EPRI JOURNAL is published eight times each year (January/February, March, April/May, June, July/August, September, October/November, and December) by the Electric Power Research Institute.

EPRI was founded in 1972 by the nation's electric utilities to develop and manage a technology program for improving electric power production, distribution, and utilization.

EPRI JOURNAL Staff and Contributors Brent Barker, Editor in Chief David Dietrich, Managing Editor Ralph Whitaker, Feature Editor Taylor Moore, Senior Feature Writer David Boutacoff, Feature Writer Eugene Robinson, Technical Editor Mary Ann Garneau, Production Editor Jean Smith, Staff Assistant

Richard G. Claeys, Director Corporate Communications Division

Graphics Consultant: Frank A. Rodriquez

© 1989 by Electric Power Research Institute, Inc. Permission to reprint is granted by EPRI, provided credit to the EPRI JOURNAL is given. Information on bulk reprints available on request.

Electric Power Research Institute, EPRI, and EPRI JOURNAL are registered service marks or trade­ marks of Electric Power Research Institute, Inc.

Address correspondence to: Editor in Chief EPRI JOURNAL Electric Power Research Institute P.O. Box 10412 Palo Alto, California 94303 Please include the code number on your mailing label with correspondence concerning subscriptions.

Cover: Lighting is often a key consideration for effective merchandising in the commercial sector. Unfortunately, many businesses do not see energy efficiency as an important concern in developing their lighting strategies. EDITORIAL Lighting for Efficiencyand Productivity

In the current economic environment, efficiency is the order of the day. More and more U.S. businesses are feeling the pressure from foreign and domestic competitors to cut costs and increase productivity-to seize a competitive edge by doing more with less. Cost-effective improvements in energy efficiency and workplace productivity can make the crucial difference between a profitable commercial concern and a marginal one. Lighting is one area of energy use that offers a largely untapped potential for significant gains in efficiency. Commercial buildings present a broad target for lighting improvements: lighting accounts for about a third of the electricity consumed in the commercial sector and can account for more than 40% of a building's peak demand requirements. Furthermore, lighting is often a major input to the cooling load, an­ other important factor contributing to a building's energy use in summer peak demand periods. Advanced lighting approaches can significantly boost efficiency in commercial settings-and, as this month's cover story notes, they can also enhance the productivity of a building's occupants. In fact, the productivity issue can override all others in the implementation of changes to the workplace environ­ ment, since labor costs exceed lighting costs by a ratio of more than 100 to 1 in a typical commercial enterprise. Although new lighting technologies clearly can provide energy savings for customers, changes in productivity must be valued as well. The technology is here-there are many well-proven lighting measures and equipment upgrades that can be cost-effectively implemented today. These include more efficient lamps, ballasts, controls, and fixtures and improved ways of configuring and installing this equipment. The electric utility industry is working hard to get customers to use new lighting strategies to increase their operating effi­ ciency. A majority of U.S. utilities have programs in place that educate customers on how to use lighting more effectively and that provide incentives for the implementation of cost-effective lighting measures and equipment. EPRI is supporting these efforts by developing customer-oriented information that characterizes state-of-the-art lighting equipment and practices. The Institute is also demonstrating best-available technolo­ gies, developing lighting design tools, and performing evaluations of how lighting systems interact with other workplace support systems. Through these programs, we are striving to help utilities and their customers get the most value out of lighting energy use, and the best news is that there is value to be had on both sides of the meter. Utilities can select the lighting technologies that, if implemented by their customers, can best satisfy the utilities' load-shaping objectives, whether they be on-peak reductions, strategic load growth, off-peak sales, or retention of baseload. At the same time, customers will enjoy the benefits of lower energy costs and an en­ hanced work environment.

Morton Blatt, Manager Commercial Program Customer Systems Division RESEARCH UPDATE

42 TOXRISK Update A new version of the EPRl-developed TOXRISK computer program is capable of estimating toxic risk as a function of period of exposure as well as level of exposure.

44 Direct-Load-Control Programs The duty-cycle approach for evaluating and predicting the impact of residential direct-load-control programs is being applied at three eastern utilities. 24 Home Automation 32 NOE Center

47 Planning and Forecasting Under Uncertainty A scoping study has identified three major approaches that utility planners can use to address uncertainty in marketing, forecasting, and demand-side planning.

49 Coal Quality Impact Model Detailed equipment performance models allow the flexible CQIM'" software to quantify the impacts of various coals on all power plant systems directly affected by coal quality.

52 Coal Ash Use Sites Field investigations are measuring long-term changes in soil, vegetation, groundwater, and surface water around such coal ash use sites as structural fills and roadbeds.

56 Gas Turbine Combustion Viewing System DEPARTMENTS 60 Calendar By allowing direct, high-resolution viewing of the hot-gas 40 Tech Transfer News path, the combustion viewing probe can give early warning 58 New Contracts 61 Authors and Articles of system malfunctions in today's advanced gas turbines. 59 New Technical Reports EPRIJOURNAL

Volume 14, Number 8 December 1989

EDITORIAL Lighting for Efficiency and Productivity

COVER STORY 4 Lighting the Commercial World Advanced lighting strategies can greatly increase energy efficiency in the commercial sector, but utilities are having difficulty getting businesses interested.

FEATURES 16 Electrification: Key to Manufacturing Productivity The history of American manufacturing shows a strong linkage over the past century between productivity gains and the implemen­ tation of electrified production techniques.

4 Lighting 24 Integrated Home Automation 16 Manufacturing Electronic intelligence that can coordinate the operation of various home systems and appliances promises greater control, comfort, and convenience for homeowners.

32 The NDE Center: Celebrating a Decade of Achievement EPRl's Nondestructive Evaluation Center, where research results become fieldworthy tools, has established an impressive record of accomplishments in its first decade.

or owners of the Peppermill Hotel Casino in Reno, Nevada, lights are serious business. They help maintain around- the-clockF allure, keeping patrons com­ fortable and alert. As one manager puts it, "Lighting is key to the overall ambi­ With about a quarter of the electricity sold in ence we're trying to create for the cus­ tomer." Understandably, then, a sugges­ this country used for illumination, industry tion that the business install energy­ efficient lights is no small proposal. But experts believe that advances in lighting sys­ the Peppermill this spring did just that, replacing about 1000 incandescent bulbs tems may have the largest potential for energy with energy-efficient fluorescents that have cut the lighting bill in half. Even savings of any electricity end use. To help more important, the switch did not tar­ nish the dazzling atmosphere. realize these savings, utilities across the nation The retrofit project, done in collabora­ tion with Sierra PacificPower, was the have established programs encouraging their result of a program designed to encour­ age energy efficiency in lighting. Utilities customers to adopt more efficient lighting. Tar­ across the country today are offering similar programs. Many, like Sierra Pa­ geting both new and retrofit markets, with cific, attract customers with rebates for energy efficiency. Other utilities offer au­ emphasis on the biggest area of potential dits, loans, design assistance, and a vari­ ety of educational tools, including dem­ savings-the commercial sector-the pro­ onstrations of energy-efficient lighting options. But while these programs have grams offerincentives like rebates and loans for helped encourage more efficient use of lighting, they have captured only a small energy-efficient conversions. But stubborn portion of what some experts say is the largest potential for savings in any en­ barriers to customer acceptance persist, such as ergy end use. And increasingly today, utilities are being called on to step up high first-costexpenses and confusionover the their efforts. Some conservationists go so far as to say utilities have both the re­ array of energy-efficient options available. A sponsibility and the clout to lead the ef­ fort to realize these savings. variety of energy conservation and lighting "The invention of the light bulb gave birth to the electric utility industry, and groups across the country, including utilities, utilities often have taken the lead in pro­ moting high-quality, energy-efficient are working hard to overcome these barriers. lighting," says Karl Johnson, EPRI project manager in Commercial Building Sys­ But just how much of an impact their efforts tems. "Certainly they have a good infra­ structure to reach the consumer, and will have on the average utility customer they've already taken advantage of it by offering incentives like rebates, audits, remains to be seen. and loans. There's more that can be done. But a major challenge is to overcome

EPRI JOURNAL December 1989 5 the significant barriers to customer, de­ "the biggest gold mine in the economy," conservation (a form of demand-side signer, and builder acceptance of existing saying efficiency improvements could planning) is becoming increasingly more high-efficiency lighting products." save U.S. consumers billions of dollars attractive than the traditional supply­ According to EPRl's statistics, lighting a year because of the reduced cost of side planning. Some utilities specifically accounts for 20-25% of the electricity maintaining and operating more efficient gear their incentives toward reducing used in the United States. In the com­ systems. In other words, he says, "this is loads during hours of peak demand. mercial sector, at least a third of the elec­ not a free lunch. This is a lunch you are A 1987 EPRI survey showed that 60% tricity used is attributable to lighting, paid to eat." of utilities surveyed used lighting for a and these figures would be even higher Utility involvement in promoting en­ wide range of demand-side management if the ventilating and air conditioning ergy-efficient lighting over the years has needs and 30% had lighting-efficiency loads attributable to heat emitted by been spurred by concerns about capacity, programs in place. A separate study lights were taken into account. Various the potential environmental impact of from the same year showed that the bulk interest groups offer different estimates new power plants, and a desire to help (92%) of the energy-efficient utility light­ for the savings potential in lighting. But customers cut costs. During the energy ing programs are directed toward com­ all agree there is tremendous oppor­ crisis of the 1970s, utilities in California mercial and industrial customers. tunity for improvement, particularly in and the Northwest pioneered the move­ Schemes ranged from incandescent­ the commercial sector. "This has been a ment to promote all kinds of energy effi­ fluorescent conversions to installing oc­ much-heated topic involving tremen­ ciency. Now, the greatest momentum can cupancy sensors and daylight sensors. dous controversy and many exaggera­ be found in New England, where util­ Financial assistance to the customer, in­ tions of savings potential," says Clark ities are working hard to keep electricity cluding rebates for installing high-effi­ Gellings, director of EPRI's Customer demand during peak hours at a level ciency equipment and interest-free loans, Systems Division. "The vastly different comfortably below capacity. There envi­ was the most popular method of imple­ estimates have made it difficult for re­ ronmental groups and utilities have mentation (offered by 64% of the util­ searchers to know where to focus their found a common cause in conservation ities). Energy audits (offered by 52%), attention." and are collaborating on strategic plans most of which are free, was the second In an effort to pool the input of the dif­ to control electricity demand. Mean­ most common method, followed by ferent groups, EPRI distributed a ques­ while, utilities without immediate capac­ studies evaluating new technologies tionnaire on the maximum potential sav­ ity crises are relying on such programs (conducted by 12%). Another 12% of ings in efficient technologies and fol­ to help offset anticipated future increases surveyed utilities provided education on lowed up with a workshop in April 1989. in demand. Wherever they are imple­ effective and efficient lighting. Gellings, who organized the workshop, mented, strategies in lighting effi­ The number of utilities with such pro­ believes this project represents the first ciency-which range from installing grams has increased in the years since time anyone has rounded up the calcula­ more efficient bulbs to using daylight to the survey, mainly because of the new tions of such a diverse group of conser­ supplement electric light-are an effec­ thrust in New England. And the Na­ vation experts. The resulting report cites tive way to help postpone building the tional Association of Regulatory Utility a savings potential from adopting effi­ next generation of power plants. Commissioners wants to encourage cient commercial lighting technologies of more utilities to follow suit. NARUC's 39-55% by the year 2000. The commer­ Utility concerns conservation committee is working on cial lighting savings accounts for about Even though lighting has been viewed as making energy-efficiency programs more 65% of the potential lighting savings in a captive market for electric utilities, in­ attractive for utilities. "Investor-owned all sectors, including industrial and resi­ novation in making this end use more ef­ electric utility companies, which supply dential. The report incorporates the ficient is on the rise. EPRI's current com­ 77% of this country's electricity, typically views of a wide range of groups, includ­ mercial lighting research plan reflects lose money when they help their cus­ ing the Natural Resources Defense this increased interest. The plan for 1989 tomers use electricity more efficiently," Council, Lawrence Berkeley Laboratory, and 1990 focuses new projects on peak says Stephen Wiel, chairman of NARUC's the Rocky Mountain Institute, public clipping and other strategic conservation conservation committee. "State regula­ utility commissions, EPRl, and others. objectives, with the commercial sector as tory commissions currently set rates us­ Amory Lovins, director of research at a prime target. Commercial lighting is ing formulas that reward utility com­ the Rocky Mountain Institute in Colo­ one of the largest single contributors to panies for increasing sales. In an effort to rado and an outspoken conservationist, peak demand at most utilities. For many, help utilities profit from conservation calls the untapped potential in lighting controlling demand through strategic programs, NARUC now is encouraging

6 EPRI JOURNAL December 1989 Why Not Upgrade?

The simple availability of energy-saving lighting equipment has not been enough to spur its wide usage. For the commercial sector, the decision to implement advanced lighting systems can be seen as a complex puzzle of perceived barriers, incentives (many offered by individual utilities), and technical questions. Simplifying such decisions through continued research and innovative marketing could significantly increase customer involvement.

• • • Bewildering Disruption Advantages array of not of products business understood • • • No Low interest Changes in perceived in the work need conservation environment

• Lower Loans for energy equipment consumption purchase • • Free Equipment energy giveaways audits

• • Savings New Lighting potential energy effects standards on HVAC • • • Human Definition Best response of overall to lighting quality lighting designs commissions throughout the country to experiment with ratemaking reforms. Getting the Word Out Among the suggestions is that utilities Utilities often offer their customers special rates, loans, or other incentives to upgrade be allowed to earn a fair return on in­ lighting systems as part of their energy conservation programs. Brochures educatEJ cus­ vestment in energy conservation pro­ tomers about the incentives and hardware options available, and some utilities even give grams, just as they make money by in­ on-site design assistance to ensure that customers make the most of advanced lighting vesting in new plants. But more changes opportunities. are needed, according to Wiel. Other in­ centives, such as bonuses for achieving conservation goals, are necessary to overcome the current disincentives, he says. Only 10 states, including Wisconsin, currently allow public utilities to make a return on investment in conservation programs. An added incentive for Wis­ consin Electric Power is a bonus for re­ ducing peak megawatts. Wisconsin Elec­ tric's programs as a result have saved more than 156 MW-about 50% of which is from lighting-since the program started two years ago. That's enough to power about 78,000 homes. The program offers rebates based on the cost of the conversion, the kilowatts shaved from the utility's system peak, and the kilo­ watthours saved. "We're willing to kick in on just about anything reasonable the customer presents to us that'll improve efficiency," says Dan Thomas, supervisor of commercial and industrial energy management for Wisconsin Electric. One project involved installing about 175,000 electronic ballasts in the fluorescent lighting systems throughout Milwau­ kee's public schools, about 150 buildings in all. The $5 million project resulted in $2 million worth of rebates. The re­ sulting energy savings will offset the en­ tire cost of the job in five years. For Consolidated Edison of New York, lighting and other conservation pro­ grams are attractive as long as the cost of saving a kilowatt remains lower than the cost of building a new kilowatt of capac­ ity, says Steve Pertusiello, a manager in the conservation department. Consoli­ dated Edison, which supplies power to a 660-square-mile area, including all of New York City, has offered incentives for lighting efficiency to commercial and

8 EPRI JOURNAL December 1989 industrial customers for the past two up to 6% of the New England Electric in­ services a substantial number of tenant­ years. Those with peak demands be­ vestment in the customer's project; and occupied commercial buildings, find tween 150 and 500 kW can have their design assistance from a team of consul­ there is little incentive for either the ten­ systems audited free. tants. The utility also will work closely ant or the landlord to save energy. The A separate, pilot program encourages with customers, using a computer soft­ landlord simply passes the energy bill on customers to devise their own plan to re­ ware program to determine the best en­ to the tenant, and the rent typically duce their lighting load and other energy ergy-efficient options for a given build­ would not reflect any savings resulting loads that affect the utility's system peak. ing. from the tenant's efforts. Most commer­ If the plan meets utility requirements, One "Design 2000" lighting project in cial buildings, in fact, are not owner­ the customer will get a $200 rebate for the works involves a 200,000-square-foot occupied. every kilowatt reduced from peak de­ computer company office building in Peter von Herrmann, president of the mand, with a maximum offer equivalent Marlboro, Massachusetts. Timing con­ Lighting Research Institute, which fo­ to half the cost of the installation. Re­ trols and occupancy sensors that turn cuses on the human response to lighting, bates are offered also for implementing lights on and off automatically will be says productivity is another important energy-efficient technologies such as incorporated in the new structure. En­ concern. For example, von Herrmann lamps, ballasts, occupancy sensors, and ergy-efficient fluorescent lamps and bal­ says, the employee cost of a square foot daylighting controls. lasts also will help save energy. The util­ of office space (assuming a salary of Pertusiello credits encouragement from ity will spend up to $27,695 to cover the $40,000 and a 100-square-foot office) the state commission. "They've been entire cost of the efficiency improve­ equals $400 a year. But the cost of light­ mandating programs and telling us which ments. ing that square foot is only about 50\t a measures we should consider to offer for Since the start of its systemwide pro­ year. If productivity is adversely affected incentives," he says. But Consolidated Ed­ grams two years ago, New England Elec­ even slightly by efficiency improvements ison has another motivation as well. With tric has spent nearly $60 million on plan­ in a lighting system, those changes will supply options limited because of strong ning, implementation, and evaluation of lose much more money than they ever opposition to new plants, the only alter­ conservation programs, about a third of could save in electricity. "We are saying native is to control demand or to buy which has gone toward lighting, reports that the productivity gain is a very im­ electricity from the independent power Peter Gibson, manager of conservation portant one. Or, at least, not losing pro­ producers. Lighting conservation pro­ and load management services. The re­ ductivity," von Herrmann says. "Other­ grams have also helped Consolidated Ed­ sulting savings total about 110 MW. Gib­ wise, you're kidding yourself in the ison defer the expansion and upgrading son says that conservation programs energy equation." of its distribution systems. typically are 30-50% less expensive than Richard Tempchin, a manager in the Somes utility programs go beyond ret­ building new power supplies. customer services and marketing divi­ rofitting existing lighting systems to tar­ sion of Edison Electric Institute, says one get the new construction market as well. Barriers for customers key problem for utilities trying to "sell" For example, New England Electric Sys­ Convincing businesses that energy-effi­ the idea of energy-efficient lighting is tem's "Design 2000" program targets cient lighting can maintain and even en­ that customers often are overwhelmed new commercial and industrial build­ hance their environment is a challenging and confused by the number of options. ings, which are expected to account for task. Even customers with significant "What's happening is the technologies much of the utility's growth in peak de­ lighting bills are hesitant to alter lighting are getting way ahead of customers," he mand in the next decade. Incorporating systems that are functioning properly. says. "There are all kinds of new op­ energy-efficient strategies during the ini­ Glenn McCormick, engineering manager tions-energy-efficient ballasts, elec­ tial design of a building is almost 10 for the Peppermill in Reno, says most ca­ tronic ballasts, dimmers, energy manage­ times cheaper than retrofitting, accord­ sinos are more concerned about their ap­ ment controls. It's not as easy as it used ing to New England Electric's calcula­ pearance. "If your casino attracts one to be to make these decisions in light­ tions. The program, which is among the person who comes in regularly, that far ing." As Tempchin points out, commer­ most sophisticated in the country, was outweighs any energy savings you're go­ cial customers, who get a wide selection developed in conjunction with the Con­ ing to get out of changing a few light of equipment from lighting suppliers, servation Law Foundation of Boston. It bulbs," he says. are more informed than residential cus­ offers financial assistance, up to the full For more typical commercial cus­ tomers. Some of the more advanced cost of upgrading to a more energy­ tomers, the barriers are just as strong. lighting technologies simply are not efficient system; a design incentive of Utilities like Consolidated Edison, which available in hardware or grocery stores.

EPRI JOURNAL December 1989 9 Strategies for the Old and the New

Energy-efficient lighting works well both with existing buildings and in new construction projects. Virtually all facilities can be retrofit with advanced hardware that will increase efficiency and improve lighting quality. Designers can be more comprehensive with new buildings, taking full advantage of daylighting and computer-based control strategies. An additional obstacle is the first-cost likely see no need to change it. But even One feature of the new lab is a demon­ expense of high-efficiency lamps. If a if customers want to change their light­ stration area with walls, ceilings, and lighting bill is a small expense for the ing, they often do not have the time to furniture that can be altered to mimic an consumer, it is hard for that person to explore the multitude of efficiency op­ actual setting. Strategies for daylight­ justify buying an energy-efficient fluo­ tions, Johnson says. ing-the use of natural light tu supple­ rescent lamp when the standard incan­ In his prior management experience ment interior electric lighting-also can descents are available at about a tenth of with Stanford University, Johnson con­ be tested in full-sized commercial envi­ the cost. Some utilities have tried to fronted many of these customer barriers. ronments. A resource library will pro­ overcome this barrier by actually giving Nevertheless, he convinced the univer­ vide the latest information on efficient lamps away. About four years ago, Texas sity to install more efficient lighting sys­ lighting technologies. In a preliminary Utilities tested the residential waters by tems campuswide. He says it is impor­ report, Seattle City Light identified the sending out coupons for free efficient tant to be sensitive to the customer's commercial sector as the fastest-growing fluorescent bulbs that could be screwed concerns when proposing any project. electricity consumer in the Northwest. In into incandescent sockets. About 25% of "We approached this and other projects addition, the report says, lighting ac­ the 16,000 residents in the targeted town by proposing to improve the quality of counts for more than half of commercial took advantage of the offer. the lighting system while saving money electricity, when heating, ventilating, and John Nordgren, a supervisor in light­ and energy. If I'd walked in and said air conditioning loads are figured in. The ing for Puget Sound Power & Light, says Tm going to de-lamp your fixtures,' project cost $1.9 million, 70% of which Puget's programs have been successful they would have hit the ceiling." Part of was funded by the Bonneville Power Ad­ specifically because they help cover the the Stanford project involved replacing ministration. seemingly high first-cost expenses. about 500 incandescent lamps in the law Meanwhile, the New York State En­ Lighting has always been a significant school with screw-in fluorescents and re­ ergy Research and Development Author­ part of Puget's programs, typically ac­ flectors. This saved about 75% of the ity is working on plans to establish a counting for more than 40% of annual lighting energy used in the original sys­ similar center in New York. Many light­ program expenses. Among the most tem and improved the environment. ing manufacturers and some utilities successful of Puget's programs is one provide similar demonstrations. But awarding grants that typically cover 50% Tec hnology transfer planners involved in the Seattle and or more of a customer's energy-efficient Getting information on lighting to those New York centers say their demonstra­ conversions. The utility also conducts who are not necessarily interested in effi­ tions are more comprehensive and do audits for the customer, estimates the ciency is even more important than de­ not limit technology to one manufac­ savings that would result from an en­ veloping new technology, some experts turer's products. David Goldstein, senior ergy-efficient system, and helps in select­ say. "There are so many high-efficiency staff scientist at the Natural Resources ing a contractor for the job. But no mat­ products already available that cus­ Defense Council, believes lighting labs ter how attractive such programs may tomers aren't using now," Johnson says, like Seattle's should be in every major sound, they remain a challenge to sell, offering the electronic ballast as one ex­ city in the country if the United States is Nordgren says. "Some customers have ample. "It is important to provide infor­ to begin to take advantage of the savings been motivated from a motherhood and mation on existing technologies while potential in lighting. But others question apple pie perspective, the view that sav­ developing the even more advanced the effectiveness of such centers. ing energy is a good thing to do," he technologies." "You often have the problem of bring­ says. "But most are inspired by great Among technology transfer efforts ing the customer to the mountain versus savings." In addition to Puget's incen­ currently under way is Seattle City bringing the mountain to the customer," tives, customers look for substantial dol­ Light's design laboratory. The 8500- Johnson says, suggesting that such cen­ lar savings on their electricity bills. "Dol­ square-foot facility, which opens offi­ ters could be more effective if they incor­ lars," Nordgren says, "are the biggest cially in December 1989, is geared to­ porated mobile exhibits. Johnson's con­ motivators." ward engineers, architects, lighting de­ cern is that most lighting systems are not Karl Johnson of EPRI agrees. "If light­ signers, contractors, manufacturers, designed by professional lighting design­ ing's less than 1% of customers' overall equipment suppliers, and their clients. ers-the people likely to make most use costs, that's the kind of attention you can The design laboratory will allow visitors of demonstration centers like Seattle's. expect them to give it." Also, he says, as to test or demonstrate energy-efficient Meanwhile, those less skilled in lighting long as an existing lighting system func­ lighting strategies in mock commercial design still have not been reached. An tions adequately, the customer will most settings. EPRI study has determined the need for

EPRI JOURNAL December 1989 11 a mobile unit and created specifications for a unit that could demonstrate a vari­ ety of lighting systems. The module, as specified, would be freestanding, so util­ ity personnel could assemble it to fit a variety of commercial and other settings. Other efforts in technology transfer are under way elsewhere. Currently in the works at the year-old Lighting Re­ search Center at Rensselaer Polytechnic Institute in New York are plans for a re­ search, education, and technology trans­ fer program. LRC, which evolved from a need identified at an EPRI workshop five he options available for improving depending on use and energy rates. years ago, is striving to become a na­ Tlighting efficiency today can be Advancements in phosphors-the tional center for lighting research and overwhelming at times, not only for the white coatings inside fluorescent lamps design. It stresses an interdisciplinary consumer but also for the utility. The that emit visible light when exposed to approach, pooling the knowledge of spe­ potential for improvement does not ultraviolet radiation-have made the cialists in fieldsranging from psychology stop with more efficient lamps. It is color quality of fluorescents virtually to computers. LRC has proposed a proj­ found also in ballasts, in fixtures, in indistinguishable from that of incan­ ect that will investigate impediments to lighting-control strategies that may in­ descents. Compact fluorescents are technology transfer in the marketplace corporate the use of natural daylight, made both with and without ballasts and assess how well utility programs are and in the complex interactions be­ attached. working, says Associate Director Russell tween the lighting and heating, cooling, Incandescent lamps Significant im­ Leslie. Officials there still are working to and ventilation systems. provements have been made in incan­ secure funding for the project. To take advantage of the full poten­ descent lamps, including the develop­ The program would include a sympo­ tial for savings while maintaining vi­ ment of elliptical reflector (ER) lamps sium to address the technology transfer sual quality, experts suggest that util­ for recessed fixtures and improve­ issue. The aim is to bring together an as­ ities adopt programs that integrate a ments in the reflective films on the sortment of people involved in decisions number of efficient technologies. Here walls of other incandescents. The ellip­ on lighting, including designers, manu­ is a sampling of some relatively new, tical shape of ER lamps, combined with facturers, utility representatives, engi­ highly efficient lighting technologies. It an internal coating of reflective alumi­ neers, and vision scientists. "All these is by no means a comprehensive list but num, directs light downward rather different groups really have trouble talk­ highlights effective technologies that than into surrounding areas. Other in­ ing to one another," Leslie says. "Yet have not yet been adopted widely. candescents now are made with films each group is putting forth information Compact fluorescent lamps Twin­ on the interior lamp walls that reflect that is vital to the functioning of the tube and quad-tube fluorescents, two infrared light onto the filament while other groups." of the several types of compact fluores­ allowing visible light to pass through. The spearhead of the LRC technology cent, are often used to replace incan­ The formerly wasted infrared energy transfer program would be Lighting Now, descent lamps. Some are made to be then heats the filament further, produc­ a serial publication to be developed to screwed into existing incandescent fix­ ing more visible light. The bulb shapes reach people from the whole range of re­ tures. They are available in wattages have been changed to both spherical lated fields. The proposed format is a ranging from 5 to 26 and offer wattage and cylindrical to help the films reflect four-page, biweekly bulletin that could savings of 60-75% when used to re­ the infrared light properly. A 60-watt be kept in a binder for reference. LRC place incandescents. They last about 10 coated incandescent lamp provides plans to develop an international com­ times longer than incandescents but about as much light as a conventional munication network to keep abreast of can cost up to 20 times as much. The 150-watt incandescent PAR lamp. ER the latest developments in lighting payback period can be less than a year, lamps are widely available, but as of around the world. Currently, Leslie says, information on such developments may not appear in professional journals until

12 EPRI JOURNAL December 1989 Lighting Te chnology

now only one manufacturer has intro­ cents. They are just beginning to because they make it practical to re­ duced incandescents with selective replace incandescents in the commer­ move half the lamps and ballasts from a reflecting films. cial sector for uses like display lighting. fixture; the decrease in light output T- 8 fluorescentlamps The 4-foot, 32- Electronic ballasts By far the most from that fixture is only 25-30%. The watt T-8 lamp has a double phosphor significant development in ballast tech­ aluminum reflectors offer the lowest coating, which offers higher efficiency, nology, the electronic (solid-state) bal­ initial cost; the silver and dielectric film longer-lasting brightness, and good last runs fluorescent and HID lamps at types have the highest reflectivity. The color quality. With a diameter of 1 inch, high frequency. The result: essentially multiple dielectric coating also can be it is half an inch thinner than the T-12 no flicker or noise and lower ballast used on HID fixtures and has been lamp, which it typically replaces. The power losses. With fluorescents, in­ shown to increase the efficiency of such T-8 lamp will fit into the standard 4-foot creased efficiency is an added advan­ fixtures by up to 30%. fixture but will not work with a T-12 tage. For example, a standard T-12 Occupancy sensors and photocell ballast. With an electronic ballast, the system with an electronic ballast is up controls These are just two examples of lamp operates at a higher frequency, to 25% more efficient than a T-12 system the types of technology being used to making it more efficient. When the with a standard, magnetic ballast. With control lighting in commercial build­ lamps are compared without ballasts, HID lamps, electronic ballasts allow ings. Occupancy sensors use infrared the T-8 proves to be about 12.5% more more control over light output during or ultrasonic signals to detect motion in efficient than the standard T-12 lamp. the lamp's life. Some electronic ballasts a room, automatically turning lights on Low-wattage metal halide lamps offer continuous, or step, dimming when motion is detected or off when Metal halide lamps, a type of high­ capability, allowing light to be adjusted none is detected. Photocell systems intensity-discharge (HID) lamp, have to the requirements of a particular detect the amount of daylight entering traditionally been available only in space. With electronic ballasts, single a given space and automatically adjust wattages between 175 and 1000. Now lamps can be dimmed as easily as the electric lighting system to maintain available in 100 watts or less, they can large groups, making possible more the desired illumination. Computer­ be used to replace incandescent lamps. options for lighting control strategies controlled lighting management sys­ While compact fluorescents typically such as daylighting. tems that incorporate several control replace incandescents of 100 watts or Optical reflectors Made of alumi­ strategies also are being used in the less, HID lamps are used in place of num, silver, or multiple dielectric coat­ commercial sector. Such systems have the higher-wattage incandescents. Like ings, optical reflectors improve the included functions that turn lights on incandescents, metal halide lamps al­ efficiency of fluorescent fixtures by up and off at programmed times, automat­ low for better focus and control than to 20%. The reflectors direct more light ically reduce lighting loads during fluorescents in that they cast light downward so that it does not bounce times of unusually high peak utility directly toward one point, rather than around inside the fixture. Such reflec­ demand, and increase power to a light­ to a general area. HID lamps are 150% tors can cut lighting energy costs for ing system as it dims with age and dirt more efficient than standard incandes- existing overlighted spaces by half accumulation. D

EPRI JOURNAL December 1989 13 Research at EPRI

PRI research has explored many improving lighting and HVAC designs. responded to a questionnaire asking E aspects of energy efficiency in One of the successful strategies them, among other things, to rate the lighting. In the process, it has devel­ tested to reduce peak load entailed con­ lighting available for specific tasks oped a broad portfolio of tools and trolling the temperature in the return and in the surrounding environment. information of value to utilities, includ­ air plenum to keep it a few degrees Participating occupants had been at ing technology assessments, surveys of cooler than is typical, and then allow­ their workstations at least one year. utility lighting incentive programs, ing the temperature to rise a few This study indicated that lighting design and analysis software, retrofit degrees during the building's peak satisfaction is strongly correlated with strategies, and a utility lighting hand­ demand period. Understanding and the distribution of light. Cases with book. Two recent projects have pro­ controlling such interactions can uniform distribution provided more duced particularly interesting results reduce peak demand due to lighting, satisfaction than cases with strongly bearing on the design of energy­ cooling, and fans by about 17%. If these nonuniform patterns, which were pro­ efficient lighting systems. more efficient strategies were applied duced by some systems using task In the first, EPRI and the Department to just 5% of existing U.S. commercial lighting. Some systems with task light­ of Energy (DOE) have been collabo­ spaces and 25% of new construction ing tended to use greater power den­ rating on a project to finddesign strat­ each year, demand reductions would sities; thus the belief that individual egies that help optimize the interaction total about 1400 MW annually. task lighting combined with lower gen­ between lighting systems and heat­ Researchers will continue testing eral (ambient) lighting is always energy ing, ventilating, and air conditioning and validation of the new design meth­ efficient and advantageous was not (HVAC) systems in commercial build­ odologies and computer model and borne out by this analysis. The reason ings. Improving this interaction not will work with the design community may be the extreme contrast between only can help save energy but also can to adopt these improved methods. the task lighting and the darker sur­ reduce peak demand for utilities, par­ In an effort to better understand the rounding vertical surfaces of the work­ ticularly during the summer months, human impact of various lighting sys­ stations studied. Another key finding when the heat from lights creates more tems, EPRI joined with the New York of this study was that there was little or work for air conditioning systems. State Energy Research & Development no correlation between lighting power The project has already involved Authority, DOE, NIST, and the Lighting densities and occupant satisfaction more than 350 tests in a fully instru­ Research Institute in funding addi­ with the lighting system.

mented test chamber at the National tional analysis of an earlier DOE study. Overall, the study provides valuable Institute of Standards and Technology This "second-level," follow-up analysis insights on different types of lighting (NIST). Data have been gathered on thoroughly examined 912 workstations systems and the need to carefully under­ some 400 variables that relate to light­ in 13 U.S. office buildings. Extensive stand the complex impacts of lighting ing levels, air temperatures, room sur­ data were collected on energy use as systems and design strategies on occu­ face temperatures, and other factors. A well as lighting levels and the contrast pant satisfaction. Additional research computer model of the chamber is between light and dark areas through­ topics identifiedby this study are being being developed for future use in out the workstations. Occupants also evaluated for further investigation. o

years after the work has been done. An­ actly what we're trying to catalyze here. outsider, this may sound like a simple other problem is that the information No one is doing that effectively now." solution. But maintaining quality while presented is so highly technical that only encouraging efficiency is no easy task. A Standards and quality a limited audience can understand it. specified maximum energy use does not "You have to get that information to the Another avenue for encouraging efficient guarantee good-quality light. Lighting people-the building managers, the ar­ lighting is to impose standards that man­ quality depends on a number of factors, chitects, the electricians, and all the date or suggest maximum energy levels, including color rendition and glare. others who specify lighting. That's ex- in watts per square foot. To an industry Jim Heldenbrand, manager of stan-

14 EPRI JOURNAL December 1989 dards for the American Society of Heat­ how well utility programs address the is­ low more effective use of daylighting, to ing, Refrigerating, and Air-Conditioning sues of quality in lighting. But ambitious a new, high-frequency surface-wave Engineers (ASHRAE), admits that stan­ programs like New England Electric Sys­ lamp that is 30% more efficient than the dards have limitations. "You cannot set tem's "Design 2000" emphasize provid­ common fluorescent. One recent devel­ standards that only the best design can ing a "high-quality environment through opment could be useful to utilities as a meet," he says. "You have to set stan­ reduced glare, better color rendition, and design tool to optimize indoor and out­ dards that, in essence, eliminate the lighting controls that are more respon­ door lighting. This tool, a software pro­ poorest practices only." It has taken six sive to occupants' needs." Feedback gram called RADIANCE, simulates actual years for ASHRAE and the Illuminating from other utilities shows that customers lighting environments so that quality Engineering Society of North America to now rely on them to help make decisions energy-efficient lighting systems can be develop the so-called 90.lP standards regarding quality. "Customers will call designed from the desktop. The simula­ aimed at new commercial buildings. The us up asking our opinion on some fix­ tions are so precise they resemble photo­ voluntary standards got preliminary ap­ ture, some bulb, or some technology in graphs. A user's options are virtually proval this summer and should be pub­ general," says Dan Thomas of Wisconsin unlimited, including the ability to move lished by the end of the year. The light­ Electric Power. "They're using us as a or change fixtures and furniture. The ing specifications are based on watts per sounding board." program uses a simulation technique square foot and are about twice as strin­ Bob Henderson, senior commercial called backward ray tracing, by which a gent as the previous standards. The De­ program specialist at Carolina Power & researcher can track a light path from its partment of Energy in July imposed its Light, stresses that today's working envi­ destination to its source, incorporating own set of standards, which are manda­ ronment calls for closer attention to qual­ such factors as specular reflection and tory for new federal buildings. Based on ity in lighting. Lighting schemes must be geometry. The program has been verified research that went into ASHRAE's stan­ designed for what he calls "seeability"­ with real-life measurements. LBL has re­ dards, DOE's standards follow a similar the ability to see well. Modern office leased this research-grade program to format, allowing, for the first time, credit equipment has introduced new compli­ universities and hopes to make it more for daylighting and controls. DOE's stan­ cations, such as glare on video display user-friendly in the future. dards also include even more stringent terminal screens. When a break room at But while such technological advances requirements for 1993. Carolina Power was converted into a ste­ may sound exciting, they also raise some But while the standards offer some nographers' office, workers complained serious questions: Will the marketplace guidance, the ultimate decision on fix­ that glare from the light fixtures inter­ ever begin to keep pace with this tech­ tures and lamps still rests with the light­ fered with their attempts to read from nology? What will it take for that to oc­ ing designer. "Everyone in the scientific computer screens. A big part of the cur? Perhaps increased utility activity is community and in the lighting industry problem was that the office was over­ part of the answer. But clearly the suc­ will agree on one statement," Helden­ lighted. Among other changes, two cess of energy efficiency in lighting brand said. "There is no scientific mea­ lamps were removed and one ballast won't depend on utilities alone. It will sure for quality of lighting. And there is was disconnected from each lighting fix­ depend on the efforts of other interest a strong need for this." Until such a mea­ ture. Acrylic lenses were replaced with groups, the success of technology trans­ sure is discovered, he says, decisions on half-inch parabolic-wedge louvers, and fer, and-perhaps most important-the quality will depend primarily on the pro­ sun screens were installed on windows. customer, be it the casino manager or the fessional judgment of lighting experts. There was little or no reflection and the supermarket owner. As Arnold Fickett, In the meantime, though, the federal lighting energy bill was cut in half. vice president of EPRI's Customer Sys­ government has adopted an amendment tems Division, puts it, "The guy who to the 1987 National Appliance Energy Future work owns Joe's supermarket isn't going to Conservation Act that requires fluores­ Technological advances in lighting are worry at all about the next power plant. cent lamp ballasts in this country to meet improving the quality and efficiency of He's going to install more efficient light­ specific efficiency levels. As of January 1, lighting every year, making options for ing if he gets something out of more effi­ 1990, manufacturers will be prohibited utilities and their customers even more cient lighting. You really have to think from producing the least-efficient types abundant. Perhaps the best peek into the about what's in it for Joe." • of ballast currently on the market. The future is offered at Lawrence Berkeley new law also bans the importation of in­ Laboratory (LBL), where researchers are efficient ballasts. developing technologies ranging from This article was written by Leslie Lamarre, science writer. Te chnical information was provided by Karl Johnson and It is almost impossible to determine new film coatings for windows that al- Morton Blatt, Customer Systems Division.

EPRI JOURNAL December 1989 15

onstantly rising productivity began in the 1970s was thus a source of poses a way of looking at productivity and has supported the nation's great concern. By the end of the decade, related issues as they are affected by the use phenomenal rise in living stan­ productivity growth in U.S. manufacturing of electricity. dards during this century. Cre­ had slipped into the minus column (-0.6% The new work builds on a widely ac­ atingC ever more output with less input per average annual growth), a disquieting con­ cepted economic view of the factors unit is the magic that has allowed Ameri­ trast to the 2.2% average annual growth accounting for long-term productivity cans to live well at home and dominate rate of the 1960s. Today productivity seems growth. Ever since the Industrial Revolu­ international markets. About half of the to be on the mend, at least in manufactur­ tion, when the steam engine firsteased the steady gains in U.S. manufacturing output ing, but caution lingers as the 1990s begin. load on groaning muscles, it has been since 1899 can be attributed to growth in A forthcoming book by Sam H. Schurr, known that capital investment in new multifactor productivity, a measure track­ Calvin C. Burwell, Warren D. Devine, Jr., machinery could increase output per ing the amount by which output exceeds and Sidney Sonenblum, prepared under worker. Careful economic studies by the the combined input of labor and capital. the auspices of EPRI's Energy Study Center, Nobel laureate Robert Solow, Moses This single statistic reflects the enormous provides some critical background on the Abramovitz, and others have documented importance of productivity improvements productivity problem and important clues the fact that this gain is not merely a trade­ to sustained economic growth. as to the future. Electricity in the American off, a substitution of capital for labor, but The slump in productivity growth that Economy: Agent of Technological Progress pro- a true net increase in output over the com-

THE STORY IN BRIEF

A historical review of American manufacturing shows a strong linkage over the past centurybetween technological progress and electrifiedproduction tech­ niques. As centralized, steam-powered belt-and-shaft drives of the late nine­ teenth centurygave way after Wo rld Wa r I to decentralized electric motor drives mounted on individual pieces of fa ctory equipment, American manufacturing productivity soared. A surprising concurrent decrease in energy intensity, even at this early stage, points up a key conclusion: electrification of processes allows manufacturers to do more with less-including less energy. Will another pro­ ductivity breakthrough result from the present use of programmable automated controls? Although productivity trends fa ltered with the economic disturbances of the 1970s, the most recent developments in electronics-including sophisti­ cated microprocessor technology for flexible, decentralized control-are expected to reestablish significant gains in manufa cturing productivity in the 1990s.

EPRI JOURNAL December 1989 17 bined input of capital and labor. The source Industrial Energy Puts On a Clean Shirt of this surplus is rising productivity-the ability to do progressively more with less Electricity came into U.S. factories as a clean substitute for the coal-fired steam engine, input per unit. mainly in printing plants and textile mills. But manufacturing operations �till featured a What, then, accounts for productivity central prime mover, rigid arrays of line shafts and pulleys, with leather belts delivering growth itself? An important factor on power to parallel ranks of machines. American society was slow to change, too. We which economists focus is technological remained horse-drawn and wood-fired, unpaved and tree-shaded. progress, mainly as embodied in new plant and equipment and in new ways of orga­ nizing production. Schurr and his colleagues begin where earlier works leave off, with a closer look at the nature of additions to capital stock and organizational changes in manufacturing, in particular as they have been influenced by the use of electrified techniques of pro­ duction. The authors rely on two streams of information: historical descriptions of the evolution of manufacturing technology over the course of the twentieth century, and quantitative measures of manufac­ turing outputs and the productive inputs of labor, capital, and energy over the same period. This evolutionary perspective per­ mits them to identify the underlying trends that characterize the long-term record. Some provocative patterns emerge. Dur­ ing the twentieth century, for example, the use of electricity in manufacturing has grown far more rapidly than any other in­ put: at an average rate of over 8% annually, compared with 1.5% for nonelectric energy, 1.3% for labor, and about 3% for capital. These figures strongly suggest that addi­ tions to the capital stock in manufacturing have been geared predominantly to the use of electricity. What the new study adds to the existing productivity-technology paradigm is a crit­ ical link between technological progress and electrified production techniques. In particular, this analysis is able to relate the ongoing process of electrification to steady productivity gains in U.S. manufacturing during most of the century, and to some surprising energy savings as well. But how does electrification actually spur productivity? Returning to the focus on capital stock as the entry point for new technology, why has the investment in elec­ trified equipment been so important? A

18 EPRI JOURNAL December 1989 partial answer is that the use of new electri­ Electric Motors Reinvent the Factory cal equipment has provided greatly im­ proved ways to effect physical and The unit drive, an electric motor at each machine, was conceived at the turn of the cen­ chemical changes in materials. But of even tury, proved in World War I, and swept the nation in the 1920s. Materials could move more _ greater importance is the fact that electri­ efficiently, and productivity soared as electricity's precision boosted throughput, cut fication has brought about changes in the losses, and even reduced energy intensity. U.S. lifestyles went big-time, the culture went organization of entire systems of produc­ ragtime, commerce climbed up skyscrapers, and everyone stepped on the gas. tion. A look at two critical developments in the history of American manufacturing shows the fundamental importance of these electrically based systemic changes.

Machine drive goes electric The introduction of electric power into the American factory around 1890 triggered changes that eventually led to an explosion of productivity, but not in one giant step. Initially, electricity was treated as just an­ other source of heat and power to supple­ ment steam and other power sources. Its larger potential was not yet clear. In pre-electric factories, machines had to be arranged in long rows parallel to the constantly rotating metal shafts that brought them their power. The shafts were driven from a central source, typically a co,'11-fired steam engine, which turned them continuously from dawn to dusk, no matter how many machines were actually in use. A large plant might contain thou­ sands of feet of metal shafts and leather belts connecting shafts and machinery, as well as thousands of drip oilers suspended above the works for continuous lubrica­ tion. Just one broken shaft could halt work on an entire floor, and a broken steam en­ gine could shut down an entire factory. The initial innovation in the electrifica­ tion of machine drive was to replace the central steam engine with a large electric motor. This first occurred in such industries as printing and textiles, where the grime from coal-fired steam engines was a con­ stant threat to product quality and the need for clean power made it worth the risk to try a new technology. In textile mills, large motors of several hundred horsepower were soon driving over 1000 looms at a time. But the system of power distribution within the factory remained the same. Ma­ chines were still fixed to an elaborate over-

EPRI JOURNAL December 1989 19 head structure of shafts, clutches, belts, capped many years of growth in the use of adoption of electrified unit drive during and pulleys. electrified unit drive. During that decade, the 1920s was stronger, in fact, than the A much more radical step was to outfit electricity use surged at an annual rate of decline during the 1970s, and it occurred each machine with its own electric motor­ more than 9%. But even more significant without the prompting of a severe energy that is, to use electrified unit drive, a devel­ as a manifestation of the triumph of elec­ shortage. The average decline of 4.3% per opment that did not take hold on a wide tric machine drive was the electrification year for the period 1920-1929 compares scale until after the First World War. Ma­ of capital during that period: electricity with about 2.4% per year for the period chines, freed from their overhead tethers, use grew at a rate 13 times that of capital, 1973-1981 (although the rate of decline in­ could be arranged in new ways-ways al­ whereas the use of nonelectric energy ac­ creased in the early 1980s). These figures lowing related operations to be placed tually declined. suggest a new way of looking at energy side by side, for example, to achieve a Accompanying this electrification was a conservation: from this perspective, con­ smoother work flow and a faster rate of surge in productivity. Gains galloped along servation over the long term is primarily a throughput. These more efficient arrange­ at an average 5.2% per year during the by-product of technological progress and ments, plus the instant availability of elec­ 1920s, the highest rate for any decade in this the rising productivity it provides. tric power throughout the factory, permit­ century and a sharp contrast to the 0.8% Overall, then, three major trends stand ted a new emphasis on continuous pro­ average annual rate during the period out: a growing share for electricity in the duction, and output began to soar. In 1899-1920. By correlating historical ac­ energy mix, rising productivity, and de­ observations on some pioneering installa­ counts with quantitative data, these gains clining energy intensity. Typical of U.S. tions, Professor F. B. Crocker noted the im­ can be traced in large part to the transfor­ manufacturing during much of the twen­ portance of unit drive as early as 1901, long mation in production systems made possi­ tieth century, these trends were all strik­ before it was widely used: " ...the output of ble by electrified unit drive. ingly pronounced during the rapid manufacturing establishments is materi­ Paradoxically, the move toward electri­ change sparked by the rise of electrified ally increased ...by the use of electric driv­ ficationalso coincided with a sharp decline unit drive in the 1920s. ing ...with the same floor space, machinery, in the intensity of energy use in manufac­ Microprocessors bring flexibility and workmen. This is the most important turing (the ratio of total energy input to advantage of all, because it secures an in­ manufacturing output). While productiv­ The need for flexibility in manufacturing crease in income without any increase in ity soared during the 1920s, energy use rel­ continued to grow over subsequent de­ investment." ative to output plummeted; the average cades. Eventually, another electricity­ Dismantling the overhead structure rate of decline was 4.3% per year. How based technology emerged that would go cleared up headroom in the plant and al­ could such energy savings occur, given the beyond the flexibility of unit drive and lowed the introduction of traveling elec­ losses that accompany the conversion of trigger another breakthrough in the orga­ tric cranes for more efficient materials primary fuels to electricity? Indeed, it nization of production systems. That tech­ handling. Belt slippage, a major problem could be expected that a growing reliance nology was the microprocessor. with line drive of multiple machines, be­ on electricity in manufacturing would lead While the electric motor provided the came minimal with individual units. Unit to the use of more, not less, total energy. motive force to operate production equip­ drive eliminated power losses due to fric­ Electricity's role as an agent of techno­ ment, the microprocessor has become the tion in rotating shafts, and it saved the logical progress provides the key. Just as key to controlling it. Just as the electric energy that had been wasted in turning electrified production techniques have al­ motor managed energy with a precision shafts when some machines were not in lowed manufacturers to produce a given and flexibility never before possible, so use. A malfunctioning motor stopped level of output with less input of labor and the microprocessor now manages infor­ only one machine, not a whole group or a capital, so too has less energy input been mation. Electricity and the microprocessor whole factory. Savings such as these were required. Conversion losses of primary have become the new production part­ sometimes more than enough to offset the energy are more than offset by the greater ners, playing the same critical roles played capital costs of additional electric motors, efficiency of electrified production tech­ by electricity and the electric motor earlier and beyond them were the enormous in­ niques. Electrifiedprocesses have allowed in this century. creases in productivity that resulted from manufacturers to do more with less-in­ Pre-electric control systems were crude, the reorganization of production flows. cluding less energy. relying on mechanical gears or hydraulic Energy conservation, then, is neither an devices to control machine tools. These Productivity surges invention of the 1970s nor necessarily the early automatic machines of the 1920s The great productivity harvest in manu­ product of deliberate striving. The decline were inflexible: only a limited and fixed facturing came in the 1920s, a period that in energy intensity associated with the sequence of control movements could be

20 EPRI JOURNAL December 1989 Microprocessors accomplished without human interven­ Decentralize Production Control tion. Virtually all progress beyond this stage has relied on electricity and electronic

Computers have maximized the power and precision of our manufacturing muscle, and control devices. microprocessors are now adding the versatility of distributed electronic intelligence­ During the 1930s, electrically- powered making available the full range of market, material, inventory, design, assembly, mechanical devices called servomecha­ backlog, cost, and other information needed to customize production and maximize nisms came into use. Designed with feed­ productivity simultaneously. Thanks to the unique contributions of electricity, today's back capability, the new devices could au­ mass market economy paradoxically offers more variety and more choices than we had tomatically measure particular machine ac­ in the past. And electricity continues to enable much technological innovation. tions, identify the difference between actual and desired actions, and respond to the difference by making adjustments to achieve the actions desired. Servomecha­ nisms were able to direct far more complex sequences of machine movements than their predecessors. But their repertoire of movements was still quite limited and had to be set in advance. A major breakthrough came with the application of reprogrammable controls, which used automated instructions to shift machine operations in the midst of produc­ tion. Initially developed during the 1940s and 1950s, reprogrammable control de­ vices came into widespread use during the 1960s. They allowed production sequences to be altered easily in midstream in re­ sponse to changes in a punched paper tape. The tape was decoded by a mechanical or optical reader attached to the machine itself. The next step was the development of stored-program digital computers. Substi­ tuted for the cumbersome system of paper tape, the computer memory could store in­ structions governing every operation of an automated machine, and reprogramming was only a keystroke away. Since the main­ frame computers of the 1960s were expen­ sive, however, justifying their cost required that each one control a number of ma­ chines. Reliability came to be a problem. When a computer broke down, a number of machines were disabled, just as multiple machines had been disabled by the failure of one electric motor during the very early days of electric drive. These problems set the stage for the mi­ croprocessor, which allowed computer control to take place at the machines them­ selves. Greater flexibility removed the

EPRI JOURNAL December 1989 21 penalty on nonstandard products and tegration. The pioneers of electronic inte­ Electric Power launched the trend toward customization gration may be acting on faith, just as for Energy Efficiency that has since been seen in products rang­ many of the pioneers of electrified me­ ing from automobiles to T-shirts. More chanical drive did. Computer technolo­ One hundred years of technological innovation significant, the microprocessor allowed gies will almost certainly bring further in manufacturing, represented by steadily computer-controlled machines to be linked organizational change and productivity increasing electrification for fast, precise with one another and with the various ac­ gains to manufacturing, but it is unclear at application of energy, has led to sharply lower tivities that support production. what pace those changes will occur. energy intensity-that is, a reduction in the Intel introduced its microprocessor, or overall energy consumption per unit of U.S. "computer on a chip," in 1971. The micro­ Growth falters economic output. processor was programmable in the same So far, the productivity advances accom­ way as much larger computers. Compared panying the adoption of microprocessor with the earliest mainframe, the microcom­ technology in manufacturing have not 0 0 puter of the mid-1970s had more comput­ been as rapid as those that occurred dur­ Energy intensity II 40 ing capacity, was 20 times faster, and cost ing the heyday of electrified unit drive in r-­a, a, only 1/10,000 as much. Thus began the dis­ the 1920s. Poor productivity gains during C 150 X (I) persal of computers to the very sources of the 1970s prompted skeptics to question D 30 C the information they collect and to the very the benefits of computerized control. c "iii processes or machines they control. This Manufacturing productivity had grown at C 20 decentralized system of control, relying on an average annual rate of 2.4% during the � 100 a hierarchy of computers, became known period 1960-1973, before the introduction 10 as "distributed intelligence." of the microprocessor-based technologies.

In the early 1980s, this new system came But just when the benefitsof the new tech­ 50 0 to the fabricating industries. In a typical nologies would have been expected to be­ 1890 1920 1950 1980 configuration, several microcomputer­ gin showing up, productivity growth controlled machine tools are linked by a dropped further, to a low average annual 120 minicomputer, and several minicompu­ rate of 0.2% (1973-1981). At the end of the Multifactor productivity ters are tied in turn to a mainframe. The 1970s, productivity had actually begun to 100 programs for the manufacture of every shrink. 0 part the firm makes are stored in the What happened? One explanation is 0 � 80 that the emphasis during this time was on II central database and transferred automat­ a, ically from the mainframe via the mini­ saving energy, not labor or capital. Be­ a, C computers to appropriate machines in the cause labor and capital are the inputs em­ X 60 network. In addition, information about ployed in multifactor productivity mea­ D the status of every machine, the volume of surements, using more of them in order to its production, and the quality of the fin­ save energy will make productivity fig­ 40 ished parts flows back to the mainframe. ures plummet. Energy conservation prior­ The possibilities do not stop there. An ities led to choices that saved energy but 20+-���-r-���-r-�--c�,---i organization can link design, engineering, failed to support overall productivity 1890 1920 1950 1980 fabrication, production management, and growth. Other possible factors bogging even product distribution and field ser­ down this growth were a large influx of Electric Power vice through a common database. Elec­ untrained workers into the workforce and for Productivity tronic integration allows factories, and the lags in the availability of new equipment. office operations that support them, to be Once these conditions passed, manufac­ The same period has seen sharply higher reorganized for maximum speed and flex­ turing productivity snapped back. The pe­ multifactor productivity, the ratio of economic ibility in response to a constantly chang­ riod 1981-1985 saw a sharp rise in the aver­ output to the combined input of labor and ing competitive environment. age annual growth rate-to 4%. Recent capital. Multifactor productivity is climbing Owing to obstacles ranging from com­ figures confirm that the slowdown of the more steeply in manufacturing alone (shown puter incompatibility to sheer information 1970s has indeed been reversed and that here) than in the total U.S. economy, where overload, very few enterprises have yet the reversal has been sustained. labor-intensive service activities play a realized the full potential of electronic in- Like productivity, electricity use in prominent role.

22 EPRI JOURNAL December 1989 manufacturing has continued to grow, but just as productivity began rising again. Like electrified unit drive decades ear­ much more slowly than it did in the 1920s. That is consistent with the historical pat­ lier, today's microprocessor technology The average annual growth rate for 1973- tern, suggesting that conservation efforts offers benefits typical of electricity-based 1985 was only 0.3%. When electrified unit got a strong boost when the new technol­ production techniques. It provides decen­ drive revolutionized industry during the ogy, which served both conservation and tralized control. It permits flexible utiliza­ 1920s, the ratio of electricity use to capital overall productivity goals, came into tion of plant and equipment as needed. It investment grew at an average annual rate widespread use in the early 1980s. allows for easy expansion in small incre­ of 8.3%. The 1970s saw just the opposite For the full period 1973-1985, then, all ments. And it allows the whole operation pattern: electricity use grew more slowly three major trends continued, but just to be coordinated and integrated into a than capital investment in the period barely. The growth in electrification and smoothly functioning unit. Both electric 1973-1981, resulting in a 3.5% annual rate productivity was quite sluggish by histori­ drive and microprocessor technology of decline in this ratio. What's more, the cal standards and failed to keep pace with have permitted significantreorganization electricity use-capital investment ratio re­ the rapid improvements in energy effi­ of manufacturing operations, with payoffs mained narrowly negative in the early ciency. So the relationship among the in both overall productivity and the effi­ 1980s, even when productivity began to three trends changed. After many years of ciency of energy use. They show how elec­ climb. marching along together, they parted tricity-based technologies can produce Do these figures mean that electricity company briefly during the 1970s. When more output than ever before with less has become less important in manufactur­ the energy shortages began to ease, how­ combined input of labor and capital per ing? Not at all. Partly they reflect recent ever, and new technology became more unit, and with less energy input as well. efforts toward energy conservation. Partly widely available, lagging electricity use As these advances continue, the histori­ they reflect the much higher capital cost of and productivity began to catch up, bring­ cal rate of increase in electricity use could new computerized equipment. But mostly ing the three trends once again more moderate somewhat, because the energy they point to the difference between the closely into step. requirements of some of the newest com­ electricity requirements of the motors that puter-related technologies are minimal. The pattern endures spearheaded the technological revolution Even so, dependence on electricity is like­ of the 1920s and the minuscule require­ The new book by Schurr and his associates ly to continue to grow. It will grow as ments of today's computer technologies. links the dramatic rise in U.S. manufactur­ the heavily energy-intensive materials-pro­ The changes in the relationship between ing productivity during the twentieth cen­ cessing industries continue to shift from electricity and capital do not signal a slow­ tury to electricity-based techniques. In do­ fuels to electricity and as computerized in­ down in the trend toward electrification of ing so, it addresses three interrelated formation management systems reshape capital, but they do point to a slowdown in questions: Why has electricity's share of the nation's productive capacity. the amount of electricity required to total energy risen so sharply throughout In short, Schurr and his associates see the power it. Nevertheless, electricity's share this century? How has this rise been re­ productivity slump of the 1970s as an aber­ of the total energy use in manufacturing, lated to productivity increases? And how ration and the recovery of the 1980s as a now about 40%, continues to increase, can it be consistent with steady improve­ return to normalcy. The pattern is well es­ largely as a result of the growing electri­ ments in the efficiency of energy use? tablished over the course of the twentieth fication of materials processing (an impor­ The answer to all three questions is the century. If the 1990s conform to that pat­ tant trend documented in the book by technological progress that is embodied in tern, electricity's share of total energy will Schurr and his colleagues, but not covered electrified production techniques. Two ex­ continue to rise in concert with healthy pro­ in this article). amples, one historical and one contempo­ ductivity gains and continuing improve­ The third major trend accompanying rary, have shown how such techniques can ments in the efficiency of energy use. technological progress has been declining exert a powerful and pervasive impact on energy intensity. Because conservation the efficiency with which all manufactur­ has been a national policy since the early ing operations are performed. To achieve 1970s, it is difficult to separate out the en­ this impact, both electrified machine drive ergy savings directly attributable to pro­ and microprocessor technology have re­ ductivity growth and the new wave of lied for their effectiveness on a unique microprocessor technology. Energy inten­ attribute of electricity: the exceptional pre­ This article. written by communicat1ons consultant Mary Wayne sity declined rapidly at an average rate of cision-in time, in space, and in scale­ was based on E!ectnc,ty m the Amencan Economy Agent of Te chno!og1cal Progress. a book by Sam H. Schurr. Calvin C. 3.6% throughout the period 1973-1985, with which energy in this form can be Burwell. Warren D. Devine. Jr. and Sidney Sonenblum that will but the drop was sharpest toward the end, transferred. be published 1n 1990.

EPRI JOURNAL December 1989 23 INTEGRATED HOME AUTOMATION

Electronic intelligence for home systems and appliances is only a step away from adaptive responses and interactions that will give us far greater control, comfort, and convenience. The integrated communication channels that make these advances possible can also enhance load management efficiency and economy on both sides of the meter. he promise of automated homes gas and electric. Baltimore Gas & Electric, But questions and doubts still abound, if is no longer pie in the sky. We Columbus & Southern Ohio Electric, and only because "home" is such a strong im­ are close to the reality of house­ Oklahoma Gas and Electric are three com­ age and at times resists change. Also, over­ hold appliances and systems panies with projects under construction. lapping and competing products add their thatT "communicate "-interacting with Manufacturers are coming to market own equivocal note. Therefore, inevitable each other and with us, not to mention with increasingly versatile electronic sys­ as the intelligent home is to a lot of sup­ with the utility, the police, and the firede­ tems that will make intelligent homes a pliers and builders-and not a few utility partment, as well as banks and brokers, reality. Two examples document the point. executives-it sometimes seems as indefi­ airlines and travel agents, school and Unity Systems (Redwood City, California) nite and remote as the often-mentioned pa­ studio broadcasters, and any number first offered its Home Manager in 1985, a perless office. of stores. We're seeing what market ana­ combination of controller, touchscreen, and Simply reeling off examples of products lysts call trend lines. They're profiling software that manages household energy and systems doesn't clearly definethe func­ steady progress, and prototypes of sever­ distribution and use and controls security tional hierarchy that takes us to intelligent al automated designs-including one and sprinkler systems. The Home Manager home environments. In fact, there are three called the Smart House, in which EPRI can monitor up to 24 zones, control airflow tiers: smart products, intelligent sub­ has a hand-are being built in several (and thus temperature) in 20 of them, and systems, and integrated home automation regions of the country. execute as many as six programmed daily systems. We already have smart products in pro­ changes in each zone. Intended mainly for First come smart products. These have fusion. Microprocessor chips are as com­ new houses, an installed Home Manager some data processing power. Products with mon as potato chips in American houses, system costs from $5000 to about $20,000. digital controls, for example, appeared in bringing a measure of intelligence to TVs, CyberLynx Computer Products (Boul­ the 1970s. Among them are VCRs, micro­ VCRs, thermostats, telephones, stereos, der, Colorado) is primarily a software firm, wave ovens, and coffee makers with digital ranges, and even coffee makers. Appliance but since 1985 it has produced a system clocks. Appliances like these rely on digital manufacturers like General Electric and called the Smarthome. Operating on radio control to automate one or more functions. Sony are routinely incorporating program­ frequencies for control signals, the system But at their present stage of development mable controls into their products. is wireless and can be owner-installed in they can't, for the most part, exchange in­ Also on the market are smart subsystems existing homes at a cost of $1000 and up, formation with any other appliance. A mi­ that enable basic, if limited, automation of a depending on the number of sensors and crowave oven, for example, is just that; it few household activities. For some years modules needed. A control unit and a can't accept remote time or power-level in­ now, the X-10 programmable controller has hand-held transmitter permit the program­ structions or even trigger a message on the been popular for operating home lighting ming of HVAC and security systems; pro­ TV screen when food is ready. All it can do and small appliances, registering impres­ gramming changes in lighting requires the is beep for the benefit of people within sive annual sales of over 2 million units. It use of a personal computer. earshot. isn't alone; many other retrofittable designs The number of products and systems As it becomes possible to exchange infor­ that communicate by means of existing available or under development today mation between different kinds of prod­ power wiring are available from hardware prompts proponents of the intelligent ucts, even those of different manufacturers, and builders' supply outlets and from such home to estimate that by 1992 there will be greater productivity in domestic manage­ chains as Radio Shack and Sears. 5000 such dwellings in the United States. ment will become commonplace. This is Moreover, entirely automated houses are Some even predict a 25% penetration of the level of intelligent subsystems. An ob­ being built, featuring specialized wiring new home construction within the next vious current example is the linkage of systems to accommodate all manner of five years. lighting and entertainment controls with present and future functions, some of those of a swimming pool or spa. As an­ which aren't even completely defined yet. Making the distinctions other example, home HVAC controls could The Smart House project, for instance, was Computer literacy certainly figures in pro­ be tied in with a security system so that the founded in 1984 under the auspices of the jections for home automation. Fostered by programmable thermostat would lock the National Association of Home Builders schools and offices, it has stimulated our doors and activate the security system (NAHB), with the participation and advice interest and confidence in residence-based while turning down the heat at some desig­ of manufacturers, utilities, and trade networks where appliances-even of dif­ nated bedtime hour. groups all over the United States. EPRI and ferent types and brands-communicate "The distinctive feature of these intelli­ the Gas Research Institute are closely in­ with each other to enhance the quality of gent combinations," says Arvo Lannus, volved, as are a number of utilities, both our domestic lives. EPRI's program manager for residential

EPRI JOURNAL December 1989 25 A Pyramid of Performance

What we like to call intelligence in the electronic operation of appliances is a matter of degree-from smart products to intelligent subsystems to integrated home automation systems. The progression is marked by greater interconnection for communication and control purposes. It's also marked by increasing sophistication, moilingfrom single appliance identities to entire household functions and services.

Smart Products Perhaps the most familiar smart product today is the videocassette recorder, programmable weeks in advance for repetitive or special-event TV recording. On-screen displays guide and verify the user's step-by-step commands, which can be entered on the buttons of a wireless controller from the ease of a sofa clear across the room.

Intelligent Subsystems Home security is the central feature of one representative intelligent sub­ system, where several appliances are linked so that the timing and mode of operation of one controls the others as well. Given a programmed combination of hallway and outdoor lights, for example, one switch turns on all those lights at bedtime-and at the same time turns down the space heating, engages intru­ -:2all sion alarms, monitors all smoke alarms, and elec­ tronically locks all exterior doors to outsiders.

Integrated Automation Systems An entire home automation system interconnects everything, making it possible not only to perform more functions in concert but also to initiate or modify virtually all of them from any control point, such as a telephone (in the house or elsewhere), a hand-held VCR control, a microwave oven touchpad, a personal computer keyboard, or any of several system controllers. With the advent of smart electric meters and utility com­ munication links, major energy-using appliances will even take advantage of low electricity rates during nonpeak hours. customer systems, "is sensing the status of weeks and months, even of seasons." acknowledges market analyst Tricia Parks one device or appliance as the basis for Veronika Rabl, EPRI's program manager of Parks Associates, who was quoted re­ controlling another. The overall service or for demand-side planning, makes a further cently in Custom Builder. But, she contends, benefit is greater than from either of them distinction. "An intelligent home is adap­ "they will become so as technology and alone. The distinction is even sharper if you tive; it takes account of changed condi­ expertise evolve." think of remote control-for example, us­ tions-the weather, for example, or my pat­ Fully integrated home automation sys­ ing the telephone to program the micro­ terns of appliance switching and use. If I tems process information quickly: their wave oven or the VCR." repeatedly open windows, the heating sys­ software is able to receive several simulta­ One of the earliest (and now least expen­ tem will learn from that and reset itself to a neous signals from different products and sive) intelligent subsystems is the X-10 line lower temperature." systems, resolve conflictsbetween compet­ of controllers and control modules, avail­ Together, their comments emphasize the ing signals, prioritize needs, and make de­ able through Sears and Radio Shack. The notion of functions more than of ap­ cisions. To realize this goal in the most com­ controllers communicate by power line car­ pliances. Instead of telephones, TVs, and plete manner, the NAHB Smart House is rier, sending programmed on/off signals to toasters, the intelligent home deals with laced throughout with multifunction cable other products via the modules at outlets or energy, communications, security, enter­ that carries low-voltage wires for telephone switches throughout the house. The most tainment, comfort, cooking, gardening, and data communications, coaxial cable for sophisticated of the X-10 controllers is pro­ transportation, utilities, and so on. audio and video signals, and power wiring grammed by personal computer but there­ Proponents of the intelligent home think (EPRI Journal, November 1986). Any ap­ after operates independently, controlling as and talk in terms of interaction between pliance from any manufacturer taking part many as 256 devices from wherever it is occupants of the home and appliances in in the Smart House development can be plugged in. A simpler X-10 controller re­ the home-and in terms of a process rather plugged into any outlet, and distributed quires no computer interface; manual pro­ than separate events. For home automation logic chips-in the appliance and at key gramming provides control of up to eight success, the supposition is that people will points in the core wiring-then establish appliances. be able to communicate easily, even casu­ the appliance identity before any current

A key distinction in the hierarchy of ally, with their integrated systems at any flows. Success for the Smart House hinges product and system intelligence is re­ number of access points (called interfaces) not only on the acceptance of its novel core sponse to unusual circumstances. For ex­ or by means of cordless remote controllers. wiring and control system but on the avail­ ample, consider the idea of a digital clock Integrated home systems today typically ability of a full range of compatible prod­ with enough intelligence (and battery have a microprocessor-based central con­ ucts. backup) to recognize power outages and trol unit, applications and system software, seasonal time changes. When the rest of our and a variety of add-on controls for cus­ Establishing communications household clocks become smart enough, tomizing installations. Examples include standards we will no longer face the bothersome Unity Systems' Home Manager and Hy­ Today's gaps in the compatibility of smart chore of correcting them manually. Instead, pertek's Home Brain. Most of the systems products and systems reveal a tension born the master clock will respond to an outage rely on power-line-carrier signals and also of long-standing competitive practices. or programmed time change and send a use dedicated wiring for such functions as Product differentiationis seen by some de­ simple reset message to all the others­ security system or HVAC control. For in­ velopers and manufacturers as a propri­ VCR, microwave, oven, alarms, and so on. stance, Mastervoice's Butler in a Box uses etary advantage and is marketed accord­ telephone wiring, and CyberLynx's Smart­ ingly. This strategy of going it alone re­ Integrated automation systems home uses radio frequencies. quires constant reevaluation as market When people talk about the intelligent or But although these are all called whole­ growth speeds or slackens or as other prod­ automated home, they're talking about far house controllers, none yet controls all ma­ uct lines compete in breadth or capability. more than being able to program the heat­ jor home electronic functions. At best, they Ultimately, of course, the expansion of ing system to lower temperatures when no integrate the functions of several selected home automation systems to accommo­ one is home or to run the dishwasher to­ appliances or a few subsystems. This is date more products will be a logical and morrow or both. Full home automation, the partly a matter of market development; necessary feature for their growing accept­ top of the pyramid, draws this definition people aren't fully aware of what can be ance and market penetration. EPRI's re­ from Arva Lannus: "It's integrated control done. Also, it's a matter of practicality in search managers are convinced that the of all major devices and systems in the adapting the new systems to existing key to cost- effective in-home communica­ home according to the schedules and use houses. "Home automation systems that tions networks will be their versatility in patterns of the residents, over a cycle of offer full integration aren't yet available," getting the most from the increasingly

EPRI JOURNAL December 1989 27

sophisticated electronic products and ap­ ance. The Smart House project, on the other pliances becoming available. The clear im­ hand, offers a proprietary communications plication is a need for standardization in protocol; manufacturers and suppliers ne­ the industry. gotiate agreements to incorporat(:! the Historically, new communications tech­ Smart House communications language nologies develop in an environment of into their designs. There is thus a commit­ competing systems, free from regulation ment to deliver products that work in the The and standardization. But even as far back Smart House environment. Automated as the early 1920s, when competition For the time being, the emerging home among radio broadcasters led strong sta­ automation industry is characterized by Home­ tions to jam the signals of weaker ones, the smart products that speak different lan­ Conversant response was government regulation of guages. If a homeowner wants a thermo­ With what was seen to be a public resource. The stat to communicate with a television, an the Future allocation of specific frequencies came to be interface-an interpreter-that can trans­ welcomed by broadcasters because it guar­ late one signal system into the language of Communication is the anteed to radio advertisers that their mes­ another is necessary. key capability that will sages would be heard free of interference. integrate today's smart Such standardization usually benefits Emerging issues in home products into tomor­ consumers in the long run; our current ex­ automation row's fully automated perience with the VHS format for home Cost is a daunting issue; the range of home systems. Heat, light, video cassettes and VCRs is an example. intelligence prices, like their capabilities, audio-video, food According to Clark Gellings, director of is mind-boggling. Some off-the-shelf X-10 preparation, communi­ EPRI's Customer Systems Division, failure systems retail for less than $100, but Sean cations, security, and of the home automation industry to adopt a Walsh, an electronics dealer in Washington, recreation facilities-all standard communications protocol in the D.C., has said that fully automating a new can respond to central­ very near future could seriously delay ac­ house costs $6000 and up. Indeed, the ized controllers now, ceptance of the intelligent home. "The NAHB Smart House core system (wiring with the help of worst thing that could happen," he cau­ and logic) is estimated to cost $10,000. gateway devices to tions, "would be a market awash in home Rewiring an existing house for a compara­ translate or reconcile automation products that are unable to ble system would cost still more. And spe­ electronic language communicate with one another." cialized whole-house controllers, such as differences. Butwith In the current circumstances, Gellings one from Custom Command Systems that the eventual standard­ points to the Consumer Electronics Bus features voice recognition, cost $40,000 ization of communica­ (CEBus) as a timely attempt to provide stan­ and up. tions, appliances will dard interfaces and a common language for Of course, home automation expense interact automatically. appliances and products from over 50 man­ will fall as market acceptance rises. Mass New data and a loca­ ufacturers, including Honeywell, Intel, production and distribution will bring tion code at one point, Mitsubishi, Motorola, Sony, Texas Instru­ prices down. However, not every system is whether from a sensor ments, and Whirlpool. "The goal is for a winner. Homeminder, a $500 low-end or the homeowner's manufacturers of water heaters, electric wireless controller jointly produced by direct input there, will door locks, cable TV systems, and the like to General Electric and the security system trigger adaptive action have a single set of standardized communi­ manufacturer ADT, failed to sell and pro­ at one or more other cations bridges by which signals can move duction was discontinued. ooints. from one product to another and be under­ Susan Cohen, a marketing administrator stood," Gellings explains. at Rolm Corp., sounds a note of caution. Sponsored by the Electronic Industries "Just because the technology is available Association, the CEBus is an example of doesn't mean that the intelligent home will open protocol, a voluntary communica­ become a reality in middle America. A tions standard available to all manufac­ number of issues need to be resolved, not turers. It's a standard with no contractual the least of which is the need to market a strings, no mechanism to compel compli- clear vision of what an automated home

EPRI JOURNAL December 1989 29 really is to the consuming public." Clark Gellings sees a number of forces shaping and articulating the future of auto­ mated houses. "People want more than Whole-House Control-Almost simply energy in their homes," he says, and he has identified what he calls the key Control is the objective of comprehensive home intelligence today. A representative players in making the intelligent home example, the Home Manager of Unity Systems, concentrates on three major functions­ market a success. They include the manu­ space heating and cooling, security, and lighting and appliances-with principal program­ facturers of entertainment products and ming and control via touchscreen. The system also accepts remote commands and repro­ major domestic appliances, of course. But gramming via telephone or computer. Special sensors monitor such conditions as there are also the suppliers of electric hard­ temperature and motion (occupancy); relays and switches actuate the responses of ware, such as circuit breakers and wiring; forced-air dampers, water valves, pump motors, solenoids, lights, and appliances. the numerous standard- and code-setting bodies of industry, government, and the electrical trade; residential developers and builders (and remodelers); the computer industry; the designers of telecommunica­ tions networks, including local telephone companies; and banks and other institu­ tions (stockbrokers and insurance com­ panies, for example) that have a commer­ cial stake in communications links to their customers' homes. Not least among these key players are electric utilities. "We don't yet know everything that these organizations need to be doing in order to make the intelligent home market a reality," stresses Gellings. "But we do know that there is real potential for electric utilities to take a role in promoting the intel­ ligent home to their customers."

Defining utility interest Quite apart from the traditional competi­ tion with natural gas as a household energy form, electric utilities have a special reason for taking an interest in electronic home automation. As Gellings notes, "Utilities are just now beginning to make the connec­ tion between their need to manage electric­ ity load and the tremendous potential for doing so that is inherent in the automated­ home concept." According to Gellings, it isn't often ap­ preciated that the intelligent home is an ideal environment in which to expand de­ mand-side load management programs. With the permission of their customers, utilities can be fully functioning partici­ pants in automated-home communications networks, not only providing better service

30 EPRI JOURNAL December 1989 to the homeowner but enhancing their own power line carrier-and later on by still to all manner of electricity uses, most obvi­ load management objectives. other media. Design specifications are now ously through improved control and effi­ Many elaborate systems and arrange­ being drawn up. ciency and reduced waste. The range of ments are already in place through which "Control is only one word for what util­ electricity uses and the precision of their utilities and their customers participate ities and householders achieve with elec­ control continue to increase, far ·more so electronically in the operation of major ap­ tronic automation," points out EPRI's than is the case with other energy forms. pliances. Research under EPRI sponsorship Veronika Rab!. "The gateway points up an­ Electric automation is a unique means of indicates that more than 3.9 million U.S. other word, opportunity-making it possi­ taking advantage of those features and ca­ residential appliances have some kind of ble for the residential customer to operate pabilities. Says EPRI's Lannus, "In a house­ communications link with electric utilities. electric appliances and systems more effec­ hold setting more than anywhere else, In more specific terms, this means that tively and economically." value is a matter of perception-of conve­ about 1.5 million central air conditioners, An example of the move toward two­ nience, comfort, speed, responsiveness." about 360,000 space-heating systems, and way communications capability is the Net­ Home automation extends and multiplies more than 1.6 million water heaters can be Comm system, a radio repeater network these features of electricity service, he says, remotely controlled by utilities. Altogether, being developed cooperatively by Met­ "because advances in electric HVAC and these appliances represent about 4 GW of ricom (of Campbell, California) and South­ appliances can so easily be integrated with customer load. ern California Edison. Installed on utility the sophisticated control and communica­ But along with load management goes a poles, the network will carry two-way com­ tions systems now coming on the scene." need for load research. It is a vital and munications between the utility and its cus­ Lannus, Rabl, Gellings, and their EPRI established practice, a necessary invest­ tomers, initially for remote reading of colleagues obviously are convinced. And ment by utilities to guide their system and newly developed electronic meters. Even­ part of their conviction comes from utility ' generation planning. Now, as the commu­ tually, Edison should also be able to initiate feedback, the growing realization that bet­ nications loops of automated houses come and terminate service remotely and add ter load management and better customer into being, utilities can tap into them for time-of-day rates for customer load man­ service are both compelling arguments for real-time data on their customers' ap­ agement. electric utilities to play increasingly active, pliance preferences and use patterns-in­ In another experimental project, South­ even major roles in marketing the intelli­ valuable for fine-tuning service costs and ern Bell and Georgia Power have already gent home. • electricity prices. tested an integrated communications sys­ Remote meter reading is another capa­ tem that provides energy management and bility offered by an integrated electronic time-of-day pricing so that residential cus­ home, saving labor for the utility and elim­ tomers can program their appliances to inating intrusion for the customer. Both take advantage of less expensive rates. will be served by two-way communica­ Home automation encourages customer tions systems and the development of what load management simply because of the EPRI is calling the electric utility gateway. attributes of smart electric appliances and The gateway will be a utility-owned and subsystems. As one EPRI researcher puts it, -installed communications device for ex­ "Efficiency goes with intelligence in this changing information between the utility, business." The HydroTech 2000 heat pump the customer's meter, and the customer's is an example-it is 50% more efficient than automation system. Thanks to a flexible conventional heat pumps and is specifi­ design featuring a modular interface, the cally designed so that its power can be lim­ gateway will be adaptable to home systems ited via utility load control signals. (See that follow the Smart House, the CEBus, or EPRI Journal, March 1988.) The unit incor­ some other protocol. Its foreseen customer porates an adjustable-speed drive and ad­ appeal is its ability to offer time-of-day vanced electronics perfected through re­ rates and then meter money-saving re­ search funded jointly by EPRI and Carrier sponses. Utility-transmitted rate changes Corp. It's one of the products that will be could even become inputs to the cus­ featured in the Smart House. tomer's automated program for energy ser­ But in addition to its load research and This article was written by John Miller and Ralph Whitaker, with vices. The gateway is expected to be acces­ management value to utilities, home auto­ background information from Veronika Rabi and Arva Lannus, Customer Systems Division. sible to the utility by telephone, radio, and mation has a distinct future in adding value

EPRI JOURNAL December 1989 31 enter Celebrating a Decade of Achievement

EPRI's Nondestructive Evaluation Center, where research results are transformed into practical and fieldworthytools, has racked up an impressive record of accomplishments in its first decade of operation-a performance recently celebrated by the people who made it happen. en years ago, EPRI's Board of Di­ Group, was selected on the basis of com­ tions. It may require the purchase of new rectors gave the green light to a petitive proposals to organize, build, and equipment, and utilities want proof that the bold project proposal: the for­ operate the facility, the first of its kind in capital outlay is justified. And it may re­ mation of a special facility with a the United States. quire retraining personnel. The center was missionT so novel it had to be defined in The need for reliable, fieldworthy equip­ set up to smooth over those obstacles and terms of what it was not. The mission was ment and procedures is especially acute in reduce the risks involved in introducing not to be R&D; that was already being man­ nuclear power plants. Periodic in-service new equipment to the field." aged at EPRI's Palo Alto headquarters. It inspections of nuclear power plant compo­ The center does this by evaluating the was not to be an applications operation; nents are required by federal law and must performance of NDE hardware, software, that was already being performed by ven­ meet strict standards established by the and procedures; by demonstrating and dors and utilities. The new facility was in­ Nuclear Regulatory Commission (NRC) documenting the performance of systems; tended to occupy a position in between. It and the American Society of Mechanical and by providing workshops and training would serve as both proving ground and Engineers (ASME). Hidden and often subtle courses to allow NDE inspectors to become training camp-a place where new tech­ flaws, such as cracks that originate on the proficient in the use of the equipment. In nologies and procedures for power plant inside diameter of pipes or in welded joints, addition, the center maintains a hotline that component inspection could be evaluated can, if left undetected, lead to forced out­ provides utilities with rapid response to and improved, and where personnel could ages to repair the affected component. urgent problems, a service that has proved learn to become proficient in their use. Forced outages aren't cheap. In addition to to be one of its most appreciated features That facility, EPRI's Nondestructive Eval­ paying for the repair, the utility may have and that has strengthened the rapport be­ uation Center in Charlotte, North Carolina, to purchase replacement power, which can tween center staff and plant personnel. has proved to be one of the Institute's most push the cost of an outage to nearly $1 While all of the center's achievements successful and popular projects. The cen­ million per day. can't be readily quantified, those that can ter's anniversary provides an opportunity Like physicians examining patients for be provide a quick snapshot of the facil­ to highlight a decade's worth of achieve­ internal disorders, power plant inspectors ity's value to the industry. As of 1988, ments that have saved utilities nearly half a examine components with X rays, ultra­ nearly 5000 individuals had participated billion dollars by turning EPRI research re­ sonic sensors, and other diagnostic tools in 386 training classes; there had been 121 sults into practical tools for the field. that reveal internal flaws without the need instances of rapid response assistance to for disassembling the component or cut­ 27 different utilities for site-specific prob­ Putting R&D to work ting into the metal. Such nondestructive lems, usually involving a plant visit by EPRI's research and development pro­ evaluation is performed to ensure that center staff; and the center had published gram has provided utilities with an array components do not contain flaws. If flaws 106 reports dealing with real utility issues. of advanced technologies, tools, and pro­ are present, their size and location must be Taken together, the center's activities have cedures to help them keep power plants determined to see if the component can be provided significant assistance to nuclear operating reliably, economically, and returned to service. utilities, and the results of its work are safely. But an unproven prototype doesn't "The NDE Center tackles the issue of being increasingly applied to nonnuclear have much practical value in the field, technology transfer," says EPRI's Gary Dau, power plants as well. EPRI has docu­ where equipment has to perform reliably who conceived of the center, directed its mented utility benefits in excess of $430 under what are often very harsh condi­ formation, and managed it for the Nuclear million on project expenditures of $28.1 tions. A power plant just isn't the place to Power Division for the past 10 years. "The million through 1988-a benefit-to-cost ra­ debug a new piece of gear or to learn new center was established to speed the transfer tio greater than 15 to 1. In addition to this procedures by trial and error. of EPRI research results into routine field handsome return on investment, the cen­ Responding to the need to get newly application-to serve as a bridge between ter has provided other, less tangible bene­ developed inspection technologies and the developers and the users of the technol­ fits.According to Dau, these include serv­ procedures into the field, EPRI's Board of ogy." Efforts to bring a newly developed ing as a focal point for objective NDE in­ Directors authorized the establishment of technology to the fieldcan encounter some formation and assistance and providing the NDE Center in 1979; the goals were to obstacles, Dau says. "Field personnel may continuity to the technology transfer effort address nuclear NDE issues in an objective be reluctant to accept a new piece of NDE on a long-term basis. manner and to provide field-qualified equipment or a new procedure for several equipment, procedures, and personnel reasons," he explains. "The technology Emphasis on realism, results training to the industry. J. A. Jones Ap­ may not be fully field-tested-there may not At the NDE Center, the emphasis is on plied Research, a subsidiary of the Jones be enough experience under field condi- solving real-world problems-with the ul-

EPRI JOURNAL December 1989 33 timate goal of giving plant personnel the component-by-component basis and saw cut are quite different from those of an practical tools and the training they need make further adjustments until they are actual flaw, so it's not a realistic test." To to do their jobs better. Comprising 67,000 satisfied that the equipment is ready for provide the needed realism, the center has square feet of laboratories, fabrication and field testing. But the equipment does not brought in retired power plant compo­ machine shops, offices, classrooms, and go into the field just yet; instead it is nents that have actual flaws, including high-bay areas, the center is well equipped brought to one of the high-bay areas and sections of reactor pressure vessels, pip­ for meeting the goal. put through its paces on a variety of sam­ ing, coolant pumps, and steam generator When a newly developed piece of NDE ple flaws. tubes. In addition, the staff has developed equipment arrives at the center, the tech­ "In a normal operating environment, techniques to implant realistic flaws in nology transfer process begins in the of­ it's very rare to encounter flaws of any simulated component sections. This as­ fice area, where staff experts evaluate the kind," says Dau. "Part of the challenge of sortment of flawed specimens housed in system and if necessary add improve­ evaluating the performance of a specific the high-bay areas is used to benchmark ments-for example, making software NDE system lies in having adequate sam­ the performance of NDE equipment, as menus more user-oriented. The process ples to test it against. The cheapest ap­ well as to give trainees hands-on experi­ then moves to the laboratories, where cen­ proach is to take a piece of metal and make ence in detecting the types of flaws they ter personnel evaluate the system on a a saw cut in it to simulate a crack. But the may encounter in the field. ultrasonic reflection characteristics of that

Bridging the Gap Between R&Dand Applications

Established to transform research results into practical tools and procedures fo r field personnel, the NDE Center provides a bridge between developers and users of advanced inspection technologies. By evaluating inspection hardware and procedures, demonstrating equipment performance, and conducting training programs, the center reduces the risks of introducing new equipment and procedures into the field. Inspection challenges into the material at a particular angle of over cracked areas to maintain the Perhaps the most significant progress incidence, and the reflected signals are strength of the pipe wall. These repairs made by the center has been in improving displayed on an oscilloscope. If the waves involve building up the outside diameter ultrasonic testing technology and proce­ encounter a flaw, the reflected signal of the pipe with a series of weld beads to a dures for detecting cracks in the primary changes and the operator interprets the thickness sufficient to compensate for the piping of boiling water reactor (BWR) signals to determine the flaw's size and crack. But the problem is that the overlaid power plants. Since the rnid-1970s, inter­ location. Reliable ultrasonic detection of material makes it more difficult to inspect granular stress corrosion cracking (IGSCC) IGSCC in BWR piping is made difficult both what's underneath. Because of concerns in welded joints in this stainless steel pip­ by the nature of the cracks and by the about inspectability, the NRC specified ing has been of major concern to utilities nature of the pipe itself. The cracks are that a utility could use weld-overlay re­ operating BWRs; indeed, the IGSCC in­ narrower than other types of cracks and pairs for no more than two fuel cycles spection challenge was one of the ind ustry usually produce lower-amplitude signals. unless NDE techniques were shown to be issues that prompted the idea for the cen­ And stainless steel has large grains that reliable; otherwise the piping would have ter in the first place. can mask or mimic subtle flaws. to be replaced, a costly and extensive pro­ To locate IGSCC flaws, an inspector or Further complicating the inspection cedure that also exposes personnel to radi­ an automated scanner passes a transducer task are the weld-overlay repairs applied ation. back and forth across a weld. The trans­ ducer sends high-frequency sound waves

FromPr ototype to Field-QualifiedHa rdware

Ensuring that newly developed inspection technology will perfo nn reliably is one of the NDE Center's main missions. Center experts evaluate inspection hardware on a component-by-component basis, document its perfonnance on realistic flawsamples, and, if necessary, add refinements until they are satisfied that the system is ready fo r fielddu ty. The equipment is then taken to one of the facility's high-bay areas, where it is tested on full-sized power plant components under simulated field conditions. In response, the NDE Center and the reports, provided on-site assistance by The nondestructive evaluation of heavy BWR Owners Group mounted a concerted rapid response teams, and developed for­ section components, such as reactor pres­ effort to develop and demonstrate effec­ mal training courses in IGSCC detection sure vessels and reactor coolant pumps, tive ultrasonic methods for monitoring and sizing. presents special problems. Their size, IGSCC through weld overlays and in­ The effort paid off in a big way. The NRC thickness, and materials make them diffi­ specting the overlay material itself. This accepted the new technology as effective cult to inspect; yet sensitive and reliable effort included the transfer to industry of and, instead of creating its own qualifi­ examinations are required to ensure the several pieces of inspection hardware, in­ cation program, now requires that BWR integrity of these critical components. cluding an advanced automatic ultrasonic piping inspectors successfully complete To examine heavy section components system called Instraspect and the ALN the center's courses. Using the center-de­ not amenable to inspection by conven­ 4000, a computer-based signal-processing veloped NDE technology and techniques tional ultrasonic or radiographic methods, system for analyzing data collected by has allowed several utilities to avoid costly EPRI developed a miniature high-energy manual and automated pipe-scanning pipe replacement. To take one example, X-ray system called MINAC (Miniature systems. To bring BWR utilities up to speed Georgia Power estimates that applying the Accelerator). This device has permitted, on the new equipment and improved pro­ improved inspection techniques at its for the first time ever, inspection of the cedures, the NDE Center staff published Hatch Unit 1 BWR plant helped the utility circumferential stainless steel welds in re­ avoid $135 million in pipe replacement actor coolant pumps. MINAC is lowered and purchased-power costs.

Rapid Response to Urgent Problems

Fast access to NDE expertise is available through the center's rapid response hotline. Utilities in need of quick solutions to critical, short-tenn problems can tap center experts for objective infonnation and assistance on NDE-related issues. Such direct assistance may involve consultation, evaluation of inspection data, a site visit by center staff, or use of the center's facilities. Rapid response activities have paid off considerably by saving man-hours and reducing or eliminating costly plant downtime. into a coolant pump by a manipulator Pacific Gas and Electric's experience with 10 times as powerful as the original sys­ arm, which rotates and tilts the device to an advanced version of MINAC illustrates tem, PG&E was able to conduct its inspec­ aim its X-ray beam at the area of interest. the system's value. During a refueling out­ tion while the plant was on-line and veri­ "MINAC has done its job so well that the age at Unit 1 of PG&E's Diablo Canyon fied that the valves were in good operat­ NRC has waived the requirement for reac­ plant, an inspection of the main steam re­ ing condition. This first application of tor coolant pump inspections," says Dau. verse-flow check valve showed that a cot­ MINAC-6 saved PG&E an estimated $2-3 "On the basis of the favorable inspection ter pin and nut were missing from the million by avoiding a week-long shut­ results, the commission now allows the valve disk. The NRC required the utility to down and the purchase of replacement inspection to be postponed until the pump verify within two weeks that the cotter power. Also, the utility will be able to com­ has to be disassembled for other reasons, pins in four similar valves in Unit 2, which pare the results of the MINAC-6 inspection representing significant savings for the was in operation, were correctly locked in with subsequent inspections to detect any utilities involved." position. Inspecting the first valve had deterioration in the valves. The NOE Center made suggestions that been easy, since it was disassembled dur­ were incorporated into an improved ver­ ing the outage; but unless PG&E could Steam turbine inspection sion of the original MINAC design, making inspect the Unit 2 valves while the plant Catastrophic failures of steam turbine the system smaller and more powerful. was on-line, it would have no choice but to components have highlighted the need for shut down the 1100-MW unit for five to seven days. Using an advanced version of MINAC called MINAC-6, which is almost

Tr aining for the Real Wo rld

Practical training courses help utility personnel become proficient in the use of advanced inspection tools and techniques. The center's curriculum is designed to meet the real-world needs of the people charged with keeping their plants in good operating condition. Formal book-and-classroom instruction is supplemented with hands-on experience in detecting realistic flawsin actual power plant components. effective NOE to detect flaws early, as well TheMe asure of Success $430 million as to predict the life expectancy of compo­ nents. The center's turbine program has evaluated and transferred to industry im­ The NDE Center's accom- proved turbine rotor inspection systems Utility benefits and also a computer program for estimat­ plishments are directly trans­ ing remaining rotor life. Making such an estimate isn't easy; in addition to NOE, the lated into benefits to member disciplines of stress analysis, heat transfer, materials science, and fracture mechanics utilities in the form of finan­ must be employed. In the past, rotor life­ time analyses were performed by turbine cial savings and an expanded manufacturers, whose assessments about whether a rotor should be repaired, re­ pool of NDE expertise. These placed, or retired tended to be overly con­ servative. benefits are ultimately passed Ultrasonic examination of turbine rotors can be performed with an EPRI-developed on to customers in the fo rm inspection system called TREES (Turbine Rotor Examination and Evaluation Sys­ of lower-cost power and more tem), which moves down the center of the rotor bore and extends transducer-tipped reliable generating plants. spider arms that make contact with the bore surface. Boresonic data from the scan are sent to a computerized data acquisition system, which produces a map of flaw loca­ tions and sizes. Together with material 4269 property data and projected turbine oper­ ational cycles, the boresonic data can be Persons trained used with the EPRI-developed computer 121 program SAFER (Stress and Fracture Eval­ NDE uation of Rotors) to predict rotor lifetime. Rapid responses GRADUAT E Several utilities are now using TREES /-1,t,vfvvvvrvt · and SAFER to perform their own rotor in­ spections and analyses, thereby reducing their dependence on the manufacturers. � ·;; This capability provides a utility with a f5 <{ second opinion on a rotor's condition and allows it to assess any conservatism in the manufacturer's analysis. As a result, util­ ities can prevent both rotor failures and premature replacement. Duke Power now uses SAFER to evaluate 10 to 12 rotors a year. The code has helped the utility to avoid premature replacement of one rotor and to extend inspection intervals; Duke expects a IO-year levelized annual saving of $2.3 million. The center's turbine program is not lim­

ited to rotor bore inspections. Another tur­ a, CX) bine inspection challenge is presented by a,

38 EPRI JOURNAL December 1989 the region where turbine blades are at­ "We've seen more and more examples nothing gets lost in this process of integra­ tached to the turbine disk rim. The combi­ of utilities applying the center's work to tion," he says. "We won't lose anything nation of high centrifugal stresses and nonnuclear power plants," says Dau. "The that the utilities have come to depend on steam corrosion can initiate cracks in the mandatory inspection requirements im­ us for. Instead we are going t9 offer them "Christmas tree" connections between the posed by the U.S. Code of Federal Regula­ more. We've got a real winner in the NOE blades and the rim. Such cracks can be tions for nuclear plants serve as a forcing Center, and we're going to build on that." detected by magnetic particle testing, but function to improve the technology. A The NOE Center has clearly played a key this method requires removing the blades similar forcing function doesn't exist for role in bringing the benefits of advanced and later installing new ones. A cheaper nonnuclear plants. But once the improved inspection technology and procedures to alternative is a manual ultrasonic inspec­ inspection techniques for nuclear plants the nuclear power industry. Moreover, the tion, but this is both unreliable and diffi­ are available, their adaptation to non­ center staff has strengthened the rapport cult to perform-the complex geometry of nuclear plants is usually straightforward. between EPRI and the end users of the the connections produces many extrane­ It's very gratifying to see more transfer of technology developed by the Institute's ous signals, and the transducer must be this type." R&D programs. Through workshops, for­ placed precisely. mal training courses, and face-to-face con­ Pacific Gas and Electric was again the Building on a winner tact with center experts during plant beneficiary of the center's work. PG&E As the NOE Center approaches its second visits, utility field personnel have come to was recently faced with the task of inspect­ decade of operation, it will be taking on a depend on the information the center pro­ ing the blade-fit area at its Geysers geother­ broader mission, according to James Lang, vides and have also been able to commu­ mal power plant, where corrosion can be a program manager for nuclear plant oper­ nicate their own needs to the center staff. problem. Recognizing the limitations of ations and maintenance, who is assuming "We've made a concerted effort to get magnetic particle and manual ultrasonic management of the center. The scope of our people into the field and to make con­ inspection, the utility elected to employ a the center's activities is being expanded to tact with the people who are actually oper­ novel approach using two technologies include nuclear maintenance applications ating and maintaining plants," says Lang. from the NOE Center. To ensure proper projects that in the past have been sepa­ Not only does the facility bridge the gap placement of the ultrasonic transducer, rately managed by EPRI. These include the between R&D and applications, he says. "It PG&E used a new personal-computer­ Nuclear Power Division's Monitoring and serves as a bridge between people-and based beam-plotting program developed Diagnostics Center, the Service Water As­ it's a two-way bridge; information flows at the center. Called PC Raytrace, this so­ sistance Program, and the Maintenance both ways. We need to know what the phisticated code traces the path taken by an Equipment Applications Center (which is field users want so we can give it to them ultrasonic beam as it passes through the located at the NOE Center but has been in a form they can use. Without effective area being examined and instantly displays funded and managed independently). two-way communication, it's very easy to the beam on the computer screen. In addi­ Consolidating these projects under a get disconnected, to lose sight of what's tion to allowing precise transducer setup to single program will, according to Lang, really needed in the field." ensure that ultrasonic signals are sent to the provide more efficient and streamlined The need to concentrate on results that area of interest, the use of Ray.trace helps in management of complementary projects count is echoed by Gary Dau as he recalls the evaluation of the data. PG&E also used and will allow EPRI to take advantage of the NOE Center's rich accomplishments UDRPS (Ultrasonic Data Recording and shared resources. "Integrating our nuclear over the past decade. "Getting research re­ Processing System), which captures re­ maintenance applications activities into sults into a form that is both usable and flected signals and displays them for anal­ one package will let us provide utilities acceptable to the people in the field is a dif­ ysis, revealing very subtle changes in the with a one-stop nuclear maintenance ficult task," he says, "but one that has tre­ signals that may be caused by cracks. shop," says Lang. "By picking up the mendous payoffs if you can make it By using Raytrace and UDRPS, PG&E phone and calling one 800 number, they'll work." • can perform 50 examinations in the time be put in touch with the person who can previously required to do a single manual answer their maintenance questions­ examination; moreover, it can perform whether they involve NOE, how to fix a examinations that were not possible or pump, how to repack a valve, or how to fix practical with earlier, manual methods. an electrical breaker." Lang emphasizes For PG&E, the net savings from these com­ that the integration will not serve to dilute This article was written by David Boutacoff, with background puter-based tools are estimated to be the NOE resources that have been the cen­ information from Gary Dau and James Lang, Nuclear Power Division. about $195,000 per year. ter's focus. "We have to make sure that

EPRI JOURNAL December 1989 39 range of cylinder thickness-to-diameter Utilities frequently rely on outside con­ ratios, crack lengths and depths, material sultants to perform flaw evaluations, TECH strengths, and load combinations. Today's which is time-consuming a.nd expensive. prescribed guidelines therefore don't con­ The Ductile Fracture Handbook gives plant sistently yield true safety margins. In fact, owners an in-house capability. "If a utility TRANSFER they often are overly conservative. Unless engineer is familiar with this handbook, plant owners can demonstrate a larger he can do the evaluations in most cases, margin of safety than is indicated by ASME saving the cost and possibly the delay as­ NEWS code procedures, evidently flawed piping sociated with hiring outside consultants," must be replaced, often unnecessarily or Gilman points out. Virginia Power, for ex­ prematurely. ample, recently saved $20,000 by using the To meet the need for unified, extended, handbook, he says. and updated ductile fracture mechanics An aerospace firm has expressed inter­ formulas for cylinders, EPRI and Novetech est in the Ductile Fracture Handbook, ac­ Analyzing Cracks in Corp. have developed the Ductile Fracture cording to Gilman. He believes that any Ductile Steels Ha ndbook. Now available is Volume 1, plant or industry using pressure compo­ ecent research in fracture mechan­ Circumferential Throughwall Cracks, the first nents can benefit from it. • EPRI Contact: R ics-the science of structural material of two volumes of formulas for evaluating Joe Gilman, (415) 855-8911 characteristics, stresses, and flaws that the integrity of flawed cylinders in nuclear compromise integrity-makes it possible plants. It is the first consolidated source Gehl Named to shift from an emphasis on brittle frac­ of formulas for evaluating throughwall to FAA Committee tures to include the realm of ductile frac­ cracks in tough ductile steels. Formula ometimes technology transfer in­ tures. The new focus better represents the types include linear elastic stress-intensity volves a kind of cross-fertilization be­ character of pressure boundary structures factor, J-integral, limit load, tearing mod­ tweenS researchers in separate industries. in nuclear power plants, whose piping and ulus, and crack-opening displacement For electric utilities and commercial avia­ pressure vessel materials are selected for and area. The volume also provides for­ tion, this process may be stimulated by the ductility-the ability to stretch and deform. mulas for obtaining fracture resistance recent appointment of EPRI Subprogram Unlike brittle fractures, which are straight­ values from pipe tests. Manager Steve Gehl to a Federal Aviation forward and easy to predict, ductile frac­ The handbook provides the structural Administration (FAA) advisory committee tures are complex; calculating their behav­ analyst with solutions that are easy to ap­ focusing on nondestructive inspection of ior requires the use of advanced com­ ply to problems involving cylindrical struc­ older aircraft-structures, engines, and puters. tures under a variety of loading conditions. various subsystems. Ten years ago, EPRI, the NRC, and others Plant owners can thus demonstrate, for ex­ Gehl's advisory appointment results undertook research to develop basic duc­ ample, that an undetected crack would be­ from his EPRI experience in quantifying tile fracture methodology. Additional re­ come evident by leakage long before be­ NDE and life assessment information and search extended ductile fracture me­ coming dangerously unstable. integrating them into power plant mainte­ chanics to cylindrical structures, a signifi­ Owners can also show where a repair nance planning and repair I replace deci­ cant development for nuclear utilities based on the current ASME code would be sions. because so many plant components are too conservative and entail premature or According to Gehl, technology devel­ cylindrical-hundreds of miles of pres­ unnecessary replacement. "Appendix A oped under EPRI sponsorship may have sure vessels, pipes, and valves, all subject of ASME Section XI specifies technically immediate beneficial use in meeting avia­ to stress from the high temperatures and adequate and safe procedures for evaluat­ tion industry inspection needs. Particu­ pressures in the nuclear primary cooling ing flaws, but compliance isn't manda­ larly promising for transfer is a flexible system. tory," says Joe Gilman, program manager eddy-current coil device developed for in­ Many of the piping and vessel flaw eval­ for structural mechanics in EPRI's Nuclear specting turbine blades that could have uation procedures derived from that ef­ Power Division. "When a utility can use application in the nondestructive exam­ fort were incorporated into ASME codes. another method for determining the integ­ ination of aircraft engines, wings, and fu­ However, the research did not cover rity of power plant structures and justify selages. many factors-in particular, the entire it, this is generally acceptable." The FAA committee work could be a

40 EPRI JOURNAL December 1989 two-way street, according to Kurt Yeager, routine maintenance, overhead, and im­ terpret the state of a plant and respond to EPRI vice president for generation and provements will be paid proportionally by any upset in a context of prescribed emer­ storage R&D. "Steve's appointment will CQ Inc. customers. gency procedures that are very complex. give us a window into airline inspection Utilities, coal companies, and manufac­ To aid their efforts and reduce the chances methods and decision processes, some of turers of coal-cleaning equipment will of error, EPRI developed the Emergency which may be translatable to utilities." have direct access to CQ Inc.'s staff and Operating Procedure Tracking System Gehl says there is a lot of ground to facilities. The subsidiary will be able to (EOPTS) and demonstrated its worth at cover. "Nondestructive examination is a perform proprietary research for its cus­ Taiwan Power Co.'s Kuosheng BWR train­ large field-not just hardware and soft­ tomers (EPRI's nonprofit status has pre­ ing facility. ware, but also training programs and the cluded this), but the results of EPRI work Results from that testing show that planning and execution of inspections and EOPTS has a marked effect on the perfor­ related maintenance. Utilities and the avi­ mance of control room crews. Crews us­ ation industry can learn from each other in ing the system displayed greater consis­ all these areas." tency, experienced fewer discrepancies, Metallurgy is the foundation of Gehl's and achieved greater success in recover­ work at EPRI-three years with materials ing from discrepancies when they did problems that affect fossil power plant occur-improvements that culminated in performance and availability and four ear­ a 60-75% reduction in response time. lier years on nuclear fuel rod performance. The computerized system monitors Gehl joined the Institute in 1982 after plant performance and status factors that seven years at Argonne National Labora­ are essential to symptom-based emer­ tory. • EPRI Contact: Steve Gehl, (415) gency procedures. Then, in an emergency, 855-2770 will be publicly available. Unlike EPRI, it coordinates all available data with high­ CQ Inc. will be a government contractor, level cues that show operators which por­ Coal Quality Center working directly for DOE and state agen­ tions of the emergency procedures are Goes Commercial cies. "This should result in a greater over­ applicable-thereby enhancing operator PRI's Coal Quality Development Cen­ all investment in R&D that directly bene­ ability to interpret and apply procedures E ter is changing its status. As of Janu­ fits utilities," says Clark HarrisoJ, long­ and, ultimately, reducing human error ary 1, 1990, it will become part of CQ Inc., time EPRI project manager at the center. and increasing safety. EPRI's first subsidiary, ready to work for EPRI will continue to fund several proj­ EOPTS takes the form of a compact, fast­ utilities, government agencies, and others ects, with CQ Inc. as the contractor. The running software module that uses expert in a profit-making mode. CQ Inc. will ongoing work includes a DOE-cofunded systems technology to integrate a safety maintain the CQDC research and develop­ demonstration of advanced coal-cleaning parameter display system with emer­ ment program for EPRI, DOE, and state technologies, investigations of trace ele­ gency operating procedures developed by agencies; but it also will independently ments in coal and methods for removing the Boiling Water Reactor Owners Group. assist utilities with coal-purchasing deci­ them, and development of a coal quality The system was installed and evaluated sions; test new coal-cleaning, -blending, expert system. EPRI will also continue its with human factors techniques for six ac­ and -handling equipment for manufac­ cosponsorship (with six utilities) of a proj­ cident scenarios: anticipated transient turers; and solve coal-cleaning-plant de­ ect on coal handleability-that is, how well without scram (ATWS), radiation release sign and operating problems for coal com­ coal moves through power plant equip­ accident due to steam line break, loss of panies. ment. • EPRI Contact: Clark Harrison, emergency core-cooling system, loss of re­ The new organization enables EPRI to (412) 479-3505 actor pressure vessel level indication, loss­ continue R&D and demonstrations even of-coolant accident with drywell/ primary though its financialcommitment will con­ Expert System Guides containment hydrogen control, and ATWS tinue to taper off from the level of a few Emergency Procedures with abnormal suppression pool level. years ago. The Institute will own all CQ critical aspect of nuclear plant safety Ta iwan Power is using EOPTS for oper­ Inc. stock but will transfer ownership of A is what control room operators do to ator training. U.S. utilities can obtain the the facility, and will license several R&D prevent transient events from developing system from General Electric. • EPRI products, to the new subsidiary. Costs for into serious accidents. Operators must in- Contact: Bill Sun, (415) 855-2119

EPRI JOURNAL December 1989 41 RESEARCH UPDATE Environmentaland Health Assessment TOXRISK Update by Abe Silvers, Environment Division

tility companies may use and produce upper limits to human risk by applying statis­ divided by the number of animals in a dose U chemicals whose effect on human tical dose-response procedures to laboratory group to yield the proportional rate of response health must be evaluated. The risk to human animal data. Often, animal results involving to the dose. With this dose-response informa­ health from exposure to a potentially toxic laboratory exposures many times higher than tion, two extrapolations are performed. The chemical is determined by two equally impor­ that of any human population must be used in dose-response relationship is extrapolated to a tant factors: the level of human exposure and making these calculations because human low-dose region to estimate a safe dose at an the potency of the chemical for causing a data suitable for risk estimation are not avail­ acceptable risk of one extra cancer in a large health effect. To ensure a safe occupational able and because the use of lower exposure population (conventionally taken to be be­ setting and to protect the public and the envi­ levels may not allow the detection of response tween 10,000 and 1,000,000 exposed individ­ ronment, utility risk managers assess both of in even a large laboratory study group. To en­ uals). A statistical model is fit to the data and is these factors for the chemicals of concern. able utilities to consider these animal data, used to calculate the extra risk of cancer for a The risk from chemical exposure decreases EPRI sponsored the development of TOXRISK small incremental exposure. The correspond­ with exposure level and becomes vanishingly (RP2310-2), a software package that facilitates ing dose is also extrapolated to a dose for small at very low levels. This concept of the risk assessment. humans. TOXRISK dose-response relationship is fundamental to TOXRISK features The basic function of is to construct toxicology. It implies that human health can be dose-response relationships for health effects, adequately protected by controlling exposure Risk assessment methods predict the proba­ generally cancer, from animal data. The user to a toxic chemical. bility of a toxic effect from exposure to a chemi­ can supply estimates of exposure to obtain over­ When assessing risk from exposure to po­ cal substance. A group of animals, usually rats all estimates of excess risk (the extra probability tentially cancer-causing chemicals, the Envi­ or mice, is exposed to a particular material at of health effects occurring as a result of expo­ ronmental Protection Agency and other regula­ various doses; the subjects are then evaluated sure). Alternatively, the user can use the esti­ tory bodies often make the health-protective at the time of death (natural or induced) to mated dose-response relationship to predict ex­ assumption that any exposure, no matter how determine if they developed cancer or some posure levels consistent with predetermined small, entails some risk. This assumption leads other disease. If cancer is of primary interest, levels of excess risk (e.g., 1 in 1,000,000). In to the calculation of what are called plausible the number of autopsied animals with tumors is either case, the user selects the method for making the animal-to-human conversion. TOXRISK facilitates the input and manage­ ment of toxicological data and automatically ABSTRACT EPRI developed the computer program TOXRISK to makes statistical and mathematical calcula­ tions. It performs standard statistical tests on the help utilities assess the potential health effects of chemicals they use data to determine dose-related effects. The soft­ ware's data management capabilities ensure and produce. A new version will soon be released that includes models that the appropriate mathematical and statistical procedures are being applied. Its improved capable of estimating riskas a function of time to exposure as wellas computer algorithms permit calculations to be performed interactively in most cases. level of exposure. Other enhancements to the program include im­ To allow the assessment of risk under a range of alternative assumptions, TOXRISK fits proved graphics capability and increased user-friendliness. several mathematical dose-response models to the toxicological data and calculates related statistical parameters. One of these models is

42 EPRI JOURNAL December 1989 the linearized multistage model, which is Figure 1 This TOXRISK screen shows an example of data for a time-to-tumor model in which some animals die from a tumor (F) and other animals die from competing causes (C). The time to death widely used by regulatory agencies. The user is expressed in weeks from the time the tumors were observed. can be confident that the implementation of 3 this model is consistent with regulatory prac­ INCIDENCE DATA, GROUP 2 of tice, since the developers of TOXRISK also de­ DATASET NAME : C:\ TOXVER2\ TEST. TXS\ TIMEl veloped the programs used by the regulatory agencies. Other models covered by TOXRISK TITLE : Test for time-to-tumor models ROUTE/DOSE UNITS : FOOD (mg/kg/day) SPECIES : RAT are K-stage models (for Kfrom one to six}, the Weibull model, the log-normal model, and the TIME # OF TUMOR TIME # OF TUMOR TIME # OF TUMOR Mantel-Bryan model. TOXRISK can produce (weeks) ANIMALS CONTEXT (weeks) ANIMALS CONTEXT (weeks) ANIMALS CONTEXT 9 1 C 35 1 F 45 2 F graphs to show how the models fit the toxico­ 36 logical data. 15 15 C 1 F 47 1 F 15 3 F 38 2 F 49 8 C A batch feature allows a large number of risk 24 1 F 39 2 F assessment approaches to be applied auto­ 26 1 C 40 1 F 30 matically to different sets of toxicological data. 1 F 41 4 F 31 1 F 42 1 F TOXRISK provides a summary of the calcula­ 34 1 F 44 3 F tions performed on a group of data sets. It can output results to a computer screen, a printer, or a computer file. determine if a tumor is present. These studies For example, Figure 1 presents data on the time Data entry is internally quality-controlled. are called time-to-tumor studies. and cause of death for rats that have been given Since the data are validated as they are en­ The data provided by this kind of study are toxic food. This information provides a dose­ tered, many common data entry mistakes are extensive, and their evaluation requires more response curve, which is then used to predict eliminated. Moreover, TOXRISK makes it easy to sophisticated models than the models in Ver­ risk to humans (Figure 2}. store and modify such animal and human pa­ sion 1 of TOXRISK. These more complex models Some of the most frequently used time-to­ rameters as body weight, breathing rate, and can be viewed as producing dose-response tumor models have been added to TOXRISK, life span and to apply them correctly in risk curves that are also a function of time. If the time along with capabilities for creating and manag­ calculations. It can also supply default values considered is a lifetime, then the usual risk as­ ing time-to-tumor data sets. These models en­ for these parameters. sessment can be performed with the models. able the risk manager to calculate risk-related TOXRISK employs several user-friendly tools, including pull-down windows and on-screen ------• TEST FOR TIME-TO-TUMOR MODELS ------• help. Also, the user can switch temporarily to a Dataset : C:\ TOXVER2\ TEST . TXS\ TIMEl DOS shell to perform other DOS applications Exposure Pattern without permanently exiting TOXRISK. Model: One Stage Weib Age Begins: 0 Age Ends : 70 Target Species : Human Weeks/Year: 52 Days/Week: 7 With all these features, Version 1 of TOXRISK Route : Food 100% Hours/Day: 24 has enabled utility managers to perform risk Animal to human conversion method: MG/KG BODY WEIGHT/DAY assessments efficiently on IBM-compatible personal computers. Risk assessment con­ Unit Potency [per mg/kg/day) (computed for Risk of 1.0E-6) cepts are constantly changing, however, as MLE = 4. 7933E-002 Upper Bound ( ql*) = 6. 5663E + 001 more sophisticated biological studies are per­ formed. Therefore, Version 2 of the program is Dose Estimates (ppb) to be released this winter. As well as enabling Extra Risk Time 95% Lower Bound MLE more complex assessments, it offers improved 1.0000E-006 70.00 7.6147E-004 1. 0431E + 000 user-friendliness and graphics capability. 1.0000E-005 70 .00 7.6147E-003 1.0431E +001 0.0001 70.00 7.6151E-002 1. 0432E + 002 Time-to-tumor studies

To xicologists have designed studies to deter­ Figure 2 This TOXRISK screen presents human risk estimates derived from sample animal data by using a linearized model. The potency of the chemical in question has been calculated, as well as mine the time after initial exposure at which a dose estimates corresponding to extra risks of 10·6 to 10·4 (i.e., one extra cancer in a population of tumor seems to occur. Animals are sacrificed 1,000,000 exposed individuals to one extra cancer in a population of 10,000). For each category of at specified times (serial sacrifice) if a tumor is extra risk, the 95% lower bound of the dose and the maximum likelihood dose estimate (MLE) not directly observable and are autopsied to are given.

EPRI JOURNAL December 1989 43 doses that are dependent on the period of able in TOXRISK Version 2 is a biologically reduced by more than 50K. exposure as well as on a specific exposure based model that can be used to show how a Version 3 of TOXRISK, expected to be re­ level. Such factors as the time from the appear­ chemical exposure affects the net growth of leased next spring, will include graphing of the ance of a tumor cell to the resulting tumor both precancerous and cancer cells. This time-to-tumor models. This new version will response can also be calculated. model is a form of the Moolgavkar model contain data for selected ch·emicals from the One of the time-to-tumor models included in based on an expression by Thorslund. National Toxicology Program CBDS (Carcin­ Version 2 of TOXRISK is the Armitage-Doll Other enhancements ogenesis Bioassay Data System), a database model, which allows the user to vary the experi­ that includes many of the lifetime studies on mental exposure level in the extrapolation pro­ The graphing procedures of TOXRISK Version 2 animals for a large series of substances. Also cess. In the course of an experimental study, are internal (unlike those of Version 1, which available will be a system for selecting studies for example, the toxicologists may revise the depend on Lotus graphics) and are much from this database to create a TOXRISK data exposure protocol to prevent severe damage faster and easier to use. Also new is the ability set. The utility manager will then have the to the animals or their early death, or to repre­ to graph lower bounds on doses associated ability, with very user-friendly software screens, sent more realistically the actual exposure pat­ with risk to a target species and computed by to choose a substance and evaluate the can­ tern that people experience. TOXRISK can now using dose-response models. The menus cer risk for a variety of data sets. Also planned derive risk estimates from such data. have been streamlined with many windows, is extensive help (including a tutorial) on the Another of the time-to-tumor models avail- and maximum RAM requirements have been use of risk assessment.

Demand-Side Planning Measuring Direct-Load-Control Impacts by Steven Braithwait, Customer Systems Division

irect-load-control (DLC) programs have for the effects of consumer behavior. This com­ alternative DLC strategies. The approach has D been adopted by a large number of util­ bined approach can be used both to measure been applied to DLC program data in three ities. A 1988 EPRI survey of residential demand­ the impact of an implemented load-control pro­ utility case studies. side management projects identified over 360 gram and to simulate, or predict, the impact of Duty cycle is a convenient means of repre- utilities with active load management pro- grams designed to cycle, or shed, the load of various residential appliances through direct utilitycontrol. These utilities reported almostl5 ABSTRACT In order to evaluate the cost-effectiveness of million points of control for air conditioners and close to 1.7 million points for water heaters. In demand-side management programs, utilities need reliable estimates addition, a number of other utilities are now assessing or testing new DLC programs. Reli­ of the programs' load and energy impacts. EPRI has recently developed able estimates of the load and energy impacts the duty-cycle approach, a method for evaluating and predicting the of these programs are crucial to optimal pro­ gram design and to the evaluation of the pro­ impact of one of the most common typ es of load management program, grams' cost-effectiveness. The duty-cycle approach residential direct load control. The duty-cycle approach, based on the

The duty-cycle approach combines an analysis of load data from metered appliances, has been applied to engineering-based approach that accounts for the way appliances operate under natural direct-load-controlprogr am data from three utilities. and utility-controlled conditions with an empiri­ cally based statistical approach that accounts

44 EPRI JOURNAL December 1989 senting energy use based on the standard Figure 1 Frequency distributions of air conditioner duty cycles for a half-hour period on a hot summer afternoon. A wide range of natural duty cycles is shown on days of no load control, with a kilowatthour measure. The average duty cycle relatively high proportion of air conditioners operating near full capacity (90-100% duty cycle). On of an appliance is the percentage of time that controlled days, a 50% cycling strategy (air conditioners cycled off for half of each half-hour period) the appliance is used during a measurement is expected to create a constrained distribution of duty cycles. The bars represent the percentage of households whose air conditioner duty cycles fall within each 10-percentage-point range. period. An air conditioner with an estimated connected load of 4 kW and measured energy 80 use over a particular half-hour period of 1 kWh, Noncontrol day for example, has an average duty cycle of 50% Control day (50% cycling) for that half-hour period (1 kWh/0.5 hour/ 4kW). The duty-cycle representation of energy use 60 is well suited to DLC program analysis because DLC programs achieve load and energy re­ l ductions by altering the natural, or uncon­ U) trolled, duty cycles of appliances. Implementa­ -cJ :g 40 Q) tion of a cycling, or shedding, strategy affects U) :::, 0 an appliance's duty cycle by limiting its oper­ I ation and scheduling it during the control period. A given cycling strategy, however-such as 20 one that cycles air conditioners off for 15 minutes every half hour- will reduce energy use only for those appliances whose natural duty cycles exceed the upper limit imposed by the strategy-in this case, 50% (15 minutes/30 10 20 30 40 50 60 70 80 minutes). In addition, the impact on appliances Duty Cycle (%) with natural duty cycles that do exceed the limit varies with the difference between the natural duty cycle and the upper limit. for control and noncontrol days because this mate the impact that would be likely to occur Because the impact of cycling is not the part of the duty-cycle distribution is not af­ under conditions different from those actually same across appliances, the duty-cycle ap­ fected by cycling. observed. proach focuses on the distribution of duty cy­ Estimation of appliance load Given the appliance-level data collected in cles across the population of appliances, impacts the three case-study utility programs exam­ rather than on the simple population average. ined in this project, the calculation of load­ The duty-cycle distribution of a population of Estimating the impact of a DLC program is control impact, facilitated by the use of the appliances is a frequency distribution of indi­ difficult because the actual load that would duty-cycle approach, can be accomplished in vidual appliance duty cycles for a specified have occurred in the absence of control is a number of ways. The most straightforward time interval. never observed. One standard approach to approach uses the empirical duty-cycle distri­ Shown in Figure 1, for example, is an actual estimating the load impact of a particular con­ bution that is calculated for particular days to uncontrolled duty-cycle distribution for Athens, trol action is to calculate the difference be­ simulate the likely impact of a given cycling Te nnessee, air conditioners for a particular tween the curtailed load on the day of control strategy. This approach requires information half-hour period, as well as the duty-cycle dis­ and the uncontrolled load on a comparison on only uncontrolled days in order to simulate tribution that would be expected to occur if day, typically characterized by similar weather the impact of alternative cycling strategies. 50% cycling were imposed (Le., if the air condi­ conditions. This comparison day approach of­ The load impact of a strategy is calculated tioners were turned off for 15 minutes each half ten uses aggregated load data at the system, by accounting for the different relative frequen­ hour). With cycling, all the air conditioners that or perhaps substation, level because data at cies associated with each duty-cycle range on would normally have duty cycles larger than the appliance level are usually not available. a control day and on a noncontrol day. In Fig­ 50% would be constrained to duty cycles of Finding days with truly comparable weather ure 1, for example, approximately 45% of the 50%. But the use of air conditioners with natu­ conditions is also difficult. Moreover, such ap­ appliances in the sample have natural duty ral duty cycles below 50% would be the same proaches do not provide the capability to esti- cycles between 90% and 100%. For each

EPRI JOURNAL December 1989 45 Figure 2 Shown are average estimated load reductions per household from air conditioner with natural duty cycles greater than 90%). The cycling at Athens Utilities, Atlantic Electric, and Florida Power. Load reductions, as expected, are average load reduction for the households af­ greater on hotter days and for more severe cycling strategies. fected by the cycling strategy (those with natu­ 1.0 ral duty cycles greater than 45%) was 12 kW. Thus, the overall average load reduction of 0.6 AUB § kW (Figure 2) reflects the fact that a number of AE FPC households' air conditioners are cycling at rates lower than 45%, even on very hot days, 'o and are unaffected by cycling. 0 .c a, en :::, Impact-prediction model 0 .c The ultimate goal of the duty-cycle research '; 0.5 is the development of an empirically based, 0 transferable model of DLC program impacts :::, 'O that can simulate potential load impacts under a, a wide range of cycling strategies, weather

0 conditions, and household characteristics. _J This model, called the impact-prediction model (1PM), will allow utilities considering DLC programs to predict the probable impacts of various control strategies by using local weather and customer data without having to 33% cycling 50% cycling 33% cycling 50% cycling collect end-use data or conduct a DLC test Hot Day Very Hot Day program. An initial application has resulted in the successful development of a weather­ based impact-prediction model, which is range of duty cycles, the implied duty-cycle and AE, the 50% cycling refers to 15 minutes described below. reduction derived from the cycling strategy cycled out of 30 minutes, while for FPC the The first step in the development of the and the relative frequency of that range is com­ cycling duration was actually 16.5 minutes, or weather-based 1PM was to represent duty­ puted. By combining this information across 55%.) The fact that the day-type definitions cycle distribution by a parameter-based distri­ ranges, it is possible to compute the average differ, particularly for the hot days, is part of the bution function. The parameters were then re­ duty-cycle reduction for the whole sample. In reason for the differences in estimated load lated to weather conditions. The beta-density the last step of the procedure, the duty-cycle reduction; however, the estimated reductions function was selected to develop parameters impact is converted to a load impact by multi­ are quite similar for the very hot day type, which for the duty-cycle distribution because of its plying by the average connected load. This represents approximately the hottest 15-20% flexibility and general conformance to typical procedure can be used to approximate the of summer days in each case. Also, the FPC duty-cycle distributions. The final step in the load impact of any cycling strategy results should be considered preliminary, not development of a transferable model will be to Figure 2 presents estimates of average air­ having been fully compared with previous util­ explain the two parameters of the beta distribu­ conditioner load reductions calculated by this ity assessments. tion (the mean and variance) in terms of local approach using data provided by the three The average load reductions shown in Fig­ weather conditions and customers' observ­ case-study utilities-the Athens (Tennessee) ure 2 depend upon a number of factors, only able demographic variables. Utilities Board (AUB), Atlantic Electric Co. (AE), some of which have been explored to date. To date, local weather variables have been and Florida Power Corp. (FPC). Load impacts The amount of load reduction contributed by successfully correlated with the beta-distribu­ are presented by day type and cycling strat­ various households and the number of house­ tion parameters in the AUB and FPC cases. The egy for the critical late afternoon period (5 30- holds falling into each group (Figure 1) are of weather-based impact-prediction model has 6:00 p.m. for AUB, 6:00-630 p.m. for AE, and particular interest. At FPC, for example, the shown itself to be an effective predictor both of 5:00-5:30 p.m. for FPC). For each of the two estimated load reduction for a 55% cycling duty-cycle distribution and of load impact. The temperature-based day types considered, strategy on the very hot day ranges from analyst supplies only the appropriate weather average load reductions resulting from 33% 0.4 kW (for households with natural duty cycles conditions, time of day, and average air con­ and 50% cycling are reported. (For AUB between 50% and 60%) to 2.1 kW (for those ditioner capacity in order to estimate the im-

46 EPRI JOURNAL December 1989 pact of any desired cycling strategy on air customer and dwelling characteristics in order able relationships between reported load im­ conditioner load. to account for differences among utility cus­ pacts and differences between utilities (EM- The load reductions predicted by the tomers, and to address other issues such as 4588 and EM-3861); and LOADSIM, a building­ weather-based 1PM closely track those calcu­ the possible increase in load, or payback, fol­ simulation model that permits calculation of lated from the empirical duty-cycle distribu­ lowing release of control. potential load-control impacts on· the basis of tions. Each of the impacts estimated by the EPRI perspective engineering principles (EM-3249). The empiri­ duty-cycle approach is well within the 95% cally based analysis of highly disaggregated, confidence interval of the IPM's impacts. This research complements some previous individually metered appliances will provide a Ongoing research includes analysis and EPRI work: periodic surveys of utility demand­ useful set of tools for utilities interested in mea­ validation of the weather-based 1PM for Atlantic side management programs (CU-6546), which suring the impact of current programs or as­ Electric and Florida Power. Further effort is include utility-reported estimates of average sessing the likely impact of a wide variety of needed to expand the methodology to include DLC impacts; an attempt to develop transfer- potential programs.

Market Assessment Planning and Forecasting Under Uncertainty by Peter Cleary and Ray Squitieri, Customer Systems Division

tility managers want analytical tools that is not the only criterion for judging forecasts. tools that provide the details needed to de­ U are accurate, easy to use, and useful in In particular, as utilities have expanded their velop integrated resource plans. making decisions. Unfortunately, what they get planning to incorporate a wide range of To better address the issues of uncertainty, sometimes fails to meet these standards. The demand-side management (DSM), marketing, in 1988 the Customer Systems Division began computer modeling community long believed and rate design options, they have recognized a project designed to help utilities treat uncer­ that bigger, more complex models were the the value of detailed end-use forecasting and tainty, communicate the results, and incorpo­ road to better planning. More time, more equa­ planning models, such as REEFS, COMMEND, rate them into the making of decisions. A scop­ tions, and more money, it seemed, would make INDEPTH, and HELM. EPRI has responded by ing study jointly funded by the division's Mar­ the new models so accurate that they would moving toward more user-friendly; PC-oriented ket Assessment Program and by EPRl's Utility supplant older methods. But large macroeconomic models failed en­ tirely to anticipate major changes in an increas­ ingly turbulent external environment. Moreover, ABSTRACT Although models used in planning and forecasting forecasting comparisons often conclude that complex methods perform no better than sim­ have become increasingly large and complex, their point-estimate ple methods. Complexity and detail do not forecasts reflect only one possible future scenario. Planners and fore­ always, it seems, lead to greater accuracy. Complexity also tends to make a model difficult casters need ameans of dealing with uncertainty. A recentEPRI scop­ to use and, more important, difficult for deci­ sion makers to understand. ing study has combined the Market Assessment Program's experience Deterministic, single-outcome forecasts ap­ pear to be less relevant in today's world of in developing powerful models to forecast electricity demand with the rapid and unexpected changes. Utility de­ cision makers now recognize the value of de­ UtilityPlanning Methods Center's efforts in applying a variety ofproba ­ veloping flexible plans on the basis of a range of outcomes rather than fine-tuning a single bilistic and scenario methods to uncertainties in supply-side issues. forecast. Utility planners also recognize that accuracy

EPRI JOURNAL December 1989 47 tainties. Scenarios have been particularly ef- veyed used a hybrid of two or even all three of Ta ble 1 fective for communicating results to decision the approaches. The best approach depends PHASE 2 OBJECTIVES AND POSSIBLE TASKS makers. In general, however, they provide no on the nature of the problem and the planning Objective Proposed Task quantitative means of evaluating different strat- environment. The following are important fac­ Simple screening tools with Develop simple spreadsheet egies; this often results in different strategies tors in selecting an approach: probabilistic component models with simulation O software corresponding to different future views of the Length of analysis. For example, a long-term world. Managers accustomed to quantitative forecast is required for evaluating capacity ex- Links to decisions Add decision capability to end-use models analysis may be uncomfortable with the sto- pansion decisions. 0 Conduct probabilistic rylike structure of scenario analysis. Level of aggregation. This depends on the approach case study The probabilistic approach uses techniques scope of the problem. The decision may be Scenario approach tailored Conduct scenario case of decision analysis to produce quantitative systemwide (capacity expansion) or narrower to utilities studies measures of possible outcomes. Results are (DSM programs, interruptible rates for indus­ Transfer of scenarios Conduct scenario most often presented through such measures trial customers). workshops of probability as the cumulative distribution o Level of detail. Does the decision maker Prepare guidelines on scenarios function; the analysis generally uses a com- need aggregate or detailed results? For exam- puter model to determine the consequences of pie, peak energy will be sufficient for some Transfer of probabilistic and Prepare guidelines and statistical approaches conduct workshops major uncertainties. forecasting problems; others will require hourly This approach works well for short-term or loads. end-use analyses. In these cases, the analyst 0 Sources of uncertainty. Uncertainty can Planning Methods Center set out to combine can identify key factors affecting the problem arise from changes in the business environ­ the UPMC's experience with uncertainty in ca­ and determine their possible values and prob­ ment, such as technical innovation, or from pacity planning and fuel contracts with Market abilities. The probabilistic approach may also random changes in the physical environment, Assessment's experience in marketing, fore­ work for longer-term problems and for ques­ such as weather. casting, and demand-side planning. tions of marketing, planning, and corporate 0 Degree of complexity. This depends on the The scoping study team ofApplied Decision strategy. It works best when the analysis can major sources of uncertainty. The number of Analysis and Barakat, Howard, and Chamber­ be tied directly to a decision. Because it pro­ uncertainties and their interdependence will lin identified three major approaches to ad­ vides explicit likelihoods for a range of possible determine which approach best suits the prob­ dressing uncertainty: scenario, probabilistic, results, the probabilistic approach makes it lem. and statistical. easy to compare different decisions. It does 0 Available data. The utility may have a large Comparison of methods not work well when the intended audience is number of data and/or many experts to draw not comfortable with quantitative analysis. on, or it may have limited informational re­ The scenario approach focuses on major The statistical approach uses quantitative sources. changes in the external environment. Sce­ techniques from statistics and econometrics. 0 Staff. The utility may or may not have staff narios are qualitative, internally consistent Results often appear in the form of variances with expertise in the appropriate method. descriptions of plausible states of the world. and confidence intervals around a forecast. 0 Objective. The utility may be trying to de­ Scenario results are most commonly pre­ This approach deals well with short-term, ran­ velop insights about what may happen in the sented in the form of stories. The company dom fluctuations, such as weather. It focuses future, or it may be making decisions directly. best known for its scenario work is Royal on uncertainty in the forecasting model param­ 0 Audience. The intended audience (man­ Dutch Shell. Utilities known for scenario work eters, which rely on historical data and do not agers, executives, or regulators) and their include Southern California Edison and Geor­ take into account shifts in the future utility envi­ backgrounds will affect the choice of method. gia Power. ronment. As a result, the statistical approach Stages of analysis The scenario approach deals well with plan­ fares poorly when uncertainties arise from ning and forecasting problems that involve long-term structural changes in the business The scoping study noted that each approach possible shifts in the external environment. environment. This approach is also difficult to has three phases: formulation, analysis, and Scenarios can combine many complex uncer­ apply when the problem involves many depen­ communication. Each phase has several tainties to produce independent but internally dent uncertainties. Moreover, it does not easily steps, though not every step appears in each consistent views of the future. Consequently, incorporate risks that must be determined by approach. In a preliminary step common to all this approach suits questions in which uncer­ expert judgment. the approaches, the analyst determines what tainties are numerous and interrelated. It can In practice, a utility is unlikely to use one type of forecast is required (e.g., short or long provide qualitative insights about key uncer- approach exclusively. Most companies sur- term), what model or models are available, and

48 EPRI JOURNAL December 1989 how the results will be used. 0 Step 4. Describe the uncertainty in the vari­ Each approach represents a selection of The study identified eight steps in the plan­ ables identified in the formulation phase. methods for the different steps. In practice, ning or forecasting process. The first three 0 Step 5. Associate values with each of the utilities often draw from all three approaches. make up the formulation phase: variables; a range of values represents differ­ The scoping study concluded by making 0 Step 1. Identify the key factors that are un­ ent conditions due to uncertainties. recommendations for future work.- Table 1 lists certain in the forecast and determine their rela­ 0 Step 6. Run the model (or reduced model) the needs identified and the corresponding tionships to generate outputs based on the values Phase 2 tasks to meet these needs. Phase 2 0 Step 2. Tie the factors to variables in the assigned in Step 5. work now in progress includes preparation of a planning or forecasting model. The final two steps make up the communi­ report on the three approaches, a videotape 0 Step 3. Develop a reduced model as a cation phase: on scenario planning, a prototype scenario proxy for the large forecasting model. 0 Step 7. Communicate the results. generation framework, and an analysis of the The next three steps make up the analysis 0 Step 8. Integrate results into the decision­ statistical uncertainty in a simple industrial phase: making process. end-use model.

Fuel Planning Coal QualityImpa ct Model

by Arun Mehta, Generation and Storage Division

oal composition and characteristics sig­ The result is the Coal Quality Impact Model that can be applied at various levels of detail, C nificantly affect nearly every facet of (COIM™), which features detailed equipment depending on the sophistication of the user power plant operation, including auxiliary performance models for all systems directly and the amount of input data available. power requirements, forced-outage rate, main­ affected by coal quality. Available from EPRI as tenance costs, net plant heat rate, and a unit's a commercial software product, COIM has CQIM capabilities ability to meet full load. Therefore, as-delivered been designed for ease of use ii\a variety of Unlike computer programs that rely on numer­ fuel costs alone cannot properly reflect the utility applications. It is a flexible evaluation tool ous general correlations, CQIM is able to pre- relative economics of alternative coal supplies. To select the most cost-effective coal supply, a utility must be able to predict the impacts alter- ABSTRACT EPRI has developed the Coal Quality Impact Model native coals would have on plant performance and overall power generation. (CQIM} to help coal procurement personnel reduce fuel costs, supple­ Available methods for analyzing coal quality impacts include large-scale test burns, pilot­ ment long-term contracts, investigate low-sulfur coals for reducing scale burns, and detailed computerized engi­ neering analyses. Full- or pilot-scale burns di­ sulfur dioxide emissions, and conduct evaluations of new coal supplies. rectly determine a unit's ability to accommodate CQIM a particular coal; they can be time-consuming helps utilities maximize the delivered heating value per dollar of and costly, however, especially if several coals fuel expenditure and minimize the operating costs associated with are being considered. Moreover, translating ob­ served performance into fuel-related costs can combustion. By using CQIM to select the most economical coal sup­ be extremely difficult. Therefore, a computer­ based analysis is often the most practical way plies on the basis of as-burned (rather than as-delivered) fuel costs, to evaluate alternative coals. In 1985 EPRI initiated the development of utilities can save millions of dollars. a comprehensive state-of-the-art computer program for predicting coal quality impacts.

EPRI JOURNAL December 1989 49 oped for each specific power plant system COIM contains a vast number of algorithms Ta ble 1 and, as applicable, each system component for supplementing data provided by the user. SAMPLE CQIM PLANT INPUT DATA COIM then translates the detailed equipment This default information is based on the data FOR SCREENING EVALUATION performance predictions into fuel-related entered by the user, the overall plant configura­ power generation costs, taking into account tion, the characteristics of the design coal, and General plant data Unit size (MW net) 500 net plant heat rate, unit capability, mainte­ established equipment design practices. Capacity factor 0.685 nance and availability, and waste disposal. COIM users can therefore selectively collect Auxiliary power requirements (MW) 40 Design net turbine heat rate (Btu/kWh) 8200 Changes in coal quality affect net plant heat and enter the most important input data in Excess air level (%) 20 rate through deviations in boiler efficiency, terms of their specific needs and experience steam attemperation levels, flue gas tempera­ and can rely on the model to supply default Major equipment data Steam generator tures, and auxiliary power requirements. COIM values for the remaining data. Of course, the Maximum heat input per plant area predicts boiler efficiency for a given coal on the more actual data used, the more accurate the ([million Btu/h]/ft2) 1.8 Maximum design gas exit basis of calculations of latent heat, sensible COIM predictions. temperature (°F) 1900 heat, and unburned-carbon losses. (These Ta ble 1 illustrates the relatively simple input 12 Tube spacing at furnace exit (in) losses are based on calculations of steam gen­ data requirements for conducting an initial Maximum design flue gas velocity (ft/s) 60 erator and air heater heat transfer perfor­ screening evaluation of coal quality. The results Fuel preparation and firing mance, which affects the flue gas temperature can then be used to determine if further more Number of exhauster mills 6 , Rated capacity of each mill (t/h) 60 and velocity profiles through the gas path.) To detailed input is necessary to more precisely Fans characterize changes in auxiliary power as a predict unit performance. As the level of input Number of forced-draft fans 2 Maximum flow per FD fan (acfm) 600,000 function of coal supply, COIM considers sys­ data detail is increased, the COIM evaluation Number of induced-draft fans 2 tem horsepower requirements as well as the becomes more sensitive to any unusual or lim­ Maximum flow per ID fan 950,000 Air heater frequency and level of use of components. iting unit-specific design features. Typical air heater leakage (%) 12 Coal supply to a plant is generally not uni­ COIM evaluates alternative coals by compar­ Corrected gas outlet temperature form in quality but varies within a specified (OF) 300 ing them with the current coal supply. Typical Precipitator range. COIM employs Monte Carlo derating coal properties for input to the model include 800,000 Total collecting plate area (ft2) analysis techniques to consider the impacts of ultimate and proximate analyses, ash compo­ Design volumetric gas flow (acfm) 2,000,000 Ash handling coal quality variation on system capability. sition, ash fusion temperatures, and Hardgrove Pressurized system for fly ash, Monte Carlo techniques are used to approxi­ grindability numbers. COIM can supplement capacity at normal operations (t/h) 50 Sluice system for bottom ash, mate the variation of quality within the coal incomplete sets of coal quality data on the capacity at normal operations (t/h) 20 supply's specified range in order to determine basis of similar coal supplies. derating magnitude and frequency as a func­ Cost results from COIM can be used to com­ Economic data Replacement energy cost tion of time. pare the economic viability of various coal sup­ (mills/kWh) 20 15 Coal quality affects equipment outage rates plies. In Figure 1 four coal supplies are com­ Limestone cost ($/t) and hence overall unit availability and mainte- pared on the basis of differential fuel-related nance costs. COIM predicts maintenance and generation costs as predicted by COIM. As the availability impacts through the use of a modi- figure shows, COIM provides a comprehensive lied UNIRAM model. This model uses historical basis for evaluating coals. diet equipment performance impacts for a data and correlations that relate com­ A notable COIM capability is sensitivity anal­ specific unit on the basis of the following input ponent-level failure mechanisms to coal qual­ ysis, which is useful in determining threshold data: ity. It uses UNIRAM methodology to translate limits for specific coal properties This type of ° Configuration data, including available predicted component failures into occur­ analysis can be very useful in establishing coal spares and system redundancy rences of system and plant derating. All fail­ procurement contract provisions for deviations 0 System and component capacities ures, regardless of their impact on availability, in delivered coal quality 0 Historical unit and system performance are considered in maintenance cost predic­ CQIM validation 0 System cycle times and auxiliary power re­ tions. quirements COIM predicts waste disposal costs for a In addition to COIM development, EPRI has 0 Historical maintenance and availability data variety of waste disposal configurations, in­ supported many activities to transfer the model 0 System and component failure mode data cluding landfill and ponding. For units with flue effectively to the utility industry and to ensure These data are used with COIM's equipment gas desulfurization, it also considers dewater­ that it remains a state-of-the-art tool. performance models, which have been devel- ing capability and FGD system performance One of the most significant efforts involves

50 EPRI JOURNAL December 1989 Figure 1 These CQIM economic results predict how the use of four alternative coals would o Establishing coal specifications and prop­ increase or decrease the generation costs associated with the current coal. The breakdown of erty range limits for specific generating units to cost components illustrates the comprehensive level of analysis possible with CQIM, which is sensitive to all significant factors in coal quality evaluation. minimize generation losses and operating problems 3 0 Quantifying the performance and economic impacts of variations in key coal properties, Coasom,bl� such as heating value, sulfur content, and ash 2 Maintenance :23 Waste disposal content

Ta ble 2 HOST UTILITY PLANTS FOR CQIM VALIDATION

Utility Unit Boiler Particulate Removal Device S02 Removal Device Centerior Energy Eastlake Unit 5 (660 MW) Supercritical Cold precipitator None East Kentucky Power Cooperative Spurlock Unit 2 (500 MW) Subcritical Hot precipitator Wet lime Houston Lighting & Power W A Parish Unit 8 (535 MW) Subcritical Fabric filter Wet limestone New England Electric System Brayton Point Unit 3 (575 MW) Supercritical Cold precipitator None Northern States Power Sherco Unit 1 (700 MW) Subcritical Particulate scrubber Wet limestone Pennsylvania Electric Conemaugh Units 1 and 2 (900 MW) Subcritical Hot precipitator None Public Service Company of Indiana Cayuga Units 1 and 2 (535 MW) Subcritical Cold precipitator None Southern Company Services Watson Unit 4 (250 MW) Subcritical Cold precipitator None Te nnessee Valley Authority Cumberland Unit 1 (1300 MW) Supercritical Cold precipitator None Union Electric Rush Island Unit 1 (575 MW) Subcritical Cold precipitator None Virginia Power Mount Storm Unit 1 (550 MW) Subcritical Cold precipitator None Wisconsin Power & Light Nelson Dewey Unit 2 (105 MW) Subcritical (cyclone) Hot precipitator None

EPRI JOURNAL December 1989 51 Equipment VAX systems; commercial versions features that promote ease of use and provide COIM's predictive capabilities. New correla­ of the program for IBM VM and MVS operating the user with greater flexibility in integrating tions for steam generator performance with systems are being finalized. Also, the EPSC has the model into utility-specific coal procurement respect to slagging, fouling, and the formation a simplified version of the program available for practices. of nitrogen oxides are planned as part of other the IBM PC. An expert system will be incorporated into EPRI programs. In addition to commercialization efforts, EPRI COIM to facilitate data input and aid in output Through improved understanding of how plans to further refine and enhance the useful­ interpretations. Also, the ability of CQIM to con­ coal properties affect unit performance and ness of COIM. In response to comments re­ sider full-scale, bench-scale, and pilot-scale fuel-related costs, a utility can potentially save ceived from the utility users group, COIM is data will be expanded. This will enable a utility millions of dollars by means of improved coal currently being converted to run on an IBM­ to use the model to address the economics acquisition practices and improved unit oper­ compatible 80386 desktop computer. This associated with a coal being test-burned. In ations. COIM will be an essential tool for devel­ program version will incorporate additional addition, such data can be used to improve oping that understanding.

Environmental Performance Assessment Coal Ash Use Sites by lshwar Murarka, Environment Division

lose to 80 million tons of fly ash, bottom Radian Corp. conducted field investigations to The field investigations at each site consisted C ash, slag, and scrubber sludge are pro­ measure changes in the chemical composition of two steps: (1) a reconnaissance effort involv­ duced annually by the electric utility industry. of groundwater, soil, surface water, and vegeta­ ing one-time, limited sampling, and (2) detailed Markets for 15-25% of these by-products of tion at two coal ash use sites: an embankment sampling over a network within a limited area. coal combustion have been developed. How­ site in Waukegan, Illinois, and a structural fill site The reconnaissance effort, conducted in ever, wider use of these residues is limited by in Little Canada, Minnesota. All field sampling, 1987, began with the installation of piezometers transportation costs, by the lack of a reliable sample analysis, and data analysis were com­ for estimating the hydraulic gradient and direc­ and quality-controlled supply (which certain pleted by November 1989. Detailed technical tion of groundwater flow. Then groundwater­ uses require), and by concerns that the resi­ reports on the two sites are being prepared. monitoring wells were installed upgradient, be- dues may contaminate groundwater, soil, and surface water. In a 1988 report to the Congress, the Envi- ronmental Protection Agency encouraged the ABSTRACT The wider use of coal combustion by-products for increased use of coal combustion wastes when that can be achieved in an environmen­ embankments, structural fills, roadbeds, and the like depends in part on tally acceptable manner. Even though coal ash has been used at some 300 sites in the United the development of information about their environmental performance States, however, measurements of the environ­ mental effects are scarce or nonexistent. To in these applications. EPRI is sponsoring field investigations to mea­ help meet the need for information, in 1987 sure long-term changes in soil, vegetation, groundwater, and surface EPRI initiated a study on the environmental ef­ fects of ash use (RP2796). This research, co­ water around coal ash use sites. Th e results of two recently completed, funded by Empire State Electric Energy Re­ search Corp., is aimed at measuring long-term limited field studies show localized changes due to the migration of ash changes in soil, groundwater, vegetation, and surface water quality around embankments, constituents. structural fills, roadbeds, and amended soils. In the first phase of the study (1987-1989),

52 EPRI JOURNAL December 1989 low the coal ash placement zone, and downgra­ Figure 1 Distribution of boron and molybdenum in broadleaf plants along the Waukegan sam­ dient along a transect in the groundwater flow pling transect. Plants growing directly on ash (locations 2 and 3) show the greatest uptake and accumulation in tissues. path. Samples of soil, groundwater, surface­ runoff water, and vegetation were collected and analyzed for water quality parameters and ele­ 90 mental composition. 80 The more detailed sampling effort was per­ formed in 1988. At each site 25 groundwater­ oi 70 monitoring wells, 12 porous-cup lysimeters, and C 60 2 pan-type lysimeters were installed beneath 0 the ash to monitor chemical changes in the C 50 saturated and unsaturated zones. Groundwater C 0 wells were screened at two depths. Samples 0 40 C were analyzed for a broad range of water quality 0 parameters (pH, Eh, alkalinity, alpha radioactiv­ £D 30 ity, and concentrations of chloride, fluoride, sul­ 20 fate, nitrate, phenol, arsenic, boron, barium, cadmium, cobalt, calcium, magnesium, chro­ 10 mium, copper, lithium, molybdenum, nickel, se­ 0 lenium, strontium, and zinc). 2.5

Waukegan site assessment oi The Waukegan site, located in northern Illinois a 2.0 few hundred meters west of Lake Michigan, has C nearly level topography, consisting of dune ·� 1.5 sands and back beach sediments deposited by the lake. A state highway segment and over­ 0 0 pass embankment were constructed in 1970- 1.0 1972 with approximately 350,000 cubic meters C

(500,000 metric tons) of coal ash from Common­ 0.5 wealth Edison's Waukegan generating station. The ash embankment is composed principally of ASTM Class F fly ash, although minor amounts 0 2 3 4 5 of bottom ash, clay, and other fill materials were Sample Location also used in the construction of the highway interchange. Several soil core samples were taken at the sites. Subsamples representing several depths of differences in rooting characteristics. Broad­ plant tissues. Figure 1, for example, shows con­ were analyzed in the laboratory to measure the leaf plants typically have taproot systems that centrations of boron and molybdenum in broad­ accumulation and/or deposition of ash constitu­ extend relatively deep into the soil. Grasses, on leaf plant tissues. Compared with samples from ents. In general, the data indicate that the sandy the other hand, have fibrous root systems that location 1 (the control location), samples from soils did not attenuate constituents leached are concentrated closer to the surface. Elemen­ locations 2 and 3 (directly on ash) had in­ from the ash. The underlying soils showed no tal concentrations in plant tissues were mea­ creased concentrations of these two elements. contamination by, or accumulation of, the ash sured for samples collected at several spatial The plant-tissue concentrations decreased at constituents. The reason is that sandy soils lack locations. Both accumulations and deficiencies locations 4 and 5, where plants grow on soil the necessary properties for adsorption, being of elements were found in plant tissues, but the overlying the coal ash deposit. The following very low in clay, cation exchange capacity, and plants had no apparent visual symptoms of ei­ elements showed increased concentrations in hydrous oxides of iron and manganese. ther type of change. tissues of plants growing directly on ash: Two types of vegetation, grasses and broad­ Chemical analysis identified the accumula­ 0 Broadleaf plants: boron, molybdenum, and leaf plants, were sampled to examine the effect tion of several elements of coal ash origin in nickel

EPRI JOURNAL December 1989 53 Figure 2 Distribution of phosphorus and zinc in grasses along the Waukegan sampling transect. magnitude higher in groundwater samples ob­ These results illustrate the reduction in availability and uptake of certain plant nutrients. tained beneath the ash. Other species from coal ash that appear to have leached into the shallow 2. groundwater zone include arsenic, barium, cal­ cium, chloride, iron, lithium, manganese, mag­ 2. nesium, potassium, sodium, and strontium. The concentration of these ash-derived constituents C 0 rapidly dissipates with distance away from the -� 1. embankment and at greater depths in the 0 C 0 groundwater below the ash.

(f) 1. Samples of soil-pore liquids obtained from lysimeters within the ash embankment showed

0 elevated concentrations of the same species 0 . that had elevated concentrations in the shallow groundwater beneath the ash. The concentra­ tions of certain ash-derived species (boron, cal­ cium, potassium, molybdenum, sulfate, stron­ 3 tium, and vanadium) were significantly higher in the soil-pore liquids, however. This contrasts 3 with the Little Canada ash fill site, where major ions were present in about equal concentrations 2 in soil-pore liquids and groundwater. The Wau­ C kegan ash embankment appears to have been 2 leached to a lesser extent than the Little Canada ·gC

C ash fill. This finding is not surprising, since the 0 paved road surface and steep embankment 0 C side slopes at the Waukegan site favor runoff and reduced infiltration into the ash. Little Canada site assessment

The Little Canada structural fill site is a level, 2 3 4 5 Sample Location currently undeveloped area adjacent to a small industrial park in a suburb north of St. Paul, Minnesota. To gether with the industrial park, the site was originally a low-lying, poorly drained 0 Grasses: aluminum, boron, molybdenum, 0 Grasses: magnesium, phosphorus, and zinc area; it was reclaimed by filling it with coal ash. and strontium At the Waukegan site, groundwater flows in The filling operation began in 1978 and was Chemical analysis also identified decreases an easterly direction beneath the embankment completed in 1980. About 39,000 cubic meters in concentrations of some elements in plant toward Lake Michigan. On the basis of mea­ (50,000 metric tons) of ash were used over the tissues. For example, phosphorus and zinc surements made in 1988, the groundwater flow approximately 1-hectare site. The ash came concentrations in grass tissues were lowest in velocity is estimated to be between 4.9 and 11.5 from three coal-fired power plants owned by samples from locations 2 and 3 (Figure 2). The meters per year. Data obtained during the four Northern States Power. The alkaline coal ash decreased concentrations in plant tissues are sampling dates show consistent trends over has cementitious properties and is classified as a direct result of the decreased availability of time for ash-derived constituents in ground­ a Class C fly ash according to the ASTM test these elements due to high pH in the soil-ash water and soil-pore water. method. mixture. These elements showed decreased Data for the wells that transect the embank­ In contrast to the Waukegan site studies, the concentrations in plants growing on ash: ment show that compared with samples from analysis of extracts of coal ash and soils from 0 Broadleaf plants: barium, calcium, magne­ upgradient wells, sulfate levels are two times the Little Canada fill site revealed that some sium, manganese, and zinc higher and boron levels are over an order of elements (cobalt, manganese, molybdenum,

54 EPRI JOURNAL December 1989 and pore water samples at the two sites over two Ta ble 1 sampling events. Concentrations of soluble GROUNDWATER DATA, LITTLE CANADA SITE (samples collected August 11, 1988) constituents released from the ash were mea­ sured in the shallow groundwater layer immedi­ Sampling Boron Sulfate Calcium ately beneath the ash zone (boron, calcium, Location* Concentration (mg/L) Concentration (mg/L) Concentration (mg/L) magnesium, molybdenum, sodium, chloride, Upgradient (average) 0.5 175 187 fluoride, sulfate, and possibly strontium); how­ Well B1 Shallow 5.8 1992 460 ever, none of the heavy metals present in the ash Deep 0.6 88 250 moved into the deeper soils, pore water, or Well B2 (shallow) 4.3 4432 500 groundwater. The results also show that the con­ Well B3 (shallow) 9.1 4844 290 centrations of ash-derived constituents that mi­ Well B4 grated into the shallow groundwater dissipate Shallow 4.3 667 390 Deep 0.6 4 150 rapidly with distance away from the ash Downgradient deposits. Shallow 0.6 21 230 The following conclusions can be drawn from Deep 0.6 74 240 the study results: 0 The release of highly soluble constituents "Distances from the upgrad1ent sampling location to the other locations are as follows: to well Bi, 60 meters; to well 82, 83 meters; to well B3, 108 meters; to well 84, 133 meters: to the downgradient location, 212 meters. from the ash occurs periodically when fluctuat­ ing groundwater encroaches on the ash. 0 Ash leachate constituents have not impacted the deeper groundwater zones below the ash strontium, and zinc) migrated from the coal ash groundwater zone. Further, the results indicate or the groundwater downgradient of the ash and were attenuated by the underlying soils. a very localized change in groundwater quality deposits. These attenuation trends are best seen in the due to the migration of water-soluble constitu­ 0 Limited attenuation of the leached constitu­ decreasing trace element (strontium and mo­ ents from the ash. ents occurs in the underlying soils. lybdenum) concentrations with depth. It is pos­ Groundwater data collected during this inves­ 0 Vegetation growing on the coal ash exhibits tulated that the trace elements coprecipitated tigation indicate that ash-derived constituents increased uptake of some constituents and de­ with manganese compounds as a response to are present in groundwater immediately be­ creased uptake of others. The reasons are (1) changing pH and redox conditions under­ neath the ash fill. However, it is not likely that changes in the concentration of elements avail­ neath the ash-soil contact zone. leaching of the ash in response to surface water able to plants as a function of the addition of ash Results from the two-year investigation show infiltration and downward percolation is the constituents to the soil, and (2) changes in soil consistent trends for ash-derived constituents ·1n cause. The lack of rainfall at the site (less than an pH induced by the ash additions, changes that the shallow groundwater and in soil-pore water. inch was recorded during the period of ground­ affect the solubility of some plant nutrients (both Compared with results for the control location water sampling in 1988) ensures very little infil­ ash-derived elements and elements native to and the deep groundwater zone, elevated tration. Field measurements show seasonal the soil). levels of potassium, sodium, calcium, magne­ fluctuation of the groundwater table, with at least On the basis of the limited studies at the sium, sulfate, chloride, boron, fluoride, molybde­ one monitoring well exhibiting a groundwater Waukegan and Little Canada sites, it is clear num, and strontium were measured in shallow­ elevation above the coal ash. Therefore, re­ thatthere is a need for a statistically sound data­ depth groundwater-monitoring wells installed in searchers hypothesize that the presence of ash­ base on environmental performance of waste the fill area. Table 1 shows selected results. In derived constituents in groundwater is due to use sites. General conclusions for all use prac­ soil-pore liquid samples obtained from lysime­ fluctuating groundwater levels in the zone near tices cannot be drawn from research at just ters beneath the ash, the same 10 species were the ash-soil contact. At normal rates of rainfall two sites. For this reason, field measurements found in approximately the same concentra­ and groundwater recharge in the site vicinity, the at a cross section of sites are planned for the tions as in the shallow groundwater samples. water table is expected to come in contact with future. In contrast, the groundwater wells installed in ash and allow leaching of ash constituents by In the second phase of studies under the deeper layer of the aquifer showed no direct dissolution reactions. RP2796, research will focus on the waste use change in the chemical composition of the wa­ Conclusions and future studies categories of embankments, roadbeds, and ter. These data provide a basis for concluding ash-amended soils. Nine sites will be selected that vertical mixing and depth of penetration Similar trends were observed for aqueous con­ to provide diverse environmental, geographic, of leachates are restricted to a very shallow centrations of elements in shallow groundwater and waste use combinations. These studies will

EPRI JOURNAL December 1989 55 measure several chemical constituents (inor- gases may accumulate. Hydrology and geo- formance at each site. Assessment manuals ganic, organic, and radioactive) in groundwater chemistry will be characterized to quantify the and data books will be prepared and published and surface water samples, as well as uptake distribution of mobile and attenuated constitu- to assist in defining parameters for the environ- and accumulation by nearby vegetation. Radon ents. Analysis of the field measurements will mentally acceptable use of coal combustion emissions will also be measured at sites where provide assessments of the environmental per- residues.

Diagnostics Gas Turbine Combustion Viewing System by Henry Schreiber, Generation and Storage Division

as turbines are a popular option for ca­ The image was scanned by a remote television the fuel nozzle and the upstream end of the G pacity increases; published estimates in­ camera. The probe looked into one of the 10 combustor. These are the types of flame anom­ dicate that 40,000-75,000 MW of new gas tur­ combustion chambers on the engine. This aly of interest to gas turbine operators, as well bine generating capacity will be added in the high-resolution probe produced an image as to engine designers. next 20 years. The large, stationary gas turbines composed of 80,000 pixels and allowed real­ The image quality obtained with this re­ offered by the major manufacturers in the 1980s time observation of combustor light-off, load search prototype probe proved highly suitable range in capacity from 78 to 150 MW in simple changes, and shutdown. During this evalua­ for gas turbine diagnostics in that it showed cycle and up to 220 MW in combined cycle. Net tion period, investigators used the probe to flame location and geometry in detail. The efficiencies are in the 32- 33.5% range for the observe combustor performance phenomena cost, however, was beyond the usual range for simple-cycle offerings and in the 44-49% range such as delayed ignition, crossfire tube oper­ an instrument that would be commercially mar­ for the combined cycles. Te chnical advances ation, flame instability, bimodal flame deploy­ ketable to utilities. EPRI therefore undertook a and improved design philosophy have in­ ment, and abnormal attachment of the flame to follow-on project with UTRC to develop a lower- creased availability and reliability to the high-80% to mid-90% range, while capital costs are in the $200- 250/kW range for simple cycle and the $400-600/kW range for a combined ABSTRACT Th e higher efficiency of moderngas turbines is due cycle. Lead times are relatively short compared with largely to the cooling and materials technology that has made it possi­ those for other generation options, and the avail­ able sizes of capacity-increase increments al­ ble to raise the temperature of the combustion gases flo wing through low the utility to add capacity parallel to load growth. The higher firing temperature required the turbine. To maintain high reliability and availability in the fa ce of to achieve the high efficiencies necessitates the use of new diagnostics. these more severe operating conditions, diagnostic instrumentation is Research prototypes needed that can provide early indication of a system malfunction, so

A proof-of-concept prototype of a combustion that corrective action can be taken before there are adverse physical viewing probe was built by United Te chnologies Research Corp. (UTRC) under EPRI contract effects on the equipment. The combustion viewing probe is one of (RP2102-2). This single-sensor probe was in­ stalled on unit 41 (a General Electric MS7001B several hot-gas-path diagnostic instruments developed by the EPRI combined-cycle unit) at the T. H Wharton plant of Houston Lighting & Power. The instrument Combustion Tu rbine Program intended to fulfill this function. consisted of a heat-resistant probe containing a 120° wide-angle lens and fiber-optic cable.

56 EPRI JOURNAL December 1989 cost probe using new fiber-optics technology Figure 1 Schematic diagram of one combustor and probe in a 10-probe array on a gas turbine unit that would still provide sufficient image defini­ at HL&P's Wharton plant. The probe looks into one of the 10 combustion chambers and picks up flame images, which are then displayed on a video monitor for real-time observation of combus­ tion and probe durability. This redesigned tor performance. combustion viewing system is intended for use as part of a permanently installed array. The TV camera array consists of one viewing probe mounted on each of the 10 combustors on a GE MS7001 Te n-image light guide engine. Gas turbine unit 33, part of a com­ termination bined-cycle unit at HL&P's Wharton plant, was chosen as the test engine for the full, 10-probe array. The probes are mounted so that they do not protrude into the combustor liner and therefore are not exposed directly to the flame. The probe Video monitor operating temperature remains close to the tem­ Fiber-optic image guide perature of the engine's compressor discharge air, which acts as a purge and coolant as it blows by the probe shaft into the combustor Probe Combustor casing dilution air hole, where the probe is mounted. Combustor liner Figure 1 is a schematic diagram of one combus­ tor and probe in the 10-probe arrangement.

Production prototype The 10-probe production-scale prototype has a 0.3-mm spherical lens optically coupled to a 0.3-mm, 2000-strand coherent fiber-optic ca­ ble, or image guide. The 2000-pixel image produced on a video screen provides sufficient detail to show flame instability, combustion de­ lay, crossfire tube activity, and general flame location. These diagnostic observations en­ able an operator to detect combustion anoma­ lies promptly. viewing system displays the flames in all 10 rough handling. Another anticipated design The 10-probe production prototype was in­ combustors simultaneously and in the same enhancement will make it possible to enlarge stalled on Wharton unit 41 and started up the relative position as the combustors on the en­ the image of any one flame and observe it as it week of May 29, 1989. The viewing system dem­ gine. This display is satisfactory for purposes changes in real time. With these design en­ onstrated its potential value immediately upon of operation and maintenance. The next phase hancements, the combustion viewing probe startup, when an ignition delay of over two of the project will be to develop a fully commer­ system will be ready to be offered to utilities minutes was observed in one of the 10 com­ cial prototype unit that incorporates system commercially. It will be particularly useful on bustors. Ignition delays produce thermal design modifications to further reduce the cost the new generation of higher-firing-tempera­ stress and allow unburned fuel to accumulate. and make the probes easier to fabricate on a ture gas turbines. EPRI will develop specifica­ This latter effect can result in a hazardous un­ commercial scale. The design of the probes tions that will enable the buyer of a new gas controlled pressurization of the exhaust duct­ will be made more rugged to make field main­ turbine to incorporate the combustion viewing work. tenance of the viewing system easier and re­ probe into the purchase specifications submit­ The current production-model combustion duce the risk of damage to the probes from ted to the gas turbine manufacturer.

EPRI JOURNAL December 1989 57 New Contracts

Funding/ Contra ctorI EPRI Funding/ Contractor/EPRI Project Duration Project Manager Project Duration Project Manager

Customer Systems Controls and Automation Te chnical Support $67,300 Automation Technology (RP2710-7) 12 months Inc.JS. Oivakaruni Power Electronics Systems Analysis $183,800 Georgia Tech Research Materials Testing tor CAES Well $59,700 Radian Corp./ B. Mehta (RP2951-9) 19 months Institute IM. Samotyj Applications (RP2894-6) 8 months Development of Advanced Off-Peak or $227,000 Vaughn Thermal Corp./ Base-Loaded Electric Water Heaters 9 months C. Hiller (RP2958-4) Nuclear Power Integrated Value-Based Planning $604,200 Putnam, Hayes & Bartlett, Service Water Assistance Program $108,200 Texas Utilities Electric Co. / (RP2982-4) 20 months Inc.JP Hanser (RP2495-13) 14 months N. Hirota Heat Recovery Chiller Field Demonstration $131,900 W. S. Fleming & New Yo rk-New England Seismic $50,000 U.S. Geological Survey/ (RP2983-8) 13 months Associates Inc. / M. Blatt Experiment (RP2556-53) 12 months J. Schneider Installation and Operation of Chiller Heat $75,700 Pacific Homes Inc./ Recovery System (RP2983-9) 13 months M. Blatt Source Parameters and Tectonics of Large $55,000 Memphis State University I Earthquakes in Stable Continental Regions 12 months J. Schneider Advanced Electric Transportation Systems $148,300 Bechtel Group, Inc./ (RP2556-54) Technology Assessment (RP3025-1) 9 months B. Banerjee Operator Reliability Experiments at the $150,700 Accident Prevention Group Assessment of Capacitor Technology and $56,700 University of Missouri at Maanshan Plant (RP2847-3) 11 months Inc. / A. Singh Applications in Power Electronics and 9 months Columbia/ B. Banerjee Control Systems (RP3088-1) Best-Estimate Methodology for Appendix K $310,400 Westinghouse Electric Relief (RP2956-1) 16 months Corp./S. Katra

Exploratory Research Management of Reactor Vessel Irradiation $69,300 Westinghouse Electric Embrittlement (RP2975-5) 8 months Corp./ T Griesbach New Chemical Alternative Refrigerants $155,900 Clemson University I Analysis of 10/23/88 Parkfield Earthquake $50,000 Woodward-Clyde (RP8000-48) 35 months P Joyner Data and Ground Motion Predictions for the 7 months Consultants I J. Schneider Computerized Corrosion Behavior $749,900 National Association of Characteristic Parkfield Earthquake Information tor Utility Applications 34 months Corrosion Engineers/ (RP2978-2) (RP8002-19) B. Syrett Severe Accident Studies (RP3000-34) $183,000 Mollerus Engineering A Comprehensive Investigation of the $220,000 University of Kentucky I 19 months Corp./ J. Haugh Inorganic Constituents of Coal and Coal 22 months S. Yunker Nonlinear Response to Strong Ground $149,600 Woodward-Clyde Derivatives (RP8003-20) Motion (RP3014-1) 3 months Consultants I J. Schneider Molecular Structure of Coal Macerals $250,000 Southern Illinois Thermal-Hydraulic Qualification and $551,700 S. Levy, Inc./ S. Katra (RP8003-21) 22 months University IS. Yunker Application of Modular Accident Analysis 21 months Characterization and Reconstruction $286,400 Arctech Inc./ S. Yunker Program (RP3044-1) of Coal-Gasifying Mixed Cultures 36 months (RP8003-23) Containment Analysis Methods (RP3048-1) $598,300 Numerical Applications 31 months Inc./M. Merila Critical Currents, Weak Links, and Flux $814,500 University of Wisconsin I Pinning Studies of Oxide Superconductors 36 months T Schneider Accident Sequences for Training $184,400 NUS Corp./A. Singh (RP8009-5) (RP3050-1) 20 months Fundamental Studies on Critical Current $936,600 Stanford University I LWR License Renewal Lead Plant -BWR $2,000,000 Northern States Power Density in Materials (RP8009-11) 60 months 0. Sharma (RP3075-1) 60 months Co. / R. Burke Development of Ceramic Superconductors $195,600 University at Buffalo Seismic Instrumentation of Mexican Power $79,000 PRISIS/ R. Kassawara for Electric Power Applications 13 months Foundation, Inc./ and Industrial Facilities (RP3090-3) 2 months (RP8009-12) T Rodenbaugh Radiation Protection /Health Physics $50,600 Analytical Resources Inc. / Improved Superconducting Ceramics for $159,000 U.S. Department of Exposure Characterization (RP3099-3) 8 months P Robinson High-Current Applications (RP8009-15) 15 months Energy I W Bakker Uncertainties in Neutron Kinetics $50,000 S. Levy, Inc./G. Srikantiah Calculations (RP3101-2) 7 months Generation and Storage Estimation of Strong Ground Motions Close $140,000 Woodward-Clyde to Earthquakes in Eastern North America 11 months Consultants/J. Schneider Market Evaluation for Fly Ash Fillers $63,100 Kline & Company Inc./ (RP3102-1) (RP2422-20) 6 months 0. Golden Simulator Fidelity Requirements (RP3107-1) $184,900 General Physics Corp./ Combustion Tu rbine Coatings Guidebook $172,100 Southwest Research 8 months R. Colley (RP2465-2) 17 months Institute/ R. Frischmuth Studies of Human Performance in $135,000 Anacapa Sciences Inc./ Mass Culture of Algae Utilizing CO2 From $509,200 University of Hawaii at Stalk Gases (RP2612-1 1) 18 months Manoa/J. Berning Inspection and Testing Tasks (RP3112-1) 12 months J. O'Brien Short-Rotation Woody Crop Trials for $80,000 Energy Performance Development of Utility Guidelines for $100,000 Quadrex Corp./ V Chexal Energy Production in North Central United 25 months Systems Inc./J. Berning Snubber Reduction Program for Nuclear 6 months States (RP2612-12) Piping System (RP3114-1) On-Line Monitoring for the Plant Auxiliary $432,700 Bechtel Power Corp./ BWR Stability Support (RP3114-7) $100,000 Computer Simulation and Power System (RP2626-3) 24 months J. Stein 11 months Analysis Inc./R. Galer Commercialization Potential of AFBC $193,500 Dearborn Chemical Void Fraction Model Improvement for CHEC $125,000 S. Levy, Inc./ V Chexal Concrete (RP2708-4) 9 months Company Ltd ./0. Golden (RP3114-18) 12 months

58 EPRI JOURNAL December 1989 Acid Precipitation and Human Health Design Guidelines for a Sodium EN-6492 Final Report (RP2661-22); $32.50 Injection Waste Management System New Contractor: Steve Hoffman GS-6486 Final Report (RP2708-1); $800 EPRI Project Manager: C. Hakkarinen Contractors Baker/TSA Inc.; !CF Te chnology Inc. Technical EPRI Project Manager: D. Golden Measuring Hydraulic Conductivity Reports With the Borehole Flowmeter Design and Commissioning of a Filter EN-6511 To pical Report (RP2485-5); $40 Module for High Te mperature and Pressure Contractor: Te nnessee Valley Authority GS-6489 Interim Report (RP1336-7); $25 Requests for copies of reports should be directed to EPRI Project Manager: I. Murarka Contractor: Rheinisch-Westfalische Te chnische Research Reports Center, PO Box 50490, Palo Alto, Hochschule, Aachen California 94303; (415) 965-4081 There is no charge EPRI Project Managers: 0. Tassicker, S. Drenker for reports requested by EPRI member utilities, U.S. universities, or government agencies. Others in the EXPLORATORY RESEARCH Use of Coal Gasification in United States, Mexico. and Canada pay the listed Compressed-Air Energy Storage Systems price. Overseas price is double the listed price. Research Planning Research Reports Center will send a catalog of EPRI Study of Fiber-Optic Sensors GS-6491 Final Report (RP2999-2); $32.50 Contractor: Energy Storage and Power Consultants reports on request. To order one-page summaries of ER-6428 Final Report (RP8004-1); $32.50 EPRI Project Manager: B. Louks reports, call the EPRI Hotline, (415) 855-2411. Contractor: Te nnessee Valley Authority EPRI Project Manager: N. Hirota Impacts of Cleaning Te xas Lignite on Boiler Performance and Economics CUSTOMER SYSTEMS Fundamental Chemical Structure of Coal GS-6517 Final Report (RP2425-1); $32.50 ER-6488 Final Report (RP8003-4); $25 Contractor: Combustion Engineering Inc. Moisture Measurements in Residential Contractor: Rockwell International EPRI Project Manager A. Mehta Attics Containing Radiant Barriers EPRI Project Manager: L. Atherton CU-6495 Final Report (RP2034-24); $100 Contractor Oak Ridge National Laboratory Proceedings: Fourth Symposium on Integrated Environmental Control EPRI Project Manager: J. Kesselring GENERAT ION AND STORAGE GS-6519 Proceedings; $62.50 EPRI Project Manager: J. Cichanowicz ELECTRICAL SYSTEMS Specifications for a Plant Electrical Systems and Equipment Workstation Leaning Brick Stack Liners GS-6430 Final Report (RP2960-1); $25 PCB Residues in Transformer Carcasses GS-6520 Final Report (RP1871-18); $32.50 Contractor: Sargent & Lundy Contractor: Battelle, Columbus Division EL-6237 Final Report (RP2028-19); $25 EPRI Project Manager: J. Stein EPRI Project Manager: P. Radcliffe Contractor: General Electric Co. EPRI Project Manager: G. Addis Use of Coal Ash in Highway Construction: Kansas Demonstration Project Lightning Flash Characteristics: 1987 NUCLEAR POWER GS-6460 Interim Report (RP2422-15); $32.50 EL-6413 Interim Report (RP243H); $25 Contractor: Kansas Electric Utilities Research VIPRE-01 : A Thermal-Hydraulic Code Contractor: State University of New York at Albany Program for Reactor Cores, Vols. 1-3 (Revision 3) EPRI Project Manager: J. Mitsche EPRI Project Manager: D. Golden NP-2511-CCM-A Computer Code Manual The Evaluation of PCDF Te chnical Assessment of the (RP1584-1); Vol. 1, $25; Vol. 2, $25; Vol. 3, $25 Analytical Methodologies Indirect-Fired Gas Tu rbine Power Plant Contractor: Battelle, Pacific Northwest Laboratories EPRI Project Manager: G. Srikantiah EL-6416 Final Report (RP2028-21); $25 GS-6470 Final Report (RP2387-3); $40 Contractor: Twin City Testing Corp. Contractor: Hydra-Co Enterprises Inc. Linear and Nonlinear Response of Structures EPRI Project Manager G. Addis EPRI Project Manager: A. Cohn and Equipment to California and Eastern United States Earthquakes Substation Evaluation Using Self-Calibrating Power Angle NP-5566 Final Report (RP2556-8); $32.50 Diagnostic Logic System (DIALOG) Instrument, Vols. 1 and 2 Contractors: Risk Engineering Inc.; Stanford EL-6429 Interim Report (RP2115-16); $32.50 GS-6475 Vols. 1 and 2 Final Report (RP2591-1); University Contractor: Andtek Inc. Vol. 1, $32.50; Vol. 2, $40 EPRI Project Manager: L. Mankoff EPRI Project Managers: A. Singh, J. Stepp, R. Contractor: Arizona Public Service Co. Kassawara EPRI Project Manager R. Nakata Decomposition Tec hniques for Below Regulatory Concern Owners Multiarea Generation and Proceedings: Second Biennial PFBC Group: Individual and Population Impacts Transmission Planning Under Uncertainty Power Plants Utility Conference From BRC Waste Treatment and Disposal EL-6484 Final Report (RP2940-1); $32.50 GS-6478 Proceedings; $77.50 NP-5680 Interim Report (RPB101-11); $1000 Contractor: Stanford University EPRI Project Manager: S. Drenker EPRI Project Managers: N. Balu, D. Curtice Contractor: Rogers and Associates Engineering Corp. Ash Utilization in Highways: EPRI Project Manager: P Robinson Delaware Demonstration Project ENVIRONMENT GS-6481 Interim Report (RP2422-3); $40 Demonstration of Reliability-Centered Contractor: Delmarva Power & Light Co. Maintenance, Vol. 2: First Annual Effects of Environmental and EPRI Project Manager D. Golden Progress Report From Experimental Design Factors on San Onofre Nuclear Generating Station Culturing and Te sting of Ceriodaphnia dubia Eighth Annual EPRI Conference NP-6152 Vol. 2 Interim Report (RP2970-3); $32.50 EN-6468 Final Report (RP2368-2); $40 on Coal Gasification Contractors: Southern California Edison Co.; ERIN Contractor: Battelle, Columbus Division GS-6485 Proceedings (RP1654); $55 Engineering and Research Inc. EPRI Project Manager: J. Mattice EPRI Project Manager: N. Holt EPRI Project Manager: J. Gaertner

EPRI JOURNAL December 1989 59 Demonstration of Reliability-Centered Maintenance, Vol. 3: First Annual Progress Report From Ginna Nuclear Station CALENDAR MAY NP-6152 Vol. 3 Interim Report (RP2970-1); $32.50 For additional information on the meetings 1-3 Contractors: Rochester Gas and Electric Corp., listed below, please contact the person 1st International Symposium on Advanced Te chnology Engineering Systems Inc. indicated. EPRI Project Manager J. Gaertner Biological Processing of Coal Orlando, Florida Contact: Stanley Yunker, (415) 855-2815 Matrix Database of DECEMBER EPRI Experimental Projects on 8-11 Small-Break Loss-of-Coolant Accidents 4-6 1990 S0 Control Symposium NP-6180 Final Report (RP2639-1); $32.50 Advanced Computer Technology 2 Contractor: lntermountain Tec hnologies Inc. for the Power Industry New Orleans, Louisiana EPRI Project Manager: J. Sursock Scottsdale, Arizona Contact: Paul Radcliffe, (415) 855-2720 Contact: Murthy Divakaruni, (415) 855-2409 Methods of Ultimate Load Analysis of Concrete Containments: Phase 3 5-6 NP-6262-M Final Report (RP2172-1); $25 Global Atmosphere and Public Health JUNE Contractor: ANATECH Research Corp. Washington, D.C. 11-13 EPRI Project Manager: H. Ta ng Contact: Ralph Perhac, (415) 855-2572 Applications of Power Production Simulation Proceedings: EPRI 7-8 Washington, D.C. Power Plant Valves Symposium II Industrial Load Shaping Contact: Mark Lauby, (415) 855-2304 NP/GS-6327 Proceedings (RP2519-1); $40 Palo Alto, California Contractor: Maintenance Equipment Application Contact: Paul Meagher, (415) 855-2420 Center EPRI Project Manager: B. Brooks JULY Sourcebook for Chemical JANUARY 1990 Decontamination of Nuclear Power Plants 29-August3 International Conference: NP-6433 Special Report; $32.50 17-18 Indoor Air Quality and Climate EPRI Project Managers: C. Wood, C. Spalaris Marketing Program Planning Toronto, Canada and Evaluation Contact: Cary Young, (415) 855-2724 Field Studies on lntergranular Attack Salt Lake City, Utah NP-6455 Final Report (RPS306-21); $32.50 Contact: To m Henneberger, (415) 855-2885 Contractor: Westinghouse Electric Corp. EPRI Project Manager: C. Welty AUGUST FEBRUARY Enriched Boric Acid for PWR Application: 28-30 Cost Evaluation Study for a Twi n-Unit PWR International Conference: Measuring 21-23 NP-6458 Final Report (RP2648-1); $5000 Waterborne TraceSubstances EPRI Food Service Symposium Contractor: Westinghouse Electric Corp. Baltimore, Maryland New Orleans, Louisiana EPRI Project Managers: J. Santucci, C. Spalaris Contact: Winston Chow, (415) 855-2868 Contact: Karl Johnson, (415) 855-2183 Relay Behavior at the Perry Nuclear Power Plant During the 1986 Earthquake in Leroy, Ohio SEPTEMBER NP-6472 Final Report (RP2849-2); $25 MARCH Contractors: NUTECH Engineers Inc., Stevenson 18-20 6-9 Associates Conference; Condenser Technology International Symposium: Performance, EPRI Project Manager: R. Kassawara Boston, Massachusetts Improvement, Retrofitting, and Repowering Fossil Fuel Power Plants Contact: John Tsou, (415) 855-2220 EPRI Workshop on Fire Protection in Nuclear Power Plants Washington, D.C. Contact: Gary Poe, (415) 855-8969 NP-6476 Final Report (RP2969); $40 EPRI Project Manager: J. Sursock OCTOBER 20-23 EPRI-EPA Symposium: Transfer and Evaluation of Burnup Utilization of Particulate Control Technology 15-17 Credit for Dry Storage Casks Incipient Failure Detection: San Diego, California NP-6494 Interim Report (RP2813-8); $25 Predictive Maintenance for the 1990s Contact: Ramsay Chang, (415) 855-2535 Contractor: Virginia Power Philadelphia, Pennsylvania EPRI Project Manager: 0. Ozer Contact: John Scheibe!, (415) 855-2850 21-23 2d National Conference and Exhibition Advanced PWR Steam Generator on Power Quality for End-Use Applications Digital Feedwater Control System Burlingame, California NP-6497-M Final Report (RP2126-4); $40 Contact: Marek Samotyj, (415) 855-2980 Contractor: Westinghouse Electric Corp. EPRI Project Manager: J. Naser

60 EPRI JOURNAL December 1989 Authors and Articles on the development of energy-efficient search. Rabi graduated in physics from HVAC equipment. Before 1978, he the Weizmann Institute of Science in worked for 12 years with General Dy­ Israel. She also has a PhD from Ohio namics on cryogenic systems and heat State. transfer analysis. Blatt graduated in Arvo Lannus has managed the Resi­ mechanical engineering from Cooper dential Program for more than seven Union, earned an MS in industrial engi­ years. He joined EPRI in 1980 after six Johnson Blatt neering at New York University, and years with Gordian Associates, where has a master's degree in business ad­ he was director of advanced technol­ ministration from San Diego State. • ogy, responsible for research on heat pumps and energy use in buildings. lectrification: Key to Manufactur­ Earlier he was on the chemical engi­ E ing Productivity (page 16) was neering faculty of Cooper Union Schurr Rabi written by Mary Wayne, communica­ School of Engineering. Lann us has a BS tions consultant, in cooperation with in chemistry and a PhD in chemical Sam Schurr, retired deputy director of engineering, both from Drexel Univer­ EPRI's Energy Study Center. sity. • Schurr was with EPRI for a total of 12 Lan nus Dau years, from 1973 to 1975 directing the he NOE Center: Celebrating a division originally responsible for both Decade of Achievement (page 32) environmental and economic studies, wasT written by David Boutacoff, Journal and from 1979 to 1988 working in en­ feature writer, who drew on informa­ ergy studies. He also worked with Re­ tion supplied by two staff members of sources for the Future on two occa­ EPRI's Nuclear Power Division. Lang sions, first as director of its energy and Gary Dau heads the Component Re­ minerals program and later as codirec­ liability Program and has been the In­ tor of energy policy research. Before stitute's principal liaison with the NOE 1954 he was with several federal agen­ Center since it opened in 1980. He initi­ ighting the Commercial World cies. Schurr has two degrees in eco­ ated an NOE program at EPRI in 1975 L (page 4) was written by Leslie La­ nomics from Rutgers. • while on loan from Battelle, Pacific marre, science writer, assisted by two Northwest Laboratories, and managed staff members of EPRI's Customer Sys­ ntegrated Home Automation (page it when he joined the EPRI staff in 1977. tems Division. 24) was written by John Miller, com­ He had been with Battelle since 1965. Karl Johnson manages research on municationsI consultant, aided by fea­ Dau has a BS in mechanical engineering energy systems for commercial build­ ture editor Ralph Whitaker and two re­ from the University of Idaho, a PhD in ings. He has been with EPRI since 1977, search managers of EPRI's Customer nuclear engineering from the Univer­ following two years as coordinator of Systems Division. sity of Arizona, and an MS in technol­ industrial and energy services for the Ve ronika Rabl is the manager of the ogy management from MIT. Palo Alto (California) utilities depart­ Demand-Side Planning Program estab­ James Lang, manager of the Plant ment. Still earlier, he worked for eight lished early this year. She had been a Operations and Maintenance Program years in energy management for Stan­ project manager for seven years, with since 1986, joined EPRI in 1979 to man­ ford University. Johnson graduated in primary responsibility for residential age research in the Steam Generator economics from Denison University and commercial load management re­ Project Office; he later worked in the and earned an MS in mechanical engi­ search. Before coming to EPRI in 1981, Nuclear Safety Analysis Center. He ear­ neering at Stanford. she worked for almost seven years at lier was with the Navy's Nuclear Power Morton Blatt, manager of the Com­ Argonne National Laboratory, much of Directorate for nine years, becoming mercial Program, came to EPRI in 1985 the time in technology assessment, in­ director of the reactor plant valve divi­ after seven years with Science Applica­ cluding a year on assignment to the sion. Lang graduated in physics from tions International, where he worked DOE Office of Energy Systems Re- Michigan State. • NONPROFIT ORGANIZAT ION Post Office Box 10412, Palo Alto, California 94303 U.S. POSTAGE PAID PERMIT NUMBER 99 REDWOOD CITY, CALIFORNIA ADDRESS CORRECTION REQUESTED