a small


169

wind turbine. The engineering of wind machines involves aerodynamics, structures, controls, electrical conversion, electronics and corrosion prevention. Government/industrial collaboration should take place at national laboratories and universities to provide better wind machines. Applied research is also being conducted at the facilities of small wind turbine companies with support from the government. Other applied research projects involves companies, universities, and national laboratories. This research and development by industry, research institutes, state and local governments and DOE will increase the contribution of small wind turbines to the power generation mix. In 2001, about 13,000 small wind turbines were manufactured in the United States. More than 50% of these were exported. It is expected that both the domestic and foreign markets for small wind turbines will continue to grow. It is estimated that small wind turbines could provide up to 8% of U.S. electrical needs by 2020. Wind may generate at 3% of U.S. electrical demand by 2020 and 6-8% of residential electricity demand. This will provide small wind turbines for a total generating capacity of 50,000MW. A study by A.D. Little sponsored by DOE projected almost 4 million small wind systems installed in grid-connected applications. A large market for small wind turbines is in rural areas where wind-generated electricity can reduce utility bills. In 2000, American homes used over one trillion kWh which was more than one third of total electricity sales. By 2020, there will be more than 15 million homes with 1/2 acre or more of land and sufficient wind to install a small wind turbine. If each of these homes installs a 7.5-kW machine, the total generating capacity would be more than 100,000-MW. About two million mid-sized commercial buildings could be outfitted with wind turbines of 10 to 100-kW. Many public facilities such as schools and government buildings could also use small wind turbines. Industrial and commercial customers who are connected to the utility grid can use wind power as back-up generators. Where the utility grid is not available, stand-alone or hybrid systems may provide electricity for homes, communities, water pumping, and telecommunications services. The Energy Information Administration (EIA) estimates that there are about 200,000 off-grid homes in the U.S. Many communities that are remote and isolated produce their elec-

170


tricity with diesel or gasoline generators. Alaska has over 90 villages powered by diesel generators, serving a population of over 40,000 people. Several hundred other remote facilities are powered by diesel generators that range from 2 to 250-kW. New wind-electric water pumping systems are available where the turbine can be located where there is good exposure to the wind and it does not have to be located near the well and pump. Remote broadcast facilities tend to use hybrid systems that combine generation from solar, wind, and diesel systems. These markets could contribute up to 25,000MW of generating capacity by 2020. The total installed capacity for small wind turbines in 2020 could be 140,000-MW. According to the EIA, the total generating capacity in the U.S. in 1999 was almost 750,000-MW and the projection for 2020 is over 1,000,000-MW of generating capacity and 4,800 billion kWh in demand. Fifty gigawatts (50,000-MW) of small wind turbines in 2020 would produce an estimated 132 billion kWh of clean electricity per year, or approximately 3% of projected total U.S. demand. At this level of capacity, small wind systems would be providing 6-8% of residential electrical demand. The EIA forecasts that the residential demand will be over 1,700 billion kWh by 2020. A growth in the domestic market from the current installed capacity of 15-18-MW to 50,000-MW in 2020 means a doubling of the market each year for several years and then a sustained sales growth of 50-55% per year. The domestic small wind turbine industry would then reach annual sales of $1 billion and employ approximately 10,000 people by 2020. Off-grid sites could contribute up to 25,000-MW of generating capacity by 2020. References Gipe, Paul, Wind Energy Basics, White River Junction, VT: Chelsea Green Publishing Company, 1999. Park, Jack, The Wind Power Book, New York: Van Nostrand Reinhold, 1981. Halper, Mark, “More Power to You: Alternative-energy Technologies Could Soon Give Your Phone, Your Car, and Your House their Own Microgenerators,” Time, Vol. 162, Bonus Section, January 2004, i24, p. 10. Winebrake, Ph.D., Editor, Alternate Energy: Assessment and Implementation Reference Book, Lilburn, GA: The Fairmont Press, Inc., 2004.

Distributed Generation, Clean Power and Renewable Energy

171

Chapter 6

Distributed Generation, Clean Power and Renewable Energy The California energy crisis of 2001 transformed the state’s outlook on energy. Suddenly you saw terms and issues like stranded costs, renewable portfolio standards and exit fees in the daily papers. The regulatory atmosphere also changed, electricity transmission and capacity constraints made state public utilities commissions and utilities start to promote self-generation. Incentive programs in several states have been started or expanded and customers are more willing to consider technologies previously considered too expensive or complex.

CHP AND PV Historically, combined heat and power (CHP) systems were considered only for very large customers or a few very specific facility types such as hospitals and municipal swimming pools. Solar PV systems and fuel cells were not considered at all. This attitude has changed as the new energy climate emerges. The terms used to describe these systems include customer-sited generation, self-generation, distributed generation, distributed resources, distributed energy, combined heat and power, cogeneration, renewable energy, clean power, and green power. Self-generation (SG) can be used to describe any technology that is sited at a customer’s facility and produces power primarily or exclusively for use on-site. This term is used in the California incentive program and is less variable than some of the other terms. The SG market ranges from 1-kW residential PV systems to 40-MW 171

172


gas turbine cogeneration plants serving universities. Systems of about 50-kW to about 1-MW are large enough to be cost-effective and small enough to be appropriate for many end-users. The primary technologies in this area are PV, reciprocating engines, fuel cells, and microturbines. Most of the CHP activity has historically been in facilities with large electrical and thermal loads. In these large systems, the primary technologies are gas turbines, large reciprocating engines and steam turbines. The market for commercial-scale PV is growing rapidly. Powerlight is a large-scale PV systems integrator with grid-connected PV installations that have seen an annual grown rate of 55% for the last 5 years. The CHP market in the 50-kW to 1-MW range may have contracted slightly over the last few years. Customer concerns over high gas prices, combined with the economy, may be causing customers to postpone decisions. SG activity in California applications in the California Self-Generation Incentive Program is summarized in Table 6-1. The total capacity of projects that were active as of January 2003 is 105-MW. New York provided incentives in 2001-2002 that supported almost 16-MW of CHP installations of 1-MW and below. Almost 1-MW of PV systems larger than 10-kW were installed under the New Jersey Clean Energy Program. Fuel cells have been a much smaller part of the SG market. Table 6-1. California Self-generation Projects ———————————————————————————————— Complete In Work Average Size (kW) ———————————————————————————————— PV 21 168 170 Fuel Cells 1 3 400 IC Engines 7 119 540 Microturbines 5 50 200 ———————————————————————————————— The California SG program is open to systems between 30-kW and 1-MW. ESCOs are interested in projects that can be installed profitably, and these projects tend to be large. The typical PV project size is increasing. A 300-kW PV installation is no longer an unusually large system with a capital cost of about $2.5 million.


173

The PV industry will continue to grow with estimates that the gridconnected PV market will exceed $3 billion by 2010. The PV industry goal is 25% annual growth in the total capacity of panels produced domestically. According to the Energy Information Administration, there is a market for 17-GW of building-sited CHP by 2010. About half of this is in existing facilities as opposed to new construction with installations of less than 1-MW. Office buildings are nearly half of this estimate, with hospitals and colleges accounting for the rest building site applications. Much of the growth in office buildings involves absorption chillers. The Federal Energy Management Program (FEMP) estimates that 9% of federal facilities could install CHP systems with an average simple payback period of 7 years. A CHP market study for New York showed the potential for almost 20,000 new CHP installations in commercial facilities with systems less than 1-MW. The total project capacity was about 3-GW in this size sector. The projections are about 200-MW in sub-1MW commercial installations by 2012. This was about 1% of the installed base of CHP installations in the state, or about 50-MW. Major barriers to self-generation include interconnection, air pollution permits, building permits and the installation of net meters.

INTERCONNECTION Interconnection remains an important barrier in California in spite of statewide efforts to streamline the process. In states without interconnection standards, an inflexible local utility can make it nearly impossible to install any type of SG. Charges for standby power and exit fees can be a problem. New York, negligible growth in systems less than 500kW is predicted, but significant growth is expected if regulatory barriers are lowered. Net metering laws are more of an issue in residential systems. Compared to commercial, in residences, the time of peak solar output may correspond with low building loads since no one may be home. Some states have net metering laws that apply to all systems but many apply only to residential-sized systems. In spite of the barriers, significant SG activity is occurring. The

174


reasons that customers install SG projects depend on the type of technology. In California, a survey to determine the factors that were influential showed that savings were the major reason. Table 6-2 shows average responses on a scale where 10 is most influential. Table 6-2. Self-generation Factors ———————————————————————————————— Savings Environment Image ———————————————————————————————— PV 4 4 3 Fuel Cells 4 5 4 IC Engines 5 3 3 Microturbines 5 3 2 ———————————————————————————————— Reciprocating engines and microturbines are not viewed as being as green as PV and fuel cells. Reciprocating engines and microturbines are installed primarily because they are a cost-effective way to reduce utility bills. Concern for the environment ranked higher than improving the image of the business for all but reciprocating engines.

INCENTIVE PROGRAMS Incentive programs are important in the increased activity of commercial scale PV projects. Several states have programs for residential systems and a smaller group have commercial programs. Commercial sized programs are offered by California, Connecticut, Delaware, Illinois, Massachusetts, New Jersey, New York and Rhode Island. Program rules, incentive levels, and funding availability change frequently. The California Self-Generation Incentive Program was authorized to spend $125M annually through 2004. Systems between 30-kW and 1MW are eligible. The incentives include up to 50% of costs, approximately $4/watt, a 15% state tax credit for systems of less than 200-kW and standby and exit fees waived. An additional $100M was available through a separate program for systems less than 30-kW. The New Jersey program had a tiered rate structure where smaller systems received a higher incentive rate than larger projects. The incentive program covered up to 50% of project cost with the following tiered structure:


• • • •

$5.50/watt $4.00/watt $3.75/watt $0.30/watt

for for for for

175

the first 10-kW the next 90-kW the next 400-kW 500-1000-kW

Oregon has two incentives available. The Energy Trust of Oregon provides an incentive for commercial PV systems less than 25-kW at $1.75/watt, but it caps out at only $20,000. There is also a Business Energy Tax Credit (BETC) for 35% of the installed cost. The BETC program has a pass-through option where tax-exempt entities can pass the credit to a partner that has a tax liability. New York has several programs for PV with incentives of $4 to $5/ watt for 15-kW installations. There is also state tax credit for some installations and a new construction incentive for building-integrated PV. Tax credits supplement cash incentives from utilities and include a Federal Investment Tax Credit that allows customers to take 10% of the installed system cost against their federal tax burden. Several states also have tax credits, with Oregon being the highest. The federal government has been the main supporter of fuel cell projects. In energy efficiency and PV most direct support has been at the state and utility level while federal support was largely limited to research and education. The Climate Change Fuel Cell Program provides up to $1.0 per watt and 33% of project cost. States that have programs to support fuel cells, include California, Connecticut, Illinois, Ohio, Massachusetts, Michigan, New Jersey, New York and Oregon. The California program offers an incentive of $2.50 per watt up to 40% of eligible costs. The New Jersey Clean Energy program originally offered incentives higher than California’s for fuel cells powered by natural gas but changed their program to cover only fuel cells operating on renewable fuels such as landfill gas. Incentive programs for technologies other than PV and fuel cells are few. There is some support for reciprocating engine and microturbine CHP in California, New York, and Oregon. California provides incentives of $1.00 per watt that cover up to 30% of the cost of microturbines and reciprocating engines running on natural gas. Oregon has a Business Energy Tax Credit that allows a credit for 35% of the installed cost of systems that exceed 56% total efficiency. This is not the standard PURPA efficiency that includes only half of the thermal output. New York provides incentives for CHP

176


demonstration projects of up to $1M per installation that can cover up to 50% of the project cost, however these projects are selected in a competition. Reciprocating engine and microturbine projects require less support in order to be cost-effective. Even without incentives, payback periods can be short in CHP applications where a large percentage of the waste heat is effectively recovered. This is less true of fuel cells. Table 6-3a shows the possible incentives for a PV 100-kW array installed on a flat office building, a popular PV application. The table compares the installation in California, New Jersey and New York. The customer is a for-profit enterprise with a tax liability. The installed cost is $8,000 per kW with a projected electricity price increase of 2% per year. See Table 6-3b. Table 6-3a. 100-kW PV Incentives ———————————————————————————————— STATE FEDERAL NET Incentive Tax Credit Tax Credit Cost ———————————————————————————————— California $400,000 $54,000 $40,000 $306,000 New York $500,000 $13,500 $30,000 $256,500 New Jersey $415,000 None $38,500 $346,500 ———————————————————————————————— Table 6-3b. 100kW PV Costs ———————————————————————————————— Projected First Year Simple Net Present kWh/year Savings Payback Value ———————————————————————————————— California 162,000 $31,000 10 years $81,000 New York 108,700 $12,300 20 years $ 8,900 New Jersey 113,500 $ 9,200 38 years $(88,000) ———————————————————————————————— While the simple payback period can be long, the Net Present Value of the investment was positive in both California and New York due to the life of the equipment and its depreciation. The NPV goes positive for the New Jersey case if the cost goes to $6,500 per kW and the electricity price changes to 3.5% per year. New York has less savings than California in spite of New York’s more generous incentive due to less sunshine and lower electricity rates in New York. All three projects


177

included large incentives and this shows the economics are difficult in states without incentives. PV projects do not have the short paybacks when compared to 4year payback lighting projects. The benefits used to commonly sell PV are utility cost savings, reduced exposure to price increases, reliability, low maintenance, no emissions and an output coincident with the highest power prices. There is also some public relations value and the potential to fulfill some organization goals in self sufficiency for power outages. Many corporations and public agencies have established sustainability goals and policies. Johnson and Johnson has a goal of reducing its carbon dioxide emissions to a level 7% below their 1990 levels. Cities like San Francisco and San Diego have made commitments to increase their use of renewable energy. Installations include PV systems at fire houses and a 1-MW system on the roof of the convention center in San Francisco. PV is viewed as more of a contributor to sustainability goals than energy efficiency measures or CHP. With their low emissions, fuel cells are viewed as almost as green as PV. PV provides a way to reduce exposure to price volatility. Installing a PV system represents a way to prepay a part of future electricity use at a known price. The volatility in the total utility cost for the facility is thus reduced. The generating nature of PV encourages buyers to think in terms of the price of generated power. Dividing the installed cost of a PV system by the lifetime energy production yields costs of produced power at less than $0.10kWh. This price compares favorably with the current cost of utility power and is a reasonable price to lock-in for a portion of electricity needs. SG projects can be analyzed using the following economics: simple payback, cost of produced energy, net cash flow forecast, life cycle cost and net present value. Most rely on cash flow projections and estimates of the cost of produced power. The cost of produced power is especially helpful in marginally cost-effective projects.

PV AND EFFICIENCY Many projects combine energy efficiency and PV. There are several reasons for this approach. A long payback item like PV is aided by

178


packaging the measure with shorter payback measures like lighting to improve the economics of the project. Many of the costs in a performance contracting project are independent of the number of energy conservation measure (ECMs) and the cost of the package. Adding a SG project to an energy efficiency-based performance contract increases the total value of the project. Longer loans terms are used for PV, due to the longer expected service life of the equipment. With 25-year warranties on the PV panels, a 15-year loan term is not unusual. Increasing the loan term of a combined efficiency and PV project over 15 years offers real cash flow benefits. Consider an energy efficiency performance contract project with a $1M installed cost, 4 year simple payback and a 10 year loan. Now, add a 100-kW solar PV system and extend the loan term for the combined project to 15 years. Without considering the tax implications, the economics are summarized in Tables 6-4a and b. Table 6-4a. Combined Project Payback ———————————————————————————————— Annual Initial Simple Savings Cost Payback ———————————————————————————————— Efficiency $200,000 $1,000,000 5 PV $12,400 $256,500 21 Combined $212,400 $1,256,500 6 ———————————————————————————————— Table 6-4b. Combined Project Cash Flow ———————————————————————————————— Annual Loan Payment Net Cash Flow (15 year loan) ———————————————————————————————— Efficiency $147,400 $ 76,600 PV $ 26,200 $ (13,900) Combined $138,000 $ 106,900 ———————————————————————————————— The net cash flow is higher in the combined project than in the efficiency project alone. This occurs in spite of the long simple payback


179

period of the PV project. The improved cash flow results from the longer loan. PV prices are falling as production increases. The installed cost of a system is now only about 10% that the cost was in 1975. There are estimates that the installed cost of PV systems will drop by 50% in 6 years and by 75% in 12 years. These decreasing costs will allow 10 year paybacks on PV projects in California without the current 50% incentive and can drop post-incentive paybacks to 10 years in many states. SG technologies are also helped by the ongoing efforts to reduce the time required for interconnection and permitting.

MODULAR GENERATION Modular systems are available and used in the Unites States and internationally. Advances should occur in modular systems and distributed small-scale generation of less than 1-MW. These systems should be cost competitive as advances occur to reduce their capital costs. Systems should be developed that can consume small quantities of organic waste or dedicated resources for the distributed generation of power and heat locally for use on-farm, on-site and in small industrial systems. These alternatives could include the integration of modular biomass systems with fuel cells, microturbines and other distributed systems. Fuel sources include food/feed/grain processing plant residue, fats and oils, nutshells, corncobs, tomatoes, carrots, fruit, rice hulls, as well as uncontaminated urban wood residue and farm animal waste. Significant development in scaled-down, skid-mounted or mobile installations and fuel concentrators may take place to increase energy density. Significant opportunities for modular systems exist in low value byproducts from grain, soy, wood and other processing systems, and in farm and forest residues where the high cost of transporting biomass to larger facilities may be avoided. Rural communities and farmers could benefit if modular systems can be deployed to offset power costs in gridbased systems. Industry standards for grid connection should be simplified and new standards developed so that modular biomass systems can be easily connected to the grid. The waste biomass from any biorefinery that has no other value will be able to be converted into electricity.

180


BIOCONVERSION Economically viable and environmentally sound bioconversion processes and technologies allow the commercial application of a wide range of biobased fuels and products. Advances in biochemical conversion processes will increase the variety of biofuels and biobased products that are cost-competitive and produced from biomass resources. The conversion of multiple sugar streams and lignocellulosic materials could provide useful fuel and value-added products. The development of enzymatic pre-treatment methods would do much to increase the efficiency of biofuels production. Bioconversion research will take place in two general categories: processing and conversion. Improved methods and technologies for processing biomass feedstocks will increase the economics and capabilities of bioconversion systems. Improving the physical and chemical pretreatment of biomass feedstocks prior to fermentation may include new enzymes and new methods for enzyme pretreatment. Traditional agriculture and forest crops, urban waste, and crop residues are a major source of readily available complex proteins, oils and fatty acids as well as simple and complex sugars that can be used as raw materials. These materials are available at low cost across the United States. The development of low-cost chemical and biological processes will include new chemistry and thermochemical synthesis that can break down these molecules and separate the resulting components into purified chemical streams. Residual biomass resources exist in the form of plant, animal, and other residues. These residues can be used to develop value-added fuels, chemicals, materials, and other products. Additional research should result in cost effective methods for processing solid and liquid residues into economically viable biomass resources. New cost-effective methods of chemical/enzymatic conversion should allow greater utilization of biomass resources.

BIODIESEL Catalytic and chemical methods for converting vegetable oils and animal fats into biodiesel are currently in use. It will be necessary to


181

improve the efficiency of these processes, to develop new processes, and to make processes more cost-competitive with non-biobased products. Additional research is needed to conquer the barriers associated with inhibitory substances in sugar streams. Methods to enable removal of catalytic inhibitors should be developed and new catalysts developed. Research on engineering and biological principles should improve feedstock separation and product purification. Biomass fermentation and hydrolysis research should increase the fermentation and hydrolysis of fiber, oil, starch, and protein fractions of crop components and processing by-products. Along with more rapid conversion of cellulose to a fermentable substrate, there is a need to develop new fermentation technologies to enable production of base chemicals and chemical intermediates from the wide range of existing crop components.

SYNGAS Syngas fermentation research should improve the catalytic synthesis of gases to chemical as well as to improve pyrolysis to produce chemicals. Processing systems should optimize both mass transfer of oxygen and nutrients for bio-organisms and fermetor environments. Biorefinery integration advances will allow biorefineries to efficiently separate biomass raw materials into individual components, and convert these components into marketable products, including biofuels, biopower and conventional new bioproducts. Biorefineries exist in some agricultural and forest products facilities including corn wet milling and pulp mills. These systems can be improved through the better utilization of residues. New biorefineries may be enhanced from lessons learned from existing facilities. Biorefineries could become markets for locally produced biomass resources and provide local and secure sources of fuels, power, and products. Optimized biorefineries will use complex processing strategies to produce a diverse and flexible mix of conventional products, fuels, electricity, heat, chemicals, and material products from biomass. Further development and deployment of the biorefinery concept for local and regional markets will take place. Additional utilization of existing biomass processing and conversion facilities will result in the development of more biorefineries. The development of new cost-competitive biomass technology

182


platforms will result in additional biorefinery products including the bioconversion of sugars to products such as polyols and other products that can be used to produce chemicals, materials, or other biobased products. The development and commercialization of the conversion of vegetable oils will produce hydraulic fluids, lubricants, and monomers for use in plastics, coatings, fibers and foams. A biodiesel/bioproducts biorefinery provides alternatives to petroleum-based chemicals, polymers, plastics and synthetic fibers. Alternatives to petroleum-based additives in the polymer industry include dyes, stabilizers, and catalysts. Rural-based biorefineries should be modular and produce high value products. Residual waste from the biorefinery would be converted into electricity and useful heat.

BIOMASS POWER By 2020 electric utilities will use biomass to generate power at four times current levels, providing 5% of all energy use in the industrial and utility sectors. Most landfills will be tapped for natural gas which will be used to generate heat and power for homes, schools, and industry. There will also be an integration of conversion systems with power generating equipment along with increased capabilities to convert low quality gas into electricity. New technologies will allow the more efficient production of biofuels, with less reliance on agricultural products and more on grasses and woody plants. Improvements in biomass gasification technology will allow the conversion of a wide range of feedstocks including residue biomass. Ten percent of the transportation fuels used in the country may be derived from biomass. Vehicle fuel includes biodiesel and ethanol developed from biomass like soybean oil, corn oil, switch grass, and other woody plants. Biomass gasification technologies must improve their cost competitiveness with other technologies. Improved operating efficiencies are needed with a wider range of sources, such as forest and agricultural residues. Gasification should be integrated with generating turbines and biorefineries. About 20% of the chemical commodities produced in the U.S. will


183

be biobased products. Companies will not be as reliant on petroleum resources to produce some chemicals. Direct combustion of biomass is currently in use but improvements could be made to improve operating efficiencies. Co-firing will be used in more industrial facilities and includes fossil fuel replacement and reduced environmental effects. Co-firing provides near-term demand for biomass feedstocks that will help develop the infrastructure needed to produce stand-alone biomass electric generation facilities and integrated biorefineries. Demonstration projects show co-firing to be a viable option although its use is not widespread. Improved operating efficiencies will increase use in industry. Improvements in technology and demonstrations from forest products will increase the spread of co-firing technology.

ANAEROBIC FERMENTATION Power and fuels can be produced from anaerobically generated gases. These include landfill gases, anaerobic digestion of animal manure and food/feed/grain products and by-products, wastewater treatment gas, sludge and sewage gases and other sources. The methane from biomass waste is a powerful greenhouse gas, with a global warming potential 21 times that of carbon dioxide. Over 600 million tons of carbon equivalent methane are produced annually in the United States. Greater application of anaerobic fermentation is needed. Reduced capital costs and improved operating efficiencies should increase the use of these systems. Low intensity methane could be viewed as a resource instead of a waste product. Systems for the use of methane from 10-300 Btu/cu feet are feasible and should be developed and demonstrated.

COMMERCIAL BUILDING TECHNOLOGY Today’s commercial buildings use diverse technologies in their construction, operation, and maintenance. A whole-buildings approach is used where all of the building components and sub-systems are considered together with their potential interactions and impact.

184


The goal is to optimize the building’s performance in terms of comfort, functionality, energy efficiency, resource efficiency, economic return, and life-cycle. The whole-buildings approach requires the integration of planning, siting, design, equipment and material selection, financing, construction, commissioning, and long-term operation and maintenance. The whole-buildings approach can enhance air quality, lighting, and other aspects of the building’s indoor environment. The natural environment benefits through energy and waste reduction and more effective land use. About 4.5 million commercial buildings in the United States account for over one-sixth of total national energy consumption, (16 quadrillion Btu) and one-third of total national electricity consumption. The consumption of electricity in commercial buildings has doubled in the last few decades and can be expected to increase by another 25% by 2030 if current growth rates continue. Commercial buildings represent a great opportunity to save money and reduce pollution across the country. Annual energy expenditures in commercial buildings exceed $100 billion. Benefits to the environment include reduced emissions of sulfur dioxide, nitrogen oxides, and carbon dioxide from fossil-fuel power generation. A 30% improvement in energy efficiency could be in the coming decades by applying existing technologies. More dramatic improvements ranging from 50 to even 80% could be achieved with new technologies.

COMBINED HEATING, COOLING AND POWER SYSTEMS These include advances in combined heating, cooling, and power systems, optimized building controls, solar and other forms of renewable energy. Energy-efficient building shells and equipment could produce commercial buildings that are net electricity generators rather than consumers. Tomorrow’s high performance commercial buildings are more likely to incorporate smarter, more responsive technologies. Commercial buildings will use smart materials and systems that sense internal and external environments, anticipate changes, and respond using the whole-buildings approach. Wireless sensors and controls will monitor


185

energy use and adjust operations accordingly. More individualized control of lighting, ventilation, and thermal conditioning will be used with stored user profiles that specify personal environmental preferences. The control will follow an individual through a building or group of buildings. More uniform protocols will allow control devices to talk to each other and communicate externally. Buildings will use performance information to self-diagnose and correct problems, and alert users to causes of marginal operation. More sound environmental practices will allow commercial buildings will be resource-efficient and will make more use of environmentally sustainable materials. The buildings will operate more efficiently, using 30-80% less energy than 20th century buildings. Some will be net electricity exporters, generating their own power using on-site technologies such as fuel cells and photovoltaics, and supplying excess power back to the grid. Sunlight will be used to produce electricity as well as for daylighting. Passive solar construction and natural ventilation will be incorporated into buildings designed for greater flexibility and adaptability to reuse, resulting in longer life. Components and materials will be designed for more complete recyclability at the end of their lifetimes. Commercial buildings will become more closely integrated with the surrounding environment. Building philosophy will shift from single, stand-alone buildings to communities. Resource management will be optimized across the entire community using strategies such as distributed power generation. More building space will double as both commercial and residential space. Fewer but better buildings will need to be constructed as a result. Communities will benefit from better land and resource use with lower investments in highways and transportation. By 2020 commercial buildings will use dynamic envelopes that can respond to changing environmental conditions. Microscale thermal conditioning can be used with individually controlled user’s preferences. Dynamic, personalized ventilation systems will utilize plug-and-play components and systems. Solid-state lighting will be used with dynamic level changing and more daylighting. Distributed energy resources will include photovoltaics, fuel cells, combined cooling, heating, and power generation. Digital wireless microsensors will personalize building controls. The focus of building finance will become long-term, taking into account life-cycle benefits.

186


BUILDING EFFICIENCY By 2020 windows should become controlled to operate with the surrounding building environment. These windows will be more energy efficient, using gels and other new materials to improve insulation. They will be able to sense energy loads in the building exterior and interiors and adjust their insulation properties based on energy needs. Holographic techniques will allow windows to direct the outside light to particular areas of the interior space, replacing artificial light with natural light. Windows will use embedded photovoltaic technology to keep the energy balanced in a building while generating power from solar energy. Windows will also provide imaging for displays. These displays could be used for signs indicating sales, weather conditions or news. Windows will be more durable and shock proof. They will be able to stand up to serious environmental changes. They will also be more adaptable and modular, so they can be substituted with newer advanced products more easily. Windows will be produced in a more environmentally responsible manner and designed for recyclability, modularity and upgrading. Building spaces will be more adaptable by 2020. The external envelope will incorporate more adaptability and flexibility. Rooms will convert more easily from one use to another with modular components that allow for movable walls. By 2020, buildings will minimize heating, cooling, and lighting loads using integrated design and non-polluting energy sources which will return excess electricity to the grid. This should save money, and reduce emissions, brownouts and blackouts. By 2020, buildings will have enhanced air quality and airflow with natural ventilation and lighting. Intelligent features will increase the adaptability and energy efficiency of the building. These features will allow the use of light and energy only when and where they are needed. In Manhattan at the intersection of Broadway and 42nd Street, is the first Manhattan office tower to use green standards such as energyefficient design, sustainable materials, and on-site power generation. This 1.6 million-square-foot 48-story building generates some of its electricity with on-site fuel cells. These natural gas fuel cells run cleanly and quietly 24 hours a day. No combustion is required and the waste products are water vapor and CO2.


187

The building has integrated photovoltaic (PV) panels on some areas of the facade. The peak output is about 15kW, enough power to operate five homes. DOE-2 software is employed for analyzing the building’s energy. It models and compares the energy savings using a variety of options. The ventilation system provides tenants with 50% more fresh air than required by code. The reduced energy costs are estimated at $500,000 annually with a payback period of five years or less.

LIGHTING ADVANCES Lighting in the future will allow more effective use of space for multitasking, so businesses can adapt workplaces currently designed for individualized, manual operations into an environment with shared resources and electronic processes. High-quality lighting systems improve employee productivity, employee retention and quality control. Advanced lighting solutions will improve health, safety, and security in the workplace and provide savings in energy consumption. Advances in lighting will give building owners and managers higher returns on capital investments. More efficient, intelligent lighting systems will be networked and allow managers greater control over building functions, minimizing operations, maintenance, and energy costs. Sensors and controls in future lighting systems will provide new levels of information about our environment and will allow us to shape that environment to improve our creativity and productivity. Advanced lighting systems will include more powerful and cost-effective sensors and controls, wireless connectivity and high-efficiency light sources. Lighting will be done in an integrated whole buildings approach that optimizes daylight to provide more efficient, high-quality lighting, heating, cooling and ventilation. By 2010, luminaries will become smarter and more integrated, communicating with control systems, performing self-diagnostics, and allowing preventive maintenance. New materials will make reflectors configurable and more integrated with the light source. Microelectronics will be used in smaller, more flexible ballasts. Sensors will provide multiple inputs to define the lighting environment for users. Controls will work with the larger building management system to optimize daylighting, thermal load management, preventive maintenance and demand load shedding. Solid-state LEDs and organic light emitting

188


polymers (LEPs) will be used. Fluorescent sources will reach efficiencies approaching 200 lumens per watt while maintaining a high color rendition index (CRI) through the use of new two-photon phosphor coatings. Low-cost ballasts will increase the flexibility of the systems and make compact fluorescent lamps more common. By 2020 the design of building systems will combine both natural and human-made lighting systems. Newer technologies will capture daylight for later transmission and distribution. Programmable flatpanel luminaries will create theatrical-type effects using advanced control systems. More efficient, reduced-mercury fluorescent sources will become available and incandescent lamps will use advanced materials that will raise their efficiency to 60 lumens per watt.

SUSTAINABLE DEVELOPMENT Sustainable development is impacting the urban environment and is relevant to energy planners, engineers and architects. Many physical examples of sustainable development exist and the role of sustainable development is becoming a guide to planning at international, national and local government levels. Sustainable development can be a guiding vision and planning model to shift the practice of dominance by special interests to a more holistic and distributed energy world. Aspects of sustainability include urban development, population trends, environmental impacts and energy concerns. Incorporating alternative energy technologies is critical to the success of urban sustainability. Visionary ideas for guiding the development of towns, cities and regions in the 1980s and 1990s started to include two evolutionary approaches, the New Urbanism and sustainable development. Both approaches attempted to address concerns about equity issues and the conservation of environmental resources. The New Urbanism attempted to recapture the urban sense for locale and community by physically reorganizing neighborhoods, improving pedestrian access, revising transportation patterns and reintroducing mixed use development. The focus was on providing alternatives to suburban sprawl. Sustainable development has a broader vision and addresses a wider range of concerns.


189

In 1987, the report of the World Commission on Environment and Development, brought sustainable development into the forefront. The report was later referred to as the Brundtland Report. It defined sustainable development as meeting present needs without compromising the needs of future generations. At the United Nations 1992 Conference on Environment and Development (also called the Rio Summit or the Earth Summit) representatives from 167 nations, including the United States, produced the Rio Declaration Environment and Development which is referred to as Agenda 21. The Agenda 21 charter deals with policies involved in reducing unsustainable patterns of production and consumption and promoting environmental protection. While energy usage is a primary contributor to global pollution, the term energy is not specifically mentioned in Agenda 21. However, the term resource is used. In the European Community’s (EC), Sustainable Cities Agenda, the principle of ecosystems thinking emphasizes the city as a complex system incorporating aspects such as energy, natural resources and waste production. The EU suggests that cities must be viewed as complex, interconnected and dynamic systems. Sustainable development has links to global issues and balance. Future generations need to reproduce and balance local social, economic, and ecological systems, and link local actions to global concerns. This includes the efficient use of natural resources. Non-sustainable development implies growth that is environmentally unsafe and consumes resources inefficiently. Implementing sustainable development at the local level is often counter to existing planning regulations. If conventional real estate development prohibits the construction of mixed-use neighborhoods, then it becomes very difficult to build these projects. Today, we have a separation of uses in most newer areas. Shopping is separate from housing and schools. Vast areas are used for employee and customer parking. New buildings, transportation systems and power distribution systems are required to meet growing demands. As new facilities are constructed to meet urban requirements, energy usage should be considered in the planning process. This type of infrastructure planning is usually performed by the utilities, often resulting in higher costs for ratepayers. Techniques such as using improved infrastructure technologies and more creative design approaches are

190


needed. Urban expansion affects world energy consumption since it requires more energy to provide critical urban services.

GROWTH TRENDS AND ENERGY Rapid population growth on a global scale has increasingly placed a growing burden on our resources. Over half of the world’s population now lives in urban areas, gaining over one billion in population the last 30 years. In 1970 in the U.S., more people lived in the suburbs than in either urban or rural areas. But by 2000, 99% of the 153 million increase in (U.S.) population from 1930 had occurred in the nation’s metropolitan areas. Las Vegas, Nevada, a city that may have the fastest growth rate in the U.S., grew from 273,000 in 1972 to 1,376,000 in 2000. More efficient use of energy can have a major impact in housing solutions, transportation systems and environmental problems. Resource conservation permits the opportunity to grow without constructing additional power generating facilities while reducing environmental impact. Power plants and vehicles account for a major portion of sulfur oxides, carbon and nitrogen emissions into the atmosphere. International efforts at mitigation of environmental impacts include the Kyoto Protocol. While the 15 members of the European Union (a total of 87 countries worldwide) have ratified the Kyoto Protocol, the U.S. has resisted, saying that implementation could cost the country up to $400 billion and 5 million jobs. However, this analysis fails to consider the employment that would have been created by efforts to comply. Energy usage is increasing worldwide. The U.S. is the world’s largest energy consumer. From 1970 to 1996, total energy consumption in the United States grew from 68 quadrillion Btus to 94 quadrillion Btus. Energy from renewable sources grew from about 3 quadrillion Btus to 7 quadrillion Btus. In 1999, total U.S. energy nearly reached 100 quadrillion Btus with transportation using 26 quadrillion Btus. Energy usage is highly decentralized while energy generation and production tends to be relatively centralized. Rapid increases in population and increases in conditioned space are major causes of increasing energy use. The need for highly conditioned space has developed into new standards for human comfort, especially in the workplace. More efficient use of energy in the built environment can have a significant


191

impact on reducing energy needs. This includes the ability to provide for growth without constructing additional power generating facilities. While technologies are available to provide more efficient use of energy, economically viable technologies are often not implemented. Power production and power use have urban and regional impact. Issues include power plant construction, lack of flexibility from multiple energy sources and improved technologies for designing and building facilities that are extremely efficient. The U.S. is experiencing an energy crisis. Many of our buildings waste energy and incorporate few conservation or reclamation features. In spite of improved design standards, many newer buildings use more energy than older ones. This is due to the design, location, and construction technologies employed and equipment utilized. Changing standards for fresh air admission into occupied space increases ventilation air. More energy is expended in producing occupancy air (cooled or heated, humidified or dehumidified) from unconditioned air. A wide range of solutions needs to be employed to solve the complex problems associated with increased energy usage. These approaches must include both supply (production) and demand (consumer usage). Usable resources are available to be utilized but everyone shares the responsibility of not wasting or misusing usable resources. Actions include education and training in the management of resources.

SUSTAINABLE DEVELOPMENT PROGRAMS Sustainable development programs include the Florida Sustainable Communities Network, DOE Efficiency and Renewable Energy Network, Civano Sustainable Community Project, Lake Tahoe Regional Plan, Manchester VT Planning and Zoning Program, and Santa Monica, CA, Sustainable City Program. Alternative communities include Seaside, Florida and Laguna West, California which use higher density, mixed-use development. Arcosanti, Arizona, is an ecologically based experimental community dedicated to alternative energy, concentrated development and a reduced role for the automobile. Kishio Kurokawa’s eco-media city is an advanced version of an eco-city proposed in a Futian, China, city center plan. It is an eco-media city park concept for the information age to

192


demonstrate urban sustainability. The town of Navarra, Spain, has a goal of providing 100% of its electrical energy by renewable energy sources by 2010. The Building Act in Finland establishes sustainable development as the foundation for land use planning. In the United Kingdom, Leicester, Leeds, Middlesborough and Peterborough have achieved the Environmental City designation as a result of their planning efforts. Ecolonia in the Netherlands is a demonstration town for ecological development. The 1997 Treaty of Maastricht included sustainable development as an objective for European Nations. Sustainability programs have provided incentives for wind-farms in Austria and pushed German automotive and electronic manufacturers to develop new ways to recycle components and improve energy resource management. Local codes and ordinances often punish the installation of sustainable technologies. One argument against implementing sustainable or alternative technologies is that investments may fail a simple payback test or some other economic standard. Investors often need a given guaranteed rate of return in order to implement new technologies. Part of the action plan of the World Summit on Sustainable Development included the use of renewable energy technologies to be increased to 15% of worldwide energy production by 2010.

DISTRIBUTED GENERATION AND ENERGY MANAGEMENT Applying energy management systems to distributed generation provides additional cost reductions. Energy consumption reductions are usually derived from strategies such as temperature resets in air and water handling systems, chiller plant optimization, HVAC scheduling optimization, and automation of lighting system operation. These activities represent a significant potential for energy savings. In the era of deregulation, electric utility rates include RTP (real time pricing), and interruptible service (IS). On-site power generation systems have been utilized by many facilities to take advantage of these rates to reduce electric costs. Energy management systems are also used to manage the operation of on-site power generation systems for power outages and to improve power quality. One example of a typical installation involves a 2,000-kW on-site power generation system and its integration with a facility energy man-


193

agement system to exploit an interruptible electric service rate. The system was installed at a 600-acre corporate campus site in central Pennsylvania. There were 28 buildings on the site, enclosing about 1,000,000 square feet of conditioned space. Of the 28 buildings, only one, the corporate data center, does not use the centralized steam plant for heating and a 2,500-ton chilled water plant for cooling. Combining the capabilities of the EMS (Energy Management System) and on-site power generation systems allows the facility to interrupt more than 70% of its on-peak summer electric load, within 2 hours after a request, using a minimal staff and causing a minimal change to those at the site. The system reduced the annual electricity costs by more than 25% since its implementation in 2001. The peak electrical demands occur during the summer months, usually during July or August. Peaks of 4,800-kW have been recorded on days when air conditioning loads peak. Electrical power is purchased from the local utility at 13,800 volts and distributed to 21 different substations. The substations vary in size from 500-kVA to 3,750-kVA and are both indoor and outdoor types. Secondary power distribution voltages of 4,160 volts, 480, 277, and 208/ 120 volts are used. There are seven on-site emergency power generators, ranging in size from 7.5-kW to 750-kW, which is the data center generator. There is also a 2,000-kW curtailment generator. The seven machines were installed to provide back-up power to building critical loads during power outages. The 750-kW data center system provides backup power to a data center UPS (Uninterruptible Power Supply), and is also used to shed about 450-kW of load during electrical curtailments. In the early 1990’s interruptible rates were made available to large commercial and industrial customers in this area, primarily those being served at 12-kV or 69-kV voltages. With the interruptible rates, customers pay reduced demand and energy charges, in exchange for an agreement to curtail power use to a target level when requested by the utility during special periods of high demand or economic hardship. The new interruptible rate contract used about 35% (1,500-kW) of the previous summer’s peak demand for firm capacity. Tables 6-5a and 6-5b show how the firm capacity was related to the estimated savings. The requirements of the rate included a two hour curtailment notice with up to 15 interruptions per year and a 10 hour maximum duration per interruption. There would be five or less interruptions per

194


Table 6-5a. Interruptible Rate Savings ———————————————————————————————— Year Usage (kWh) Cost/kWh Savings ———————————————————————————————— 2001 (Jan.-June) 11,659,000 $ .062 2001 (July-Dec.) 8,697,000 $ .046 $ 139,200 2002 20,439,000 $ .048 $ 286,200 2003 (Jan.-June) 9,336,000 $ .049 $ 121,400 ————————————————————————————————

Table 6-5b. Interruptible Rate Comparison ———————————————————————————————— Rate Firm kW Savings $kWh ———————————————————————————————— Firm 100% 0% $.062 Interrupt 3,000 $.060 2,500 6.6% $.058 2,000 14.7% $.053 1,500 19.6% $.049 1,000 26.2% $.045 0 36.0% $.044 ———————————————————————————————— month with a buy through allowed for economic interruptions and penalties for not meeting the firm level. Manual curtailment procedures were developed to provide equipment shutdown to comply with the agreement. These included turning off discretionary lighting, non-centralized air conditioners, pilot line machinery and some test equipment. Major electrical loads like chillers, air handlers, and pumps were also affected by the procedures, sometimes causing short-term discomfort in occupied buildings.

POWER MONITORING The initial real time power measurement system was a six point personal computer-based system. This replaced watt-hour metering at key substations. The system was expanded to 17 measurement points,


195

including chiller plant and data center electrical power loads. The PowerLogic system allows monitoring in real time of critical loads. This information is needed for managing interruptions and to ensure meeting the firm power target. With the interruptible contract in place, an on-site generator with enough power could continue the operation of key equipment, including the central chiller plant. Maintaining lower electricity costs was achievable with this generation. In 2001, a 2-MW generation system was purchased, complete with step-up voltage substation and paralleling switchgear. The system successfully operated during the initial utility mandatory curtailment called in July, 2001. The generator has a 2,000-kW standby, 1,850-kW prime rating with a 3,000-HP Detroit Diesel, 16 cylinder engine which is fuel injected and turbo charged. A Marathon, 4-pole alternator provides 480 volt output and Kohler PD-100 integrated controller provides LCD operator interface, automatic synchronism and paralleling. A water cooled, 150 gallon glycol loop is used with a 100-HP direct in line fan. The fuel consumption is 133 gallon/hour at 100% load. The Johnson Controls system has grown to incorporate controls for approximately 8,000 hardware and software points, controlled by 700 individually programmed processes. These processes control all HVAC functions for fourteen buildings, the central chiller plant and the boiler operation. In the boiler room, the system monitors all critical mechanical systems, burner management systems and steam parameters. It is Ethernet based and accessible from eight operator workstations. Curtailment is a four-step process that is accessed from any operator workstation. The process initially sub cools air-conditioned spaces by turning off all reheat systems. Next, dewpoint setpoints are adjusted upward so that discharge air temperatures go up, allowing chilled water temperatures to rise. The next step starts to override all chillers to an off state, except one 1,000-ton machine which is manually controlled. Building chilled water pumps are slowed (VFD operated) or shutdown. This reduces the chiller plant load from about 1,600-kW to 700-kW, before any electrical generation is placed on-line. After these steps, on-site generation is started about an hour before the power target must be reached. Depending upon the actual power load before the curtailment, only two generators are normally needed to reach the target power level. First, the data center generation takes place,

196


contributing about 450-kW. About 30 minutes before deadline, the 2MW generator is placed on-line. The power measurement system data are used for any adjustments needed to meet the target. Additional building lighting, air handlers and smaller water pumps may be shutdown. Prior to July, 2001, electricity costs averaged $.061/kWh. After the first curtailment with generation, the firm demand level was reset to 1,000-kW, under terms of the interruptible contract. Costs averaged $.046/kWh for the rest of the year. The estimated savings for the first 2 years after installation exceeded $500,000. The $700,000 investment yielded a 37% ROI, less than a 3-year simple payback. The system reduced overall costs per kWh by 25%, compared to non-interruptible rates. Another DG facility in Mobile, Alabama, consists of 41-MWs of electrical generation (10 Wartsila gensets). The facility operates in an island mode as the sole source of electricity for a hydrocarbon processing plant. This facility is designed to start large compressors at the hydrocarbon processing plant. These compressors use 6,000 HP electric motors. The DG facility also provides heat to the plant using engine exhaust gas to heat incoming oil before it is processed.

REDUCING CO AND VOC Another project at the Sweetheart Cup factory in Owings Mills, Maryland, included oxidation catalysts for CO and VOC control. The plant equipment used Wartsila 5.76-MW gensets and 25,000 lb/Hr fired steam recovery burners. NOx was controlled to 6 PPM and CO to 80 PPM. One problem at this facility was the fitting of the new DG plant onto a limited area. The limited space required a narrow four elevation building to fit the required equipment onto the available plant space. As a part of Northwest Airlines expansion at Detroit International Airport, Metro Energy LLC installed a distributed generation facility. Equipment included Wartsila Model 18V34SG gensets rated 5.76-MWs at full load. The systems used oxidation catalysts for CO and VOC control and closed loop heat rejection systems to provide engine jacket water and air cooling. The airport expansion included a new control tower that is located very close to the distributed generation facility. Water vapor from com-


197

bustion or cooling towers was a concern since there would be some potential for a visible plume. The low heat rate system is based on the use of closed loop cooling systems instead of cooling towers. The Chicago area has seen several distributed generation projects because of the rate structure of the local utility. The University of Illinois-Chicago provides part of their electrical power through distributed generation. One University of Illinois facility in Chicago not only produces electricity and some thermal product, but it uses a fired incinerator to reduce CO and VOCs from the genset gas exhaust. This allows the facility to meet environmental requirements without additional back-end cleanup systems. There are two 4.1-MW gensets provided by Wartsila. Factors to be considered for distributed generation include electric costs and rate structure, natural gas rates, maintenance requirements and costs, electric deregulation impact and potential changes in operation.

COOLING, HEAT AND POWER SYSTEMS Distributed energy generation systems that combine cooling, heat and power (CHP) or cogeneration can make significant contributions to mitigating power constraints. These systems can meet increased energy needs, reduce transmission congestion, cut emissions, increase power quality and reliability and increase the overall energy security for a facility. The U.S. Department of Energy’s market assessment estimates that CHP could be successfully applied in almost 10% of large federal facilities (Table 6-6). This would annually conserve 50 trillion Btus of energy, reduce CO2 emissions by almost 3 million metric tons, and reduce utility bills by $170 million. A combined heat and power system generates electricity (or shaft power) and uses the heat from that process to produce steam, hot water or hot air. The most common application involves a prime mover (gas turbine or engine) with a generator to produce electricity and capture the waste heat for process steam and space heating. Boiler steam may be passed through a turbine to generate electricity in addition to serving other thermal applications. One of the simplest systems involves replacing steam pressure-relief valves with a low-cost backpressure steam

198


Table 6-6. CHP Potential at Federal Buildings ———————————————————————————————— Hospital

Industrial

Office

———————————————————————————————— Total Mfeet2

141

115

514

Mfeet2 buildings with CHP payback

Emergency and Backup Power Sources: Preparing for Blackouts and Brownouts

UNIX Backup and Recovery

Unix Backup and Recovery

Power Sources and Supplies: World Class Designs

Backup

Backup

Power Sources and Supplies World Class Designs

Simple Computer Transfer and Backup

Pro Data Backup and Recovery

Pro Data Backup and Recovery

Preparing for Dental Practice

Principles and Practice of Emergency Neurology: Handbook for Emergency Physicians

Principles and Practice of Emergency Neurology Handbook for Emergency Physicians

Principles and Practice of Emergency Neurology: Handbook for Emergency Physicians

Backup & Recovery: Inexpensive Backup Solutions for Open Systems

Backup & Recovery: Inexpensive Backup Solutions for Open Systems

Preparing and Delivering Technical Presentations

SPICE for Power Electronics and Electric Power

SPICE for Power Electronics and Electric Power

Emergency Doctor and Cinderella

Emergency and Security Lighting

SPICE for Power Electronics and Electric Power

Sources for Ancient History (Sources of History)

Particles, sources, and fields

Particles, sources, and fields

Particles, sources, and fields

Particles, sources, and fields

Emergency Doctor and Cinderella

Emergency Doctor and Cinderella

Emergency and Backup Power Sources: Preparing for Blackouts and Brownouts

UNIX Backup and Recovery

Unix Backup and Recovery

Power Sources and Supplies: World Class Designs

Backup

Backup

Power Sources and Supplies World Class Designs

Simple Computer Transfer and Backup

Pro Data Backup and Recovery

Pro Data Backup and Recovery

Preparing for Dental Practice

Principles and Practice of Emergency Neurology: Handbook for Emergency Physicians

Principles and Practice of Emergency Neurology Handbook for Emergency Physicians

Principles and Practice of Emergency Neurology: Handbook for Emergency Physicians

Backup & Recovery: Inexpensive Backup Solutions for Open Systems

Backup & Recovery: Inexpensive Backup Solutions for Open Systems

Preparing and Delivering Technical Presentations

SPICE for Power Electronics and Electric Power

SPICE for Power Electronics and Electric Power

Emergency Doctor and Cinderella

Emergency and Security Lighting

SPICE for Power Electronics and Electric Power

Sources for Ancient History (Sources of History)

Particles, sources, and fields

Particles, sources, and fields

Particles, sources, and fields

Particles, sources, and fields

Emergency Doctor and Cinderella

Emergency Doctor and Cinderella

Recommend Documents