United States Environmental Protection Agency Air and Radiation (6204-J) EPA430/B-95-012 Jurse 1995 -4 '-> &EPA Conservation Verification Protocols Version 2.0 A Guidance Document for Electric Utilities Affected by the Acid Rain Program of the Clean Air Act Amendments of 1990 EPA 430 B 95 012 c.2 ACID RAIN PROGRAM ------- If you are interested in obtaining more information on the Acid Rain Program, or in receiving a copy of the User's Guide to the Conservation Verification Protocols. >: which includes a reporting form and instructions, call: ACID RAIN HOTLINE 202-233:9620 Monday through Friday, 9:00 a.m. to 5 p.m.. Eastern Standard lime :' or write to: i Acid Rain Division (6204J) U.S. EPA 401 M Street, S.W. Washington, D.C 20460 ------- ,-* s , r H) : j'. .'_>;,£ >i;> ;'^_i ^v:nctc;i,bc'' 20460 CONSERVATION VERIFICATION PROTOCOLS Version 2.0 V A Guidance Document for Electric Utilities Affected by the Acid Rain Program of the ' Clean Air Act Amendments of 1990 by < Barry D. Solomon* Alan K. Meier** &EPA ACID RAIN DIVISION U.S. ENVIRONMENTAL PROTECTION AGENCY * U.S. Environmental Protection Agency ** Lawrence Berkeley Laboratory ------- ------- ACKNOWLEDGMENTS *"i^^i-^« -v^»» The Conservation Verification Protocols (CVPs) were developed as a team effort! Instrumental was the support provided by Alan Meier and his staff at the Lawrence Berkeley Laboratory (L6L). Also instrumental was the Conservation Verification Subcommittee of EPA's Acid Rain Advisory Commit- tee, ably chaired by Stan Hulett, which provided valuable guidance, advice, and direction for the content of the CVPs. Membership of the Subcommittee included Tom Buckley, Cary Bullock, John Fox, Larry Frimennan, Bill Harding, Jeff Harris (of .the Northwest Power Planning Council), Liz Hicks, Eric Hirst, Stan Hulett, Marty Kushler, Steve Nadel, Bob San Martin, Vince Schueler, Sam Swanson, and Steve Wiel (now with LBL). Other people who have given important input to the CVPs include Joe Etc, Chuck Goldman, Jeff Harris, Barbara Litt, and Ed Vine of LBL, and Dan Blank, Marilyn Brown, Paul Centolella, Pat Quran, Bill Gavelis, Phil Hanser, John Hoffman, Marvin Horowitz, John Hughes, Phil Hummel, Ken Keating, Joe Kruger, Anne Gumerlock Lee, Brian McLean, Bill Miller, Rick Morgan, Steve Offutt, Gil Peach, Ralph Prahl, Renee Rico, Judy Tracy, Uoy d Wright, Roger Wright, and Cathy Zoi. ------- ------- Table of Contents Executive Summary.... .' 1 Purpose and Philosophy Behind the Approach 5 Section 1. Linkage With the Acid Rain Program 9 Background 9 Conservation and Renewable Energy Reserve (CRER) ;.. -9 Reduced Utilization (RU) Provision ; 10 Criteria for Deferring Conservation Verification to States 11 Verification by EPA 12 Forms and Data Requirements for Users : 12 First Full-Year Savings 13 Section 2. Verifying Energy Savings ..15 Introduction '.. , 15 Path I: Monitored Energy Use .'. .15 Reference Case ...» 15 Service Adjusted Energy Use and Savings ; 16 Gross and Net Savings Distinctions : 17 Confidence in Estimated Savings: Satisfying a 75% Hypothesis Test 17 Acceptable Measurement Procedures 1B Duration of Metering... 19 Path II: Stipulated Savings 19 Stipulated Savings for Specific Measures 19 Best-Practice Engineering Estimates : 20 Subsequent-Year Savings > 20 Option 1: Monitoring Subsequent-Year Savings 22 Option 2: Inspection for Presence and Operation of Subsequent-Year Savings... 22 Option 3: Default Subsequent-Year Savings 22 Supply-Side Efficiency Improvements 23 Appendix A Definitions ; 25 Appendix B ; Service-Adjusted Savings and Hypothesis Test 27 Appendix C t-Statistics for 75% Confidence One-Tailed Hypothesis Testing 31 Appendix D - Stipulated Savings, Procedures, Lifetimes, and Conversion Factors 33 Engineering Estimates of Energy Savings 33 Reduced Transmission and Distribution Losses Due to Demand-Side Efficiency Improvements : -35 ------- Table of Contents (Cont.) Energy-Efficient Refrigerators... Residential Water Heating Conservation Measures.. Ground Source Heat Pumps For Homes (New and Retrofit) Higher Efficiency Lights in Office Buildings , De-Lamping in Commercial Buildings Exit Sign Light Replacement Higher Efficiency Street Lights '. Higher Efficiency Motors for Constant Load Applications Amorphous Metal Distribution Transformers Stipulated Lifetimes for Common Conservation Measures Useful Lives of Residential Energy Conservation Measures Useful Lives of Commercial and Industrial Energy Conservation Measures. Stipulated Net-to-Gross Factors for Stipulated Measures .36 .37 .39 .40 .42 .43 .44 .45 .46 .47 .48 .49 .50 ------- Tables and Figures Figure 1. Overview of Verification Options r 2 Figure 2. Relationship Between EPA's CVPs and Acid Rain Program Allowances 6 Table 1. Qualifying Criteria and Applicability for Electric Utilities Using Energy. Efficiency Incentives Under Title IV of the CAAA 10 Table 2.. Bonus Allowances Awarded from the Conservation and Renewable Energy Reserve '. 13 Figure 3. Steps to Calculate Service-Adjusted Savings 16 Figure 4. Stipulated Savings Path '. 20 Figure 5. Subsequent-Year Verification Options 21 ------- ------- Executive Summary The Conservation Verification Protocols (CVPs) .have been prepared by EPA as part of its mis- sion to implement the Acid Rain Program of the dean Air Act Amendments of 1990. The Acid Rain Program has a goal of reducing SO2 emis- sions by 10 million tons from the 1980 level. Be- ginning in the year 2000, most electric utilities' that bum fossil fuels will have to hold enough SO2 emission allowances to cover their emissions. Energy efficiency programs can be an impor- tant component of utility acid rain compliance strategies because they can help meet customer demand for electricity with reduced emissions of SO2. In addition, the Acid Rain Program has two explicit conservation incentives, which are linked through opportunities to earn or save a new tradable commodity, the emission .allow- ance. The verification of energy efficiency sav- ings under these incentives is thus essential to the credibility of the market approach. Most in- vestor-owned utilities that qualify for these in- centives will probably have their energy savings verified by procedures specified by their State Public Utility Commission (PUC). The primary users of the CVPs under the Acid Rain Program are expected to be public power utilities. The CVPs are designed to be rigorous with- out being burdensome on the utility or the regu- lator. The CVPs help to ensure the cost-effective- ness of utility conservation programs and SO2 emission reduction measures, as well as the reli- ability of energy savings from those measures. Conservation Incentives in the Acid Rain Program The Conservation and Renewable Energy Re- serve (CRER) is a special pool of 300,000 emis- sion allowances that is available to utilities that meet electric demands with either conservation or renewable energy generation. To qualify for the CRER, a utility must have at least one affected unit, a least-cost plan, and net income neutral rated for conservation measures (to be certified by the U.S. Department of Energy). In addition, the utility can only receive credit for demand- side energy conservation measures. The Reduced Utilization provision applies to the 110 Phase I power plants that must lower their emissions in 1995-1999. In contrast to the CRER, this provision allows for supply-side efficiency measures and does not require least-cost plan- ning or net income neutrality. There is no pre- scribed limit to the number of allowances that these utilities can save through conservation. Verifying Energy Savings: The CVP Approach The CVPs allow for two general savings paths (see Figure 1): Monitored Energy Use or Stipu- lated Savings. The Monitored Energy Use Path is the preferred verification approach, and its goal is to measure energy use in such a way as to infer net energy savings, i.e.. the savings attrib- utable to the utility conservation program. The Stipulated Savings Path includes procedures for estimating savings as well as simple equations and standard values for estimating stipulated en- ergy savings from a limited number of corner-' vation measures for which expected energy sav- ings are well understood. This path also includes criteria for developing program-specific engi- neering estimates that may be used by a utility in limited cases. Finally, the CVPs also include guidelines for verifying the persistence of energy savings from conservation measures. Path I: Monitored Energy Use The preferred verification approach is to infer energy savings through the measurement of en- ergy use. The key features of this verification path include: Specifying a reference case. ------- Adjusting for differences in factors such as weather, operating hours, and production rates. , . Determining net energy savings. Establishing the appropriate statistical confi- dence in the savings. The reference case is the set of conditions and levels of service from which the energy savings are to be estimated. To develop the reference case, the energy use of a group not participating in the conservation program is often used. Raw energy savings, i.e.. simple differences in metered energy use, nearly always require ad- Figure 1. Overview of Verification Options Utility Conservation Programs Stipulated Savings Standard Equations and Values Monitored Energy Use Engineering Estimates Service-Adjusted Savings Gross Savings Net-to-Gross Discounting Net Savings FirsMfear Savings Subsequent-Year Savings justments to account for differences in metering periods, weather, indoor temperature, and lev- els of service. The CVPs require these adjust- ments for both the reference case and the cus- tomers participating in the conservation pro- gram. The CVPs require estimation of net energy savings, 1&, the savings attributable to the util- ity conservation program. A standard method for estimating net savings involves comparing the gross savings of customers participating in the program to the change in energy use of a simi- lar group not participating in the program (the "comparison group"), some of whom may have installed similar conservation measures on their own initiative. The comparison group will nor- mally be considered the reference case. If this approach is not followed, "the CVPs allow for other ways to develop a reference case and to derive net savings. For example, net savings may be estimated indirectly through market research, surveys, and inspections of non-participants. There will be uncertainty about the magni- tude of energy savings from utility conservation programs because of the many factors that can bias estimates of energy savings. An objective of the CVPs is to award allowances for savings that occurred with reasonable certainty. The CVPs require that the savings be expressed in terms of the utility's confidence, based on statis- tical analysis, that the true savings were equal to or greater than those for which it applied. Based on the state of the art of energy conservation evaluation, a 75% confidence (or 1-tailed hypoth- esis test) is required. Energy savings that have been adjusted for levels of service and that satisfy the 75% hypoth- esis test are considered "first-year savings" if the reference case captures net savings (e.g.. through use of a comparison group as discussed above). Subsequent-Year Savings A major uncertainty in utility conservation pro- grams is how long the energy savings persist over time. Three options are available for verifying savings in subsequent years: monitoring and inspection, inspection only, and a default. ------- If monitoring and inspection is used to verify persistence of savings, a utility must continue to meet the 75% hypothesis test used to estimate the first-year savings. The utility must measure energy use in the 1st and 3rd subsequent years. After the third subsequent year, the utility may continue to receive full credit for its savings for the lifetime of the measure. If the inspection only option is used, a util- ity can receive credit for at least 75% of the en- ergy savings in subsequent years,for up to half of the measure's lifetime. In the case of "passive conservation measures" (gsg., wall insulation), a utility may receive credit for 90% of the savings for the full lifetime of the measure. If the utility discontinues any form of moni- toring or inspection, the default option provides for 50% of the first-year savings to continue in subsequent years for half the lifetime of the en- ergy conservation measure. Path II: Stipulated Savings While the CVPs are oriented toward measure- ment of energy savings, in some cases utilities may use simple values or algorithms that have been provided for a limited number of technolo- gies. In other cases, they may develop their own engineering estimates. In such cases, a utility may convert gross savings into net savings through simple multiplication by a default fac- tor, which varies between 0.50 and 0.95, or by documenting an actual "net-to-gross" ratio. Sub- sequent year savings for this path are estimated by the same procedure used in the Monitored Energy Use Path (see previous discussion). The rationale for the Stipulated Savings ap- proach is that the performance of some measures. is well understood and may not be cost-effective to monitor. Use of stipulated savings in these limited circumstances should prove very attrac- tive to small utilities and ones with new conser- vation programs. The CVP contains 15 measures for which there are stipulated savings: energy? efficient refrigerators, pick-up of old refrigera- tors, ground source heat pumps for homes, of- fice building lighting upgrades, de-lamping in commercial buildings, exit sign lights, street lights, water heater insulation blankets, anti-con- vection valves, pipe insulation, low-flow showerheads, water-faucet aerators, heat pump water heaters, constant load motors, and amor- phous metal core transformers (a supply-side measure and thus ineligible for CRER allow- ances). Finally, Stipulated Energy Savings are also provided for a utility to take credit for trans- mission and distribution energy losses that are avoided by demand-side energy conservation measures. ' In certain circumstances in which extensive measurement and analysis are not cost-effective or feasible, an engineering estimate of energy savings will be acceptable. These circumstances include any of the following situations: Measurement costs would exceed 10% of pro- gram cost. No customer accounts for more than 20% of total savings. Energy savings are expected to be less than 5% of use of the smallest isolatable circuit (e.g.. residential lighting efficiency improvements). ------- ------- Purpose and Philosophy the Approach Purpose The U.S. Environmental Protection Agency (EPA) has prepared the Conservation Verification Pro- tocols (CVPs) as part of its mission to implement the Acid Rain Program authorized by Title IV of the Clean Air Act Amendments (CAAA) of 1990. Two cornerstones of the Acid Rain Program are SO2 emission allowance trading and energy effi- ciency improvements as part of compliance strat- egies of affected electric utilities. These corner- stones are linked through two explicit incentives for energy conservation that are described in more detail in the Acid Rain rules: the Conser- vation and Renewable Energy Reserve (CRER) discussed in the Allowance System Rule (40 CFR Part 73); and the Reduced Utilization (RU) pro- vision for Phase I affected units discussed in the Permit Rule (40 CFR Part 72). These two incen- tive programs allow electric utilities to earn or save allowances through qualified conservation measures. The main goal of tile CVPs is to credit elec- tric utility conservation programs for energy sav- ings that they generated as a part of an SO2 emis- sions reduction strategy. An ancillary goal is to describe consistent procedures for measuring and verifying these savings. The CVPs are intended to describe good practice for impact evaluators of electric utility energy conservation programs. They are not in- tended to be prescriptive. While the formal fed- eral verification requirements will end when the CRER and RU programs expire, users may want to continue to follow the CVPs or to use them in other applications such as State-level conserva- tion programs. A good verification program should be rig- orous without being burdensome on the utility or the regulator. The CVPs will have the benefit of helping to'ensure the cost-effectiveness of util- ity conservation programs and SO2 emission re- duction measures, as well as the reliability of actual energy savings from these programs. Eventually, energy savings should be as reliable an option for providing energy services to util- ity customers as energy production is currently. Incentive for Measurement The CVPs described herein are intended to pro- vide stronger motivation for monitoring energy savings from conservation measures and pro- grams. The CVPs embody an incentive struc- ture designed to encourage more and improved measurement and less reliance on estimation of energy savings, while allowing for simple veri- fication procedures through engineering esti- mates and stipulated savings (i.e.. standard al- gorithms for calculating energy savings from specific measures) when those approaches best fit the circumstances of the utility. The stipulated savings approach, while generally much less ex- pensive and time consuming to implement, is de- signed to be conservative and will often result in less energy savings credit than will a conven- tional verification procedure. As a result, utili- ties will more typically analyze, representative customer groups to determine the energy sav- ings. - The CVPs have been designed to create a single set of protocols with sufficient flexibility to accommodate Acid Rain Program provisions as well as future applications by other federal or State agencies. Thus, the CVPs are self-standing and independent of the individual regulatory programs. This necessitates language linking the individual allowance provisions and the CVPs. Section I specifies the administrative details unique to each regulatory program, such as tun- ------- Figure 2. Relationship Between EPA's CVPs and Acid Rain Program Allowances I EPA'sAcId Rain Program | State Energy Conservation Program EPA's Conservation Verification Protocols Revise List °f Stipulated Energy Savings Update Technical Review Committee ing, reporting and auditing, etc. Figure 2 illus- trates the relationship between the allowance schemes and the CVPs. ' The Need for Flexibility Energy conservation is a diverse activity. No general protocols for verifying energy conserva- tion savings can anticipate every kind of conser- . vation technology, program, or activity that will be undertaken by utilities. Procedures for veri- fying the energy savings must therefore be flex- ible enough to accommodate verification of the common conservation measures as well as new efficiency options. For example, in the past many utility programs have focused on commercial and residential buildings. Evaluation and verifica- tion techniques developed for the residential and commercial sectors, however, will not be directly transferable to the industrial sectors. For these reasons, the CVPs will give general guidelines for verifying energy savings rather man specify- ing the verification procedure for each kind of measure. The CVPs have been written to be accessible to all electric utilities: large and small, public and private. They can be used by utilities well versed in evaluation of energy conservation programs as well as by those who are new to this field. Cost-Effective Evaluation The cost of savings verification will depend on the kind of energy savings program and verifi- cation procedures used by the utility. If the an- ticipated extra cost of evaluation reverses the cost-effectiveness of the conservation program, a less stringent evaluation will be permitted. For purposes of discussion, it is generally expected that 5-10% of the program cost will be devoted to savings verification. No specific measurement technology is required; a utility can use what- ever approach will achieve the verification re- quirements at lowest cost. It is anticipated, how- ever, that most utility verification programs will rely on a mixture of utility billing data, inspec- tion of conservation measures, simple submetering, and stipulated savings (to be ex- plained below). In limited cases engineering es- timates of energy savings can be used by a util- ity. A rapidly increasing body of literature quan- tifying the field energy savings from various con- servation measures is emerging. In some cases, the savings are sufficiently reliable that a com- prehensive verification program may not be cost- effective. A list of these measures, with stipu- ------- lated energy savings or a stipulated energy sav- ings formula, has been developed by EPA and appears in Appendix D. While the list of measures with stipulated savings is not comprehensive, it will reduce the monitoring and verification burden and allow utilities to focus measurements on demand-side management (DSM) programs where impacts are less predictable. The provision for stipulation also offers some energy savings credit to conser- vation programs that were undertaken without adequate verification plans. The list and, the stipulated savings will be subject to periodic, pro- fessional reviews and revisions. This document uses considerable specialized vocabulary, some of which has not gained gen- eral acceptance. A list of definitions of the most important verification terms is presented in Ap- pendix A. ( Outlook for Verification The practice of verification of energy conserva- tion savings based on program evaluations is rap- idly evolving. While an increasing number of State governments are instituting Verification standards for their investor-owned utilities, even the leading States have been reassessing their protocols. Section 111 of the Energy Policy Act of 1992 will accelerate these activities, since it amends the Public Utility Regulatory Policies Act of 1978 and requires that integrated resource planning include the verification of energy sav- ings achieved through conservation. The Agency strongly believes that the CVPs need to remain flexible in order to allow utilities ample freedom. to evaluate their energy conservation savings in the most cost-effective manner., ------- ------- Section Linkage with the Acid Rain Program Background The purpose of this section is to provide back- ground and context for the CVPs by linking them with EFA's Acid Rain Permits and Allowance System rules, which were promulgated on Janu- ary 11,1993 (40 CFR §§ 72.43 and 72.91 and Part 73, SubpartF, respectively). Application require- ments and applicability criteria will be discussed, along with the differences between the Conser- vation and Renewable Energy Reserve (CRER) and the Reduced Utilization (RU) provision and who is responsible for verification of energy sav- ings. For a more complete discussion of these issues, refer to 40 CFR §§ 72.43 and 72,91 and Part 73, Subpart F. The overall goal of the Acid Rain Program is to achieve significant environmental benefits through reductions in emissions of SO2 and NOX, the primary causes of acid rain. To achieve this goal at the lowest cost to society, the Program employs both traditional and innovative, mar- ket-based approaches for controlling air pollu- tion. In addition, the Program encourages en- ergy efficiency and pollution prevention. Title IV of the CAAA sets as its primary goal the reduction of annual SO, emissions by 10 mil- lion tons below the 1980 level. To achieve these reductions, the law requires a two-phase tight- ening of the restrictions placed on fossil-fuel-fired power plants. Phase I begins in 1995 and affects 110 mostly coal-burning electric utility plants lo- cated in 21 eastern and midwestem States. Phase n, which begins in the year 2000, tightens the annual emissions limits imposed on these large, high-emitting plants and,sets restrictions on smaller, cleaner plants fired by coal, oil, and natu- ral gas. The Program affects existing plants with an output capacity of 25 megawatts or greater and new utility units under 25 megawatts that use fuel with a sulfur content greater than 0.05%. The Acid Rain Program creates a new trad- able commodity, the SO2 emission allowance. The verification of energy efficiency savings under the CRER and RU programs is essential to the credibility of the market approach. Once these programs end, EPA's requirements for continu- ous emissions monitoring of SO2 will provide automatic verification of emissions reduction (40 CFR Part 75). Conservation and Renewable Energy Reserve (CRER) The CRER is a special pool of 300,000 allowances taken from the Phase n allocations. The allow- ances in the pool are available to utilities that meet electric demands with either conservation or renewable energy resources. Congress estab- lished this Reserve to provide an early "jump start" to energy efficiency and renewable energy strategies for reducing SO, emissions. There are several criteria that an electric utility must meet in order to qualify for a share of the bonus al- lowances in the CRER, which are issued on a first- come, first-served basis (see Table 1): A utility must own or operate one or more af- fected power plant units or be the partial owner or operator of an affected unit (see 40 CFR § 73.82(a) and 40 CFR § 73.10 for a listing of these affected units and EPA's yearly base al- lowance allocations). Beginning July 1,1993, a utility with at least one Phase I affected unit may apply to the CRER for allowances for verified energy sav- ' ings (post-hoc) that occur from January 1,1992 to December 31,1994, or until all the allow- ances in the Reserve are allocated, whichever occurs first (40 CFR §§ 73.82(g)(l), 73.80(b)). Beginning July 1,1993, a utility with only a Phase n affected unit (or units) may apply to the CRER for allowances for verified energy savings (post-hoc) that occur from January 1, 9 ------- 1992 to December 31,1999, or until all the al- lowances in the Reserve are allocated, which- ever occurs first (40 CFR §§ 73.82(g)(l), 73.80(b)). 1 A utility must have a "least-cost plan" or plan- ning process for meeting future electric needs, which may consider social and environmental externality costs (40 CFR § 73.82(a)(4)). 1 Investor-owned utilities must be subject to a rate-making process that ensures net income neutrality, including a provision allowing the utility's net income to be compensated in full for lost sales attributable to its conservation program (40 CFR § 73.82(a)(9)). The U.S. De- partment of Energy began certifying the net in- come neutrality of State Public Utility Com- missions' (PUCs1) electric rate-making proce- dures on January 1,1993 (40 CFR § 73.82(g)(2)). Bonus allowances may be awarded condition- ally pending DOE certification (40 CFR §73.84(c)). Table 1. Qualifying Criteria and Applicability for Electric Utilities Using Energy Efficiency Incentives Under Title IV of the CAAA FEATURE Who May Apply? * Utilities Must Have a Least-Cost Plan and Net Income Neutrality? Type of Efficiency Measures That Qualify Emission Rate at Which Allowances Earned/Saved How Many Allowances Available? Timing of Program CRER BONUS ALLOWANCES Owner/Operator of Any Affected Unit Yes Demand-Side Measures Only 0.004 Ibs/kWh (0.4 Ibs/mmBtu) 300,000 Total 1992-94 (Phase I Utilities) 1992-99 (Phase II Utilities) REDUCED UTILIZATION Owner/Operator of Phase I Affected Units No Both Demand- and Supply- Side Measures Average Emission Rate of Phase II Units During Phase I Year No Limit 1995-1999 (Phase I Utilities) Utilities applying to the CRER must use quali- fied demand-side energy conservation mea- sures (listed in Appendix A(l) of 40 CFR Part 73, Subpart F) that were installed on or after January 1,1992, or measures that do not ap- pear on the list but that qualify because they meet the requirements of 40 CFR § 73.81. Qualified measures must be consistent with a utility's least-cost plan or least-cost planning process; must be funded in whole or in part by the utility; must increase customer end-use ef- ficiency; must not increase the use of any other 'fuels (other than renewables, industrial waste heat, or industrial waste gases); and must not be programs that are solely informational or educational (40 CFR §§ 73.81(a)(2)(i), 73.81(b), 73.82(a)(3)). A qualified electric utility will earn allowances from die CRER based on a formula that assumes an emissions rate of 0.4 pounds of SO2/mmBtu (or .004 pounds/kWh). Verified conservation savings of 500 MWh, for example, would earn one allowance. Reduced Utilization (RU) Provision In establishing the Acid Rain Program, Congress recognized that during Phase I utilities might cir- cumvent SO2 emission reductions by shifting load to units not regulated in Phase I. To avoid this potential problem, the rules implementing the CAAA require that during Phase I, a unit whose utilization is reduced below its 1985-87 baseline level must surrender allowances representing a shift of emissions to unregulated units, unless such under-utilization is accounted for in one of several ways. One way to account for under- utilization is through energy conservation. The RU conservation option may be especially valu- able to some utilities, since it allows them to ac- count for reduced load at Phase I units (and avoid surrendering allowances) by receiving credit for system-wide energy savings. Unlike the CRER program, under RU a Phase I affected unit would effectively receive credit for conservation at the 10 ------- SOj emission rate of the units to which load would otherwise have been shifted (primarily Phase U units). An affected utility could thereby potentially save allowances at a much greater rate than could be earned based on the f onnula- basedCRER. The criteria for a utility to use energy con- servation under the RU provision (40 CFR § 72.43) are less restrictive than the criteria for using the CRER (see table 1). The RU provision is applicable only to electric utilities with Phase I affected units that file RU plans with EPA pur- suant to § 72.43. Such utilities may use both de- mand-side and supply-side (isfi., power genera- tion, transmission, or distribution efficiency) measures that result in verified energy savings to reduce utilization of Phase I units below 1985- 1987 baseline levels. The supply-side conserva- tion measures are listed in Appendix A(2) of 40 CFR Part 73, Subpart F; supply- and demand- side measures installed as early as January 1,1988 may qualify, as long as the energy savings that occur during the RU period of 1995-1999 because of those measures are verified. \ferificationofenergysavingsisrequiredfor each year in which energy conservation measures contribute to the RU. The savings from any con- servation measures (other than improved unit ef- ficiency measures) instituted within a Phase I unit's utility system may be attributed to the Phase I unit.1 An RU plan must be filed with EPA by November 1 of the year in which RU oc- curs. An initial estimate of energy conservation savings must be filed in the annual compliance certification report by March 1 of the following year. The utility unit's designated representative must submit the verification results as part of a confirmation report by July 1 of each year for which an affected unit's annual compliance cer- tification report is submitted to EPA. Thus the first possible verifications would take place in the first 6 months of 1996 for energy savings that occur in 1995, following submittal of an RU plan by November 1995. The EPA Administrator may grant, for good cause shown, an extension of the time to file the confirmation report, however. If the energy savings do not result in utilization of Phase I units below baseline levels, such savings do not have to be verified. In contrast to the CRER program, a utility using the RU provision does not have to meet the least-cost planning or net income neutrality requirements, and since it is freeing up its own allowances, there is no .limit to the amount of conservation for which it may receive credit. Fi- nally, unlike the demand-side measures, which are installed at the customer side of the meter, the supply-side conservation measures at the power plant or in the transmission and distribu- tion system will result in energy savings that are all directly caused by the electric utility. Conse- quently, "net" energy savings from supply-side conservation measures can be assumed by util- ity applicants to equal "gross" energy savings, as discussed in the next section. Separate applica- tion forms for RU plans will be available at a fu- . hire date. Criteria for Deferring Conservation Verification to States The primary users of the CVPs are expected to be public-power utilities. Most investor-owned utilities in States that qualify for the CRER will probably have their energy conservation savings verified by procedures specified by the PUC, and EPA recognizes that it would be burdensome to require these utilities to use two separate verifi- cation procedures. Therefore EPA will defer to the conservation verification requirements of PUCs based on criteria discussed in the Allow- ance System Rule (40 CFR Part 73, Subpart F) and summarized below. EPA will require State PUC verification of energy savings claims of electric utilities if the FUC uses periodic evaluation of energy savings to determine any type of performance-based rate adjustment for conservation programs (40 CFR § 73.82(c)(l)). For RU conservation (40 CFR § 72.43), use of qualifying State PUC verification is optional. EPA believes that the verification performed by State regulators is likely to be fairly rigorous when the evaluation of energy savings ' If certain conditions are met, a unit that is part of a holding company system may be attributed savings from conservation within any part of the holding company system. See40CFR§.72.43(b)(2){iv). 11 ------- will be used to determine the cost of the conser- vation incentives to the ratepayers. In the case of both "shared savings mechanisms" and "lost revenue adjustment mechanisms," regulators must evaluate energy savings achieved by util- ity conservation programs in order to determine the proper adjustment to a utility's rates neces- sary to recover lost revenues or to provide a fi- nancial award. Under this approach for deferring conserva- tion verification to a State, the use of engineer- ing estimates of energy savings is allowed. The requirement to periodically modify performance' based rate adjustments through the evaluation of net energy savings, however, will help to en- sure that a State is committed to good evalua- tion practices. EPA recognizes that evaluation of energy savings is a young and rapidly evolv- ing field and that many States and utilities begin their evaluation programs with engineering es-. timates. Most evaluation experts agree that while engineering estimates are useful, they should be supplemented with statistically valid measure- ments of the decrease in energy use resulting from utility energy conservation programs (Ls^ net savings). These measurements should reflect both the actual number of measures installed and an ex post estimate of the energy savings achieved per measure. EPA believes that this approach toward con- servation verification and deferral to State PUCs is consistent with the approach of the Acid Rain Program, which encourages flexibility but insists upon appropriate measurement standards. While a minimum standard for deferring con- servation verification to States is being estab- lished, EPA is not prescribing standard verifica- tion methods or techniques. States have the flex- ibility to design their own verification protocols, as many have. Because of the diversity of evalu- ation methods and the ongoing evaluation ac- tivities of State agencies, electric utilities, and other professionals, EPA believes that this ap- proach will be more effective in ensuring good practice than would mandating specific methods or requiring universal use of the CVPs. Verification by EPA The CVPs are a guidance document for electric utilities, and they may be revised periodically. Moreover, as explained earlier in this report, many utility applicants to the CRER will not have to use the CVPs. Consequently, use ol the CVPs is not a requirement, though the reader should note that under EPA's Allowance System Rule applicants to the CRER must demonstrate to the satisfaction of the EPA Administrator that in the case of qualified energy conservation measures, energy savings have actually been achieved. EPA will consider the CVPs as a source of guidance for reasonable conservation verification proce- dures. The rule allows, but does not require, that applicants that do not qualify to have verifica- tion performed by their State public utility com- mission verify savings through use of the CVPs. Forms and Data Requirements for Users Users of the CVPs should summarize their veri- fication results on the CVPs reporting forms for the CRER and RU. These forms and instructions' are included in the User's Guide to the CVPs.2 Ap- plicants to the CRER should send the appropri- ate forms and any attachments (or a State PUC- approved conservation verification, where appli- cable), to EPA at the address provided at the end of the form instructions. Applicants should call the Acid Rain Hotline (202-233-9620) to receive a copy of the reporting forms. EPA began its review of the verified conser- vation savings claims when the CRER applica- tion period opened on July 1,1993, on a first- come, first-served basis. EPA reviews applica- tions to determine whether they meet the require- ments of Subpart F of 40 CFR Part 73. Varifica- tion claims may be based on the CVPs, a State PUC-sanctioned method, or other methods. If other methods are used, however, the time pe- 2U.S. Environmental Protection Agency, The User's Guide to the Conservation Verification Protocols, EPA 430-B-92-002, Washington, D.C., April 1993. 12 ------- riod required lor EPA's verification of an appli- cation is likely to be longer than if the CVPs or a State PUC-sanctioned method is used, because EPA would need to evaluate the alternative method as well as the application claim. Defi- cient verification claims will lose their place in line for bonus allowances if they must be sent back to the applicant for revision (see 40 CFR § 73.84(b)). Nonetheless, there is no deadline for sending in applications to the CRER (although the program has a limited duration), nor is there a prescribed reporting period for conservation results. The CRER allowances awarded to date are shown'in Table 2. Table 2. Bonus Allowances Awarded From the Conservation and Renewable Energy Reserve Utility No. of Allowances Centerior Energy Cleveland Electric 2 Toledo Edison 4 Cincinnati Gas & Electric 11 City of Austin (Texas) , 97 Dayton Power & Light 4 ESI Energy/Florida Power & Light 109* Long Island Lighting 535 Minnesota Power 8 New England Power Granite State Electric 24* Massachusetts Electric 436* Narragansett Electric ' 129* New York State Electric & Gas 142 Niagara Mohawk , --177 Northeast Utilities Connecticut Light & Power 173 Western Massachusetts Electric 30 Orange & Rockland - 46 Otter Tail Power 42* Portland General Electric 277 PSI Energy 41 Puget Sound Power & Light 1,002 Rochester Gas & Electric 7 Sierra Pacific 635* United Illuminating 47 Wisconsin Public Power 3 Total ' Some or all of total is for renewables. 4,161 A detailed evaluation of an energy conser- vation program requires a large amount of data and statistical analyses. In an attempt to mini- mize the reporting burden for verification, EPA is requesting applicants to submit general sum- mary data (as is indicated on the CVPs Report- ing Forms). The stipulated measures require fur- ther collection of information, which is not re- quired to be submitted to EPA but which should be maintained in a utility's files in case an audit is initiated by the Agency (40 CFR § 73.82(c)(3)). An audit of a conservation verification might be used to determine whether an incorrect number of allowances was disbursed to a utility or to collect more detailed information about the per- formance of energy conservation programs and SQj emissions offsets or reductions. Adequate verification programs by utilities should mini- mize the number of audits conducted by EPA. If such an audit determines that an applicant's en- ergy savings claims were overstated during the verification, EPA may reduce the allocation of al- lowances or require the surrender of allowances by the applicant (40 CFR § 73.82(c)(4)). First Full-Year Savings A special accounting problem may arise during the first calendar year in which a conservation measure is installed and operated. The energy savings during the first calendar year is typically for only a part of one year rather than a full year, since the conservation measures typically will not all be installed on January 1. This creates some confusion about the savings during the first year. For reporting to EPA, it is necessary to report the actual, partial year savings that coincide with the calendar year upon which the CRER allowance award will be based. (The procedure for calcu- lating the part-year savings is described in The User's Guide to the CVPs.3) After the program's first year, the'actual annual savings and calen- dar-year savings will be synchronized for the measures that were installed in the previous year and whose savings persist without attrition. The CVPs, however, use the concept of the first-year JU.S. Environmental Protection Agency, The User's Guide to the Conservation Verification Protocols, EPA430-B-92-002, Washington, D.C, April 1993. . 13 ------- savings for calculations of subsequent-year sav- ings. For example, the default subsequent-year verification option (see Figure 5) allows 50% of the first-year savings. The utility may thus need to calculate a first full-year savings to provide a basis for estimation of subsequent-year savings. The utility can have the option, however, of us- ing the full calendar-year savings (which may be the program's second year) as the basis for cal- culations of subsequent-year savings. 14 ------- Section Verifying Energy Savings Introduction As indicated earlier, the philosophy of the CVPs is to present general verification procedures rather than specific requirements. This approach offers utilities the maximum amount of flexibil- ity in developing verification methods. Never- theless, some consistency of method is needed in order to permit quantification of the energy savings and objective evaluation of energy sav- ings claims. Two general savings verification paths are described: Monitored Energy Use and Stipulated Savings. The two paths were illus- trated earlier in Figure 1 and are discussed in detail below. This section also discusses require- ments for determining the persistence of energy savings and procedures for quantifying the im- pact of supply-side efficiency improvements. Path I: Monitored Energy Use The preferred verification approach is to mea- sure energy use in such a way as to infer energy savings. The following sections describe the key steps in developing a method for converting en- ergy consumption measurements into energy savings. These steps are: \ * Specifying a reference case against which to measure savings. Adjusting for several factors, such as weather, indoor temperature, and industrial production levels. * Determining net energy savings, i.e.. the en- ergy savings resulting from the conservation program that would not have happened with- out the utility intervention. Establishing confidence in the energy savings. The discussion concludes with an overview of acceptable measurement procedures and the ap- propriate duration of metering. No particular measurement technique is prescribed, however. The intent of the CVPs is to suggest what may generally constitute good-practice verification procedures. The discussion begins with specifi- cation of the reference case from which to mea- sure the energy savings. Subsequent steps de- scribe the adjustments and other methodologi- cal issues. Reference Case The "reference case" is the set of conditions and levels of service from which the energy savings are to be estimated. The principal element of the reference case is measured energy use. By defi- nition, the reference case consists of the untreated (pre-retrofit) group's energy use adjusted to ser- vice levels from the treated group. This differs from many definitions of reference cases, where the untreated energy use and service levels es- tablish the reference case. In the CVPs, service levels from the treated case are preferred as the reference because service levels often change as a result of the retrofit and are more likely to pre- vail for the subsequent years. (A formulation of this definition is given in Appendix B.) In simple situations such as a conservation program in single-family homes, the reference case is the energy use of the pre-retrofit condi- tion or comparison group. The reference case also includes levels of service that can affect the level of energy use (and hence estimates of en- ergy savings) such as weather, operating hours, and industrial output. For an insulation program of electric-heated homes, the reference case might consist of the space heating energy use of the homes prior to retrofit, adjusted to the long-term average number of degree-days. More complex situations will rely on data from surveys or in- spections. For example, the reference case for a building lighting retrofit program might consist of a similar group of buildings unaffected by the program. In addition, the operating hours and illumination levels should be surveyed in order to adjust for yearly variation. 15 ------- The specification of the reference case will also determine if the utility is estimating gross or net energy savings. If the reference case cap- tures the impact of the conservation program (rather than only the technology), then the util- ity may follow the Net Savings path as shown in Figure 1. For example, if the reference case is a similar group of customers that did not partici- pate in the program, then the utility is estimat- ing net energy savings. Service-Adjusted Energy Use and Savings Raw energy savings, that is, simple differences in metered energy use, will nearly always require Figure 3. Steps to Calculate Service-Adjusted Savings Monitoring Path. ^ r energy use for treated and untreated cases (through monitoring) 1 Step 2. Specify reference case from which to measure energy savings (energy use and levels of service) Step 3. Calculate service-adjusted energy use tor treated case Step 4. Estimate Service-Adjusted Savings (SAS) satisfying hypothesis test at the 75% confidence level Does the reference case capture net savings' adjustments to account for differences in meter- ing periods, weather, and levels of service. These adjustments to derive the "service-adjusted en- ergy use" (SAEU) and "service-adjusted savings" (SAS) are described below, while the general ap- proach is shown in Figure 3. The major steps involve calculating the SAEUs for the reference .case and for the treated case. The difference be- tween the SAEU of the reference case and the SAEU of the treated case is the SAS. The energy consumption of many activities depends on the level of service required. For ex- ample, a house's space heating energy use de- pends on, among other factors, the outside and inside temperatures; a factory's energy use de- pends on the number of products manufactured in mat period; and a restaurant's energy use de- pends on the number of customers served. Some of the most important service factors are: Severity of heating or cooling season. Indoor temperature. Operating hours. Conditioned floor'area. Production rate (or mix). Illumination levels. A relationship between the level of service and energy use can usually be developed. For example, space heating energy use is a function of the heating degree-days, and lighting energy use depends on the number of operating hours and the lighted floor area. These relationships can often be derived through simple regressions of energy use and the level of service determined through measurements, surveys, and other data. For example, RCG/Hagler, Bailly, Inc.4 presents one formula quantifying the relationship be- tween energy use and various factors determin- ing the level of service, but other functions may be needed for industrial or special commercial applications. Although there is some uncertainty, these relationships may be used to normalize energy use in two periods with different levels of service. A step-by-step description of how to determine service-adjusted savings appears in Appendix B. 4RCG/Hagler, Bailly, Inc., Impact Evaluation of Demand-Side Management Programs, EPRICU-7179, Electric Power Research Institute, Palo Alto, CA, February 1991. ------- When estimating energy savings, energy use should be adjusted to the treated ("post-retrofit") levels of service In all cases except where there are weather variables or where a long-term av- erage differs from the measurement period. This practice accounts for the fact that energy-effi- ciency improvements are often made in build- ings at the same time that service levels are im- proved. . Gross and Net Savings Distinctions The intent of the CVFs is to estimate the net en- ergy savings from conservation programs. The utility may specify the reference case such that it captures the net savings. For example, the refer- ence case may be the energy use of an untreated group of similar customers. In this case, the net energy savings for the first year ("first-year sav- ings") emerge directly from the SAS calculation. The net effects can also be estimated indi- rectly through market research, surveys and in- spections of non-participants. Based on these methods, a utility can determine a "net-to-gross" conversion factor. (Multiplying the gross sav- ings by the "net-to-gross" conversion factor yields net savings.) A special case of net savings arises when the ' treated group of customers is effectively the en- tire population (leaving no untreated group available for comparison). This might occur when a utility targets the sole factory manufacturing a certain product. In this situation, the utility may still receive net savings credits by constructing a plausible scenario of what the customer would have done without intervention by the utility program. This scenario might be based on mar- ket research, surveys, economic models, or in- dexes of investment activity. The net-to-gross factor is usually less than one, g.g.. if a certain fraction of "free riders" ac- cept the utility incentives even though they in- tended to do the measures anyway, at the same time or in the future; program customers may also increase their energy use (the snapback ef- fect). In some circumstances, however, the net- to-gross factor may be greater than one. These situations may involve customers that implement the conservation measures as a consequence of the utility program, even though they receive no direct benefits. Another example is when pro- gram participants adopt additional energy effi- ciency measures. For example, a utility rebate program to encourage consumers to buy more efficient refrigerators might cause the manufac- turer to reduce shipments of ineligible units to . that utility's service area. Thus, the average effi- ciency of all refrigerators available to customers is increased, even if a customer chooses not to participate in the utility rebate program. This phenomenon (sometimes called "free drivership," "spillover," or "market transforma- tion") 5 may amplify the impact of a utility's re- bate program and may even raise the net-to-gross factor above one. Values greater than one can be used if the utility documents and quantifies the basis of the net savings estimate. See chapters in the report by Hirst and Reed6 (especially that by Saxonis) for a more detailed discussion. The documentation might include analyses of retailer inventory patterns before and after rebates were implemented, customer surveys, or other indi- cators of shifts in the market caused by the util- ity rebates. Confidence in Estimated Savings: Satisfying a 75% Hypothesis Test There is always uncertainty in the savings asso- ciated with energy conservation measures. Un- certainty occurs over the reliability of energy sav- ings estimates, the interaction effect of multiple conservation measures on total energy savings, and the persistence of energy savings over time.' For purposes of the emissions allowances, the objective is to award allowances for savings that occurred with reasonable certainty. The CVPs require that the savings be expressed in terms of the utility's confidence that the true savings were equal to, or greater than, those for which it ap- plied. In other words, in order to qualify for 100 allowances using the CVPs, a utility must be rea- sonably confident that the true savings equaled or exceeded 100 allowances. The utility must 5 Xenergy, final Report: Spillover Scoping Study, prepared by Xenergy, Inc. and Easton Consultants, Burlington, MA, January 1995. 'E. Hirst and J. Reed, eds.. Handbook of Evaluation of Utility DSM Programs, ORNL/CON-336, Oak Ridge National Laboratory, Oak Ridge/TN, 1991. 17 ------- demonstrate that the true savings are equal to or greater than the requested allowances with a 75% confidence. Sampling procedures to calculate the savings with the 75% confidence test are pre- sented in Appendix B, while the critical values for the t-statistic are shown in Appendix C.. The procedure described in Appendix B, while differing from that usually employed by electricity rate regulators (who typically require confidence plus or minus precision), offers utili- ties more flexibility, the ability to use smaller sample sizes, and the opportunity to claim some legitimate savings even when the evaluation it- self was not as successful as planned.7 The CVPs take this approach because while it is not based on usual statistical standards, it reflects the state- of-the-art for reasonable impact evaluation of savings from utility conservation programs. The Service-Adjusted Savings satisfying the 75% hy- pothesis are considered the "first-year savings" if the reference case captured the net effects (see Figure 4). If only gross savings were measured, then a further adjustment is needed (see above for a discussion of gross and net savings distinc- tions). j* ' Acceptable Measurement Procedures Energy savings must be derived from the differ- ence between a reference case and a treated con- dition. A variety of different methods are accept- able for measuring this difference. For whatever method is chosen, a utility may use a variety of analytical techniques, such as billing analysis, econometric modeling, submetering, or some combination of techniques. .Several common methods are: Measurement of a treated group's energy use pre-/post-installation of conservation mea- sures compared to a comparison group's pre-/post-installation. Measurement of a treated group's energy use pre-/post-installation of conservation mea- sures (Lp.. the treated group is its own com- parison group). Post-installation measurement of the treated group's energy use compared .to the untreated group's energy use fiue.. cross-sectional analy- sis). Flip-flop of measures (old device is still in place and can be switched on for a short period). The conventional verification procedure for a retrofit project involves measuring energy use before and after the conservation retrofit. In gen- eral, however, if this procedure does not include analysis of a comparison group, it yields gross energy savings (rather than net) because it does not control for what would have happened to energy use if the retrofit had not occurred. Post- treated versus post-untreated groups might be used to estimate savings from a conservation pro- gram applied to new building construction. This procedure could also supplement a before/after measurement to provide net energy savings es- timates. Alternatively, building simulations of energy use are often used for these conservation programs (see Appendix D discussion of Engi- neering Estimates). A flip-flop procedure allows for alternating measurements of the original equipment arid the replacement unit. A good candidate for flip-flop measurement is replacement of a resistance wa- ter heater with a heat pump water heater. A heat pump water heater can be easily switched from heat pump to resistance heating mode at weekly or monthly intervals until an adequate monitor- ing period has been reached. Certain energy management and control systems (from simple timeswitches to sophisticated Energy Manage- ment and Control Systems) can also be tested with a flip-flop procedure, that is, with the con- trols disabled. External controls would periodi- cally switch from one system to another, thus permitting estimates of energy savings during only one heating season. A flip-flop procedure is generally best applied when only one or two technologies are involved. Combinations of these procedures are also acceptable (and may even be necessary for esti- mating net energy savings). For example, a net ' For a comparison of this procedure with other determinations of statistical confidence, see P. Hanser and D. Violette, "DSM Program Evaluation Precision: What Can You Expect? What Do You Want?" Proceedings, Fourth National Confirmee on Integrated Resource Planning. National Association of Regulatory Utility Commissioners, Washington, D.C., September 13,1992, pp. 299-313. ------- savings analysis of a retrofit program will require measurement of the energy use of participants before and after the treatment, in addition to the energy use of a group of non-participants. There are many different methods to com- bine monitored data and survey data to estimate energy savings from utility efficiency programs. One popular method is the use of "Statistically Adjusted Engineering estimates (SAE)." The principles (and some of the variants) are de- scribed by RCG/Hagler, Bailly, Inc8 The SAE (and related, hybrid methods) is an acceptable method for estimating savings as long as the analysis includes energy use in the pre- and post- retrofit periods and reflects the reference service levels. Duration of Metering Metering is necessary to determine both the first- year energy savings and the extent to which those savings persist over the measure's lifetime Very different considerations apply to these periods. First-year energy savings calculations must be based on a sufficiently long period to cover any natural cycle of activities that influences en- ergy savings. For example, the "natural cycle" for a space heating retrofit is one winter. Briefer metering periods are acceptable under the CVPs, however, if they can be reasonably extrapolated to the whole year. For conservation measures applied to constant loads (or those automatically controlled), instantaneous power determinations are satisfactory. Once all of these procedures are completed by the utility, the result is the calculation of first- year energy savings. Path II: Stipulated Savings The CVPs are oriented towards the estimation of savings based on measurement of energy use. In certain cases, however, the utility may use the simple Stipulated Savings methods given in Ap- pendix D to earn credit for the applicable con- servation measures instead of relying on a moni- toring program. These Stipulated Savings meth- ods are of two different types: (1) algorithms for calculating energy savings for specific measures; and (2) a set of criteria for using best-engineer- ing practices. Stipulated Savings for Specific Measures The rationale for the list is mat the performance of some conservation measures is well under- stood and may not be cost-effective to monitor. Use of Stipulated Savings in these limited cir- cumstances should prove especially attractive to small utilities and ones with new conservation programs (although the energy savings assump- . tions are often conservative). The Stipulated Sav- ings list contains the following kinds of measures: Common measures whose savings are well- documented in the literature (e.g.. energy-effi- cient refrigerators). Measures where instantaneous metering is an acceptable proxy for first-year savings (g,g.. exit sign lights). Measures where short-term metering is an ac- ceptable proxy for first-year savings (e.g.. transformers). In some cases the laboratory test, "label," or other pre-specified efficiency parameter may be a reli- able indicator of energy use or savings. For stipu- lated measures, the utility may calculate energy use and savings based on these values rather than through metering. For example, in a refrigera- tor rebate program, the utility could calculate en- ergy savings by calculating the differences be- tween the maximum energy use dictated by the federal standard and the labeled consumptions of the refrigerators purchased in its high-effi- ciency program. This verification procedure would be much less costly than metering because the utility would need only to tabulate labeled values and count installations. 8 RCG/Hagler, Bailly, Inc., Impact Evaluation of Demand-Side Management Programs, EPRICU-7179, Electric Power Research Institute, Palo Alto, CA, February 1991. 19 ------- Figure 4. Stipulated Savings Path Use stipulated fii utility-researched net-to-gross factor List of stipulated net-to-gross factors (Appendix 0} ed tors ) Utility-researched 1 net-to-cross factor 1 (subject to approval)] First-year savings calculated [ Subsequent-year sayings calculated 1 The approach for calculations of stipulated energy savings is outlined in Figure 4. The cal- culations in the stipulated measures result in gross energy savings. The utility may convert the gross savings into net savings through simple multiplication by a default factor (the "net-to- gross conversion factor") given in the list of stipu- lated net-to-gross factors in Appendix D; alter- natively it can use its own adjustment factor based on its own research. The stipulated net- to-gross factors range from 0.50 - 0.95 and are based on experience with utility conservation programs. It should be noted, however, that net energy savings estimates for similar conserva- tion programs can vary widely, and if a utility develops its own estimate through internal re- search it may be able to claim larger net savings than if it uses the stipulated value. The utility may use its own estimate if it can demonstrate the legitimacy of its analysis to EPA: Subsequent- year savings for stipulated measures are esti- mated using the same procedure as is used in the Monitored Energy Use path (see the discus- sion earlier in this section). Best-Practice Engineering Estimates In certain circumstances, measurement of energy savings is not feasible or practical. These circum- stances include any of the following situations: < No single customer accounts for more than 20% of total savings. * There is an extremely small anticipated energy savings per customer, spread over many cus- tomers. (An example might be a residential program involving distribution of compact fluorescents.) Such savings may be difficult to observe in a billing analysis or submetering. The cost of monitoring and verifying energy savings is high. A utility may still want to ap- ply for verified energy savings even though verification through monitoring would cost significantly more than 10% of the total pro- gram costs. Determination of energy savings through energy monitoring is the preferred means of verification. When this is not feasible, in limited circumstances engineering estimates of savings may be substi- tuted, with an adjustment factor to account for the actual realization rate of projected energy sav- ings. The precise conditions in which engineer- ing estimates can be used are specified in the list of Stipulated Savings in Appendix D. Subsequent-Year Savings The persistence of energy conservation savings over time is uncertain. This uncertainty stems from both the eventual deterioration in a conser- vation measure's technical performance and be- ------- havioral factors such as the energy user's shift- ing preferences and maintenance practices. The limited research that has been conducted on per- sistence of energy savings indicates that conser- vation programs often overestimate measure per- sistence, particularly in the commercial sector and for residential compact fluorescent lamps.9 Three options are available for verifying sub- sequent-year energy savings: monitoring, inspec- tion, and a default. Figure 5 illustrates these options, which are applicable to both the Stipu- lated Savings and the Monitored Energy Use paths. Various discounting factors have been in- serted to make monitoring attractive. By moni- toring, a utility can obtain credit for a greater frac- tion of the savings and for a longer period. In contrast, the default option greatly restricts the allowable savings. This reflects decreased cer- tainty that savings persist when they are not monitored by the utility. For all three options, the net-to-gross conver- sion factor is calculated once for the first-year savings. The same conversion factor is used in all subsequent years to convert estimates of gross savings. For example, if a utility through exami- nation of a comparison group (or other accepted technique) determines that 15% of the treated group would have implemented the measure even without a utility incentive (a net-to-gross conversion factor of 0.85), then this factor car- ries to all subsequent-year savings. Thus it is not necessary to follow the comparison group be- yond the first year. All three options make certain assumptions about the lifetimes of specific conservation mea- . sures. The lifetime of a conservation measure is the median length of time (in years) that a mea- sure produces energy savings (see Appendix D). Stipulated lifetimes are provided for the most common conservation measures. When a stipu- lated lifetime is not given, the utility must esti- mate a. lifetime. If significant degradation in per- formance occurs, then the effective lifetime is Figures. Subsequent-Year Verification Options First-year I savings | ring on for eand (ton jit 100% of observed hypothesis test / measure >^ fr/ raipJmadfvo \ Biennial verification In subsequent years 1 and 3 (including inspection) ' Savings for remainder of physical lifetime are average of last two measurements No '"\. JSrtS^ V^ C"8881** XraHwroa^r measure) | Y|S (Active measure) ' ^ 50% of first-year savtnos physical libdiiM 90% of first-year IfBSme 75% of ftstyear savings fof units present and operating tor rtaS of physfcaJBeame (bJenNaltapBcbom) tor hall ot I _\ shortened. Ameasure may be physically present for many years beyond the time it is saving en- ergy at its designed level. Utilities may shift from one verification path to another in subsequent years. It is expected that these shifts will be from the monitoring op- tions to less extensive verification procedures. These shifts are permissible; once a shift is un- 9See, £,&, E. Hirst and I. Reed, eds., Handbook of Evaluation of Utility DSM Programs, ORNL/CON-336, Oak Ridge National Laboratory, Oak Ridge, IN, 1991; K. M. Keating, "Persistence of Energy Savings," in E. Hirst and ]. Reed, eds.. Handbook of EvtAvaiion of Utility DSM Programs, ORNL/CON-336, Oak Ridge National Laboratory, Oak Ridge, TN, 1991, pp. 8949; Northeast Utilities, A Survey of Other Utilities for Measure-Life and Persistence-of-Savings Issues, Conservation and Load Manage- ment Department, Northeast Utilities, Hartford, CT, March 1992; L. A. Skumatz, et al. "Bonneville Measure Life Study: Effect of Commercial Building Changes on Energy Using Equipment," SRC No. 7619-R2, Synergic Resources Corporation, Seattle, WA, December 1991; E. LJ Vine, "Persistence of Energy Savings: What Do We Know and How Can It Be Ensured?" EnergyThe International Journal, Vol. 17, No. 11,199Z PP- 21 ------- dertaken, however, a utility cannot reverse itself and return to a stricter option and receive con- tinuous credit from it. Option 1: Monitoring Subsequent-Year Savings The monitored savings verification option, which includes a requirement for inspection of installed measures, credits the utilities with the greatest possible savings. This recognizes the greater con- fidence in the existence and accurate estimate of the savings derived from monitoring. In addi- tion, the utility receives credit for,savings mat occur after it terminates monitoring as long as the utility completes two subsequent-year veri- fications. The monitoring option requires that the veri- fied savings estimated by the utility continue to meet the 75% hypothesis test used for the first- year savings calculation. As mentioned previ- ously, conversion of gross savings into net sav- ings can be done through the first-year default conversion factor or by the utility-researched fac- tor, and the first-year values can be used in all subsequent-year savings. A utility may, if it wishes, re-evaluate the net-to-gross conversion factor and use updated values. Verifications should be conducted every other year (biennially) after the first-year verifi- cation for a total of three verifications after the retrofit. Savings in off-years (i.e.. when no verk fication is performed) will be calculated at the inspection or default rate for no subsequent-year savings verification. When the next monitoring verification is performed, however, the utility may recoup uncredited savings that occurred in the off-year, based on the average of the last two measurements. More frequent verifications are also acceptable. The savings from the fourth year onwards are calculated as the average of the two subsequent-year verifications until the end of the measure's lifetime. Option 2: Inspection for Presence and Operation of Subsequent-Year Savings Surveys of the treated group (or representative samples) must be undertaken to confirm that the conservation measures are both present and ac- tively operating in subsequent years. A measure is considered to be inoperative if its performance is found to be significantly impaired, that is, about half of the first-year savings are not occur- ring. For example, clogged coils in a commer- cial refrigeration system and broken time docks for a thermostat setback control are reasonable evidence of performance deterioration beyond an acceptable limit. The first-year savings must be reduced by the fraction of sites found to have non-function- ing measures or where measures were removed. For example, if an "active conservation measure" was determined to have been removed from 10% of the sites inspected, then the subsequent-year verified savings for that year must be reduced by 10%. Following this adjustment, where ap- plicable, the utility may claim 75% of first-year savings for half of the measure's physical life- time (see Figure 5). Certain conservation measures do not need regular maintenance and are unlikely to be re- moved once they are installed. In mis sense, they are "passive conservation measures" whose en- ergy savings are likely to continue with a high degree of certainty (assuming correct installa- tion). For calculation of subsequent-year savings, passive measures will receive 90% of first-year energy savings without further inspection. Utili- ties will receive these savings for the measure's physical lifetime. A utility may use the passive measure path shown in Figure 5 if it demonstrates that the measure can be reasonably expected to remain in place and perform to first-year levels without requiring active maintenance or operation. Ex- amples of passive measures include: building shell insulation, pipe insulation, and window improvements. Option 3: Default Subsequent-Year Savings If the utility decides not to continue any moni- toring or inspection, it is still eligible for the de- fault savings. These savings consist of 50% of first-year savings but are limited to half of the measure's lifetime (as discussed previously). 22 ------- REVISED P. 23 Line Loss Adjustment Verified savings from demand-side energy conservation measures or programs may include savings from transmission and distribution line losses mat are avoided because of improved end-use efficiency. The utility may provide documentation of line losses that are avoided by a conservation program, or it may utilize the stipulation procedure outlined in Appendix D. Such savings may be eligible for either CRER allowances or reduced utilization credit. Supply-Side Efficiency Improvements Supply-side efficiency improvements can also be verified with these protocols. Although supply-side measures are not eligible for CRER allowances, they are eligible for credit with respect to annual reconciliation for reduced utilization of Phase I generating units (see 40 CFR 72.90). Thus, documenting supply-side efficiency improvements can provide substantial value to a utility that has underutilized Phase 1 units in a particular year. Common supply-side improvements include boiler and turbine improvements, heat rate improvement programs, high efficiency transformer switchouts, and other types of transmission and distribution (T&D) efficiency improvements. Under these protocols, supply-side efficiency measures are divided into two categories. The first of these is transmission and distribution efficiency improvements, where both the inputs and outputs are electricity. In general, energy savings from T&D measures or programs should be verified using the same procedures outlined in earlier sections. A reference case must be specified utilizing a statistically valid sample of installations, with careful attention to the levels of service for the equipment. The second category is efficiency improvements in generating facilities, where fuel or heat is converted into electricity. The lack of a large sample of similar generation efficiency measures precludes the use of statistical tests applied to other measures under these protocols. Instead, efficiency improvements at a generating unit can be measured in a straightforward manner by determining the improvement in conversion efficiency, or heat rate, through a systematic analysis of savings year fuel inputs and electricity outputs relative to a base year.1 If the efficiency improvement attributable to a particular measure or program cannot be precisely determined, measures or programs that collectively contribute to verified heat rate improvements may be grouped into a generic "heat rate improvement program" category for each generating unit. Although generation efficiency can be defined and measured in a number of ways, the "nominal" heat rate is the appropriate measure for purposes of these protocols. Nominal heat rate is the ratio of the unit's total fuel input to its total generation output during the relevant year, consistent with data reported on Form EIA-767. The maximum credit which may be earned for all measures at any generating unit is the reduction in fuel input attributable to the change in heat rate since the base year at utilization levels which occurred during the savings year. Further information on credit for supply- side efficiency measures with regard to reduced utilization of Phase I generating units can be found in The User's Guide to the Conservation, Verification Protocols (Version 2.0) and the Energy Conservation and Improved Unit Efficiency Report (EPA Form EC1 and instructions). For purposes of reduced utilization submittals, the base year is 1987. , ------- ------- Appendix A Definitions The following definitions are used in the CVPs. Several types of energy savings are used in this document, all of which are assumed to be elec- tricity. Since each type of energy savings has a different meaning and application, the.defini- tions are listed together below. Definitions for other terms are listed alphabetically following the energy savings terms. Energy Savings ' Raw, metered energy savings: The simple dif- ference in metered energy use between the treated customers and the comparison group during the relevant periods. Essentially no pro- cessing has been undertaken, so the periods need. not be similar and no normalizations have been performed. (The "savings" can even be negative.) This term is rarely used in this document but is defined to emphasize the distinctions in the defi- nitions of gross and net energy savings (below). Gross energy savings: The difference in energy use after adjusting for changes in levels of ser- vice and periods of measurement. Typical ser- vice adjustments include differences in weather (degree-days), operating hours, factory output, customers, illumination levels, conditioned floor area, inside temperature, etc. Periods of measure- ment for the treated and comparison groups will often differ and need to be normalized to a com- mon length (typically a year). Gross energy sav- ings seek to quantify the savings from the tech- nologies applied, so the comparison groups con- sist entirely of similar installations mat did not receive the conservation measures. Net energy savings: The energy savings attrib- utable to the energy conservation program. Al- ternatively net savings equals gross savings mi- nus free ridership and increased energy use from the snapback effect, plus any free drivership sav- ings. A standard method for estimating net en- ergy savings involves comparing the gross en- ergy savings for the treated group to the change in energy use of a similar group not participat- ing in the program, some of whom may have implemented similar conserving (or increasing) measures on their own initiative. First full-year savings: The energy savings ac- tually achieved by conservation measures after their installation for a calendar year or another 12 month period, based on measurement, esti- mation techniques, or bom. First-year energy savings: A specific period of a conservation program's energy savings. Soon after installation, the conservation measures be- gin saving energy.' For purposes of verification, however, the savings are divided into the "first- year" and "subsequent-year" savings. The "first- year" begins after the conservation program is being deployed in a routine manner and the con- servation measures are operating in their normal modes. « Service-adjusted savings (SAS): The reduction in energy use from the reference case due to the conservation measure. The SAS include adjust- ments for changes in service that may have oc- curred between the untreated, and treated groups. Subsequent-year energy savings: The energy savings occurring in the annual periods follow- ing the first year. Stipulated Savings: Energy savings for speci- fied measures (or procedures) that a utility may use instead of savings determined through moni- toring. Other Terms Active conservation measure: A conservation measure that requires regular maintenance and that may be removed once installed. 25 ------- Date of installation: The date on which the con- servation measure begins normal operation. Levels of service: Conditions describing the quality of energy-related services provided by an electric utility at the point of end use. , Lifetime: The median length of time (in years) that a measure produces energy savings. Lighting circuit: The electrical service to a group of lights serviced by its own circuit breaker or panel. In large buildings, lighting circuits are often 277 Volts; in smaller buildings 120 Volts is more common. In complex buildings, the light- ing circuit is the combination of circuits servic- ing the lights of a distinct area, such as one or more floors. Net-to-gross conversion factor: Anumber that, when multiplied by the gross energy savings, yields net energy savings. The factor is typically (but not always) between 0 and 1 and is deter- mined by analysis conducted by an electric util- ity company. Passive conservation measure: A conservation measure that does not require regular mainte- nance and is unlikely to be removed once in- stalled. Reference case: The energy use and set of des- ignated levels of service from which the savings are estimated. Service-adjusted energy use: The estimated energy use of a given group after adjusting to different levels of service. Start date: The date on which the program be- gins normal operation. 26 ------- Appendix B Service-Adjusted Savings and Hypothesis Test General Approach for Service-Adjusted Savings The goal of the service-adjusted savings (SAS) is to estimate the energy savings resulting exclu- sively from the conservation measures and not from any changes in levels of service. It begins with the assumption that the estimate is based on the difference in energy use between an un- treated and a treated group. The approach seeks to adjust energy use in the untreated and treated ' groups to similar levels of services before esti- mating the difference. It is sometimes convenient to think of the untreated case as the "pre-retro- fit" and the treated case as the "post-retrofit" case. This analogy applies for very simple situations but requires modifications when the levels of service change or the verification involves cross- sectional analysis (such as in situations involv- ing the estimation of energy savings from new construction programs). Service-Adjusted Energy Use (SAEU) The energy use for a given installation can gen- erally be described as a function of the level of services supplied. Some of the common service levels determining the variation in energy use are: Severity of heating or cooling season. -Indoor temperature. Operating hours. Conditioned floor area. Production rate (or mix). Other levels of service may be more appro- priate for special end uses or applications. The key aspect is that changes in these levels of ser- vice are responsible for differences in energy use. The relationship between service levels and en- ergy use can be expressed as a function: For lighting energy, use, the function might be of the form, E. = a + JJH., where a and p are con- stants and H. is the number of hours per year the site operates. (These coefficients might have been derived from a regression of submetered data.) The coefficients, a and p, reflect a level of effi- ciency of delivery of that service which is unique for the hardware and operating procedures in place. This relationship must be derived through monitoring and (where appropriate) market re- search data for both the treated and untreated conditions: The coefficients in the functional relationship will be different for the two situations because they reflect the efficiency at which the service is be- ing delivered. In the lighting case, for example, the coefficients a and P will be smaller in the treated case because less energy will be required to obtain the same level of service. The "reference case" is the energy use' and set of designated levels of services from which the savings are measured. The energy use asso- ciated with the reference service levels is called the service-adjusted energy use, SAEU^ By defi- nition, the reference case consists of the untreated group's energy use adjusted to service levels from the treated group (with exceptions described below): This differs from many definitions of reference cases, where the untreated (pre-retrofit) energy use and service levels establish the reference case. In this protocol, service levels from the treated , case were chosen as the reference because ser- vice levels often change as a result of the retrofit and treated case service levels are more likely to prevail for the subsequent years. For example, a building exit light retrofit will probably include retrofitting lights that were not operating (i.e.f 27 ------- burned out) in the first place. In die course of HVAC retrofits, ventilation rates are often raised to meet code requirements. In both cases, the post-retrofit conditions are a more appropriate baseline from which to measure the savings. Calculation of Service-Adjusted Savings The "service-adjusted savings" (SAS) is the dif- ference between the SAEUs: SAS * SAEUnl-SAEUtatt The SAS must satisfy the 75% confidence hypoth- esis test (discussed later) so that SAS will always be less than or equal to the difference in SAEUs: SAS & SAEUat- In most cases, the service levels will change in the untreated and treated groups. For the light- ing retrofit example, the lights may have oper- ated 3000 hours before the retrofit and 4000 after the retrofit. The reference level of service is thus 4000 hours. TheSAEl/ .is therefore an estimate BCi of the energy use of the original inefficient lights operating at the higher number of hours occur- ring after the retrofit: SAEl/w, =/, (4000) \ The SAS for this situation is: SAS = SAEU^- =/«r<«W) -f^ (4000) Note that this assumes no other changes of ser- vice level occurred in the treated case. If the light- ing retrofits also included adjusting illumination levels, this factor should be included in/^ and /neat- A possible relationship would be: where L is the average level of illumination, in lumens, as determined by a series of representa- tive spot measurements. The SAEUref and , are calculated similarly. Weather Adjustments The severity of the heating and cooling seasons fluctuates considerably from year to year. Simi- larly, other meteorological factors, such as wind speed and insolation, have long-term, average conditions that may differ from the period of measurement. Using the measurement periods to project subsequent-year savings can lead to obvious errors. Fortunately, long-run average . data for these services are available and should be used for the reference case. The appropriate reference service levels for space heating would be the long-term average number of degree-days rather than those for the treated year. The difference between the conditions dur- ing measurement and the average means that the reference case must contain some levels of ser- vice that differ from the treated period. As a re- sult, both the untreated and treated groups will need service adjustments. A new set of services must be defined such that: where Sivg becomes the S^ Energy use often depends on a combination of service levels determined by the user and the weather. For example, a building air condition- ing system's energy use might be determined by the building's operating hours and the cooling degree-days experienced: In this case, SAELJref will contain service levels from both the treated and average conditions: The SAEUtiat requires a similar adjustment: ------- As a result, the service-adjusted energy use will reflect a combination of service levels derived from the measurement periods and long-term average values. Aggregation The above procedure for calculating the service- adjusted savings can also be applied to a group of sites by aggregating the individual results or pooling data and performing a single analysis on the aggregate data. The pooled approach may be necessary when level-of-service information (such as that derived from surveys) is not avail- able for individual sites. Sampling and Hypothesis Test In all cases, the verified savings must satisfy a 75% confidence hypothesis, that is, the utility must have 75% confidence that the actual sav- ings equal or exceed the estimated amount. As indicated above: . SAS * SAEUmf- The SAS must be lowered until the 75% confi- dence in the difference SAEU . ~ SAEIJ., actu- Irl DNi ally occurs. The precise technique to evaluate the 75% confidence hypothesis will depend on the sampling procedure and experiment design. Ideally, all participants in the treated and un- treated groups should be included in the verifi- cation. Here, simple tests for confidence in the savings can be applied. (Note that uncertainty in results exists even when 100% of the popula- tion is evaluated.) the high cost of monitoring, surveys, and data management will often prevent evaluation of 100% of the participants. If comprehensive evaluation is not possible, then a statistically rep- resentative sample of participants is acceptable. If a sample of n installations is monitored, out of a population of N, then where SASu . is the average SAS for the sampled installations. In this case, a hypothesis test must be used to determine the population mean, SAS ^^ that can be extrapolated to have occurred within a certainty of 75%. ' A variety of hypothesis tests may be used.11 A calculation using a one-tailed t-test is presented here as an example. The SAS is calculated as follows: SAS. popaktioa iSAS SAS 021 where tOK is the sample standard deviation of SAS, and is the critical value of the t-statistic for a one-tailed hypothesis test at the 75% confidence level for n - 1 degrees of freedom (determined from a table). A table of critical values of the t- statistic is included as Appendix C. The total service adjusted savings that may be claimed for the population is then N x SAS ubtioo. When the Sample is the Population In certain cases, the evaluation covers every pro- gram participant. This occurs when, for example, the program targets only a few large customers, or where the evaluation has access to billing data for every participant. These situations eliminate any uncertainty in the savings due to sampling because the availability of data for the total popu- lation eliminates the need to sample. The only remaining uncertainty results from the proce- dures for estimating savings due to changes in services (that is, adjustments for differences in weather, operating hours, etc.) When the utility verifies energy savings through analysis of the entire population of af- fected customers, the adjustment to achieve 75% confidence is not required, ^Instead, the utility may calculate the service adjusted savings as-. suming 100% confidence. In a utility weather- ization program, for example, the most likely service adjustment is for differences in degree- days between the two retrofit years and the long- term average. After this adjustment is made, the 11 P. W. John, Statistical Methods m Engineering and Quality Assurance, John Wiley & Sons, New York, 1990: I 29 ------- first-year savings result, which can be used in,. subsequent calculations for net-to-gross and sub- sequent-year savings. A verification plan designed to include all the affected customers may in the end obtain usable data for less than the entire population. Sample attrition is due to unavoidable events such as terminated customers, data collection errors, participants opting out, etc. This verifi- cation path can still be used as long as 90% of the original energy use is captured in the evalu- ation. (This corresponds to 90% of the custom- ers except where a few, large participants drop out of the evaluation.) An additional 10% attri- tion per year is permitted if the utility chooses to continue monitoring energy savings. Step-by-Step Procedure The following steps summarize the procedure for calculating the service-adjusted savings: 1. Identify reference case and levels of service de- termining energy use. This usually happens after the pre-retrofit phase of metering is com- pleted! 2. Identify levels of service requiring average conditions (usually weather-related). ,' 3. Quantify relationship between service levels and energy use based on energy data, manu- facturing data, surveys, etc. 4. Calculate SAEU n{ 5. Calculate SAS at 75% confidence level. ------- Appendix C T-Statistics for 75% Confidence One-Tailed Hypothesis Testing n-l T~ "2" 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 i:000 ~816~ .765 .741 .727 .718 .711 .706 .703 .700 .697 .695 .694 .692 .691 .690 , .689 n-l 18 20 21 22 23 24 25 27 26 28 29 30 40 60 120 .688 .688 .687 .686 .686, .685 .685 .684 .684 .684 .683 .683 .683 .681 .679 .677 .674 31 ------- ------- Appendix D Stipulated Savings, Procedures, Lifetimes, and Conversion Factors Engineering Estimates of Energy Sayings Technology Description In certain circumstances, an engineering estimate (including a building simulation) of energy sav- ings will be acceptable in lieu of measurements. An engineering estimate typically requires use of recognized calculation procedures and tech- nical specifications of materials and equipment.12 Engineering estimates or building simulations of savings often differ from monitored savings be- cause of incorrect assumptions, especially regard- ing initial energy use and operating schedule." Therefore, when measurement is not,used, greater emphasis must be placed on verification of actual measure installation and correct opera- tion. Engineering calculations may be used in any one of the following situations: Monitoring costs would exceed 10% of pro- gram cost (including cost of the conservation measures). No single customer accounts for more man 20% of total savings. Energy savings are expected to be less than 5% of electricity use of the smallest isolatable cir- cuit. For example, energy savings from resi- dential lighting conversions will appear on many different circuits. In this case, the small- est isolatable circuit is the whole house. Ifen- ' ergy savings are expected to be less than 5% of total household electricity consumption, then engineering estimates may be used. Energy Savings Calculations Even within engineering calculations, there ex- ists a range of methods and associated confidence in the estimates.14 Calculations performed with few verified assumptions will be less reliable than enhanced engineering estimates, .which rely partly on site-specific information, such as oper- ating hours, initial energy use, etc. Thus, engi- neering calculations drawing upon site-specific information will receive greater credit than those relying on national averages or "typical" values derived from handbooks or other references. A simple engineering calculation that does not draw upon site-specific information to ad- just assumptions will receive only a 50% realiza- tion rate of predicted savings. The formula is given below: Verified savings = (0.5) x (Predicted Savings) The inclusion of site-specific adjustments will improve the confidence in the enhanced en- gineering estimate or building simulation. It is impossible to predict which site-specific adjust- ments are available and will be used. Therefore, " Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods for Estimating the Impacts of Demand-Side Management Programs, 3 vols., EPRITR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994; American Society of Heating, Refrigeration, and Air-Conditioning Engineers, "Energy Estimating Methods," in 1993 Fundamentals, ASHRAE, Atlanta, GA, 1993. a M Pels and K. M. Keating, "Measurement of Energy Savings from Demand-Side Programs in US Electric Utilities," Annual Renew of Energy and the Environment, Vol. 18,1993, p. 57; S. M. Nadel and K. M. Keating, "Engineering Estimates vs. Impact Evaluation Results: How Do They Compare and Why?" in N. E. Collins, ed., Energy Program Evaluation: Uses, Methods, and Results, Argonne National Laboratory, 1991, pp. 24-33. . 14 M. Pels and K. M. Keating, "Measurement of Energy Savings from Demand-Side Programs in US Electric Utilities," Annual Review of Energy and the Environment, Vol. 18,1993, p. 57; RCG/Hagler, Bailly, Inc., Impact Evaluation of Demand-Side Manage- ment Programs, EPRI CU-7179, Electric Power Research Institute, Palo Alto, CA, February 1991. 33 ------- credits will be available for different adjustments. Combinations of these adjustments can be used to obtain greater .proportions of the predicted savings (up to a 100% realization rate). The for- mula is given below: Verified savings = (0.5 + Credit #2 + Credit #2 + Credit #3) x (Predicted Savings) Up to four adjustments may be selected as long as the sum of the credits does not exceed 0.5. The amount of credit for each adjustment is given in the. table below: Adjustment Credit Verification of installation. Calibration of savings based on actual site's utility bills. Calibration of savings based on another program's results. Calibration of savings based on site-specific operating schedules. Short term measurements. One-year later verification of installation. 0.15 025 0.10 0.15 0.15 0.10 For example, an engineering estimate pre- dicted 400,000 kWh savings in the first year of a commercial lighting program. This prediction included adjustments for verified installation of the fixtures and lights and a further downward adjustment to reflect the lower savings observed in another, similar program. The verified sav- ings calculation would then include 0.15 for verification of installation and 0.1 for calibration based on another program's results. Thus, Verified Savings (0.75) x 400,000 kWhlyear 300,000 kWhlyear Verification Requirements / The following requirements must be satisfied in order for the engineering estimates to qualify as verified:' Each submittal must include brief documen- tation explaining why the verification is eligible for engineering estimates of savings. Each submittal must include a brief descrip- tion of the technologies employed in the con- servation program and the assumptions and algorithms or procedures used to estimate en- ergy savings. Each submittal must indicate the credits se- lected. If the third credit is used, an SAS calculation must be made to normalize for any differences in energy services between buildings in the ap- plicant utility's program and the reference pro- gram (see Appendix B). 34 ------- Reduced Transmission and Distribution Losses Due to Demand-Side Efficiency Improvements Technology Description v * Electric Utilities in developed countries typically experience 2 - 8% annual losses in transmission and distribution, although peak losses can be sig- nificantly above that.15 The exact percentage de- pends on the network design and other transient factors, such as ambient temperature and system loading.16 In addition, if the utility has more ef- ficient transformers on all or part of its system, the avoided transmission and distribution losses could be as much as 25% lower. Reduced energy demand at the end use leads to a reduction in transmission losses. These sav- ings will not normally be captured in the verifi- cation schemes required by the CVPs, even though some savings are to be expected. This measure gives credit to utilities for the savings. Energy Savings Calculations The savings are proportional to the total savings due to the verified conservation programs (with the exception of amorphous core transformers).17 Transmission and Distribution Savings for Residential and Commercial Programs =0.07 x Transmission and Distribution Savings for Industrial Programs = 0.035 x Verification Requirements The initial Verified Savings must not be greater than the submittals for savings from the other stipulated and monitored savings. The utility should also indicate whether its distribution transformers are made of conventional silicon- steel or high-efficiency amorphous-core metal. 15 N.-J. Bergsjo and L. Gertmar, "Superconductivity and the Efficient Use of Electricity," in T. Johansson, B. Bodlund, and R. Williams, eds., Electricity, Lund University Press, Lund, 1989; R. Loftness, Energy Handbook, Van Nostrand Reinhold, NY, 1984. 16 Barakat & Chamberlin, Inc., End-Use Technical Assessment Guide (End-Use TAG) Volume 4: Fundamentals and Methods, EPRI CU-7222, Electric Power Research Institute, Palo Alto, CA, 1991; R. Dorf, ed.. The Electrical Engineering Handbook, CRC Press, Boca Raton, FL.1993. , , ' 17 Y. Bae, Portland General Electric Company, "1988 Line Loss Study/' Portland General Electric Co., Portland, OR, 1988. 35 ------- Energy-Efficient Refrigerators Technology Description t Refrigerators consume more electricity than any other end use in American homes. As a result of federal efficiency standards and utility incen- tives, new units consume less than half that of the 20-year old models now being retired. A va- riety of utility programs have been created to re- duce refrigerator energy use, including rebates to purchase especially efficient units and to pick up older, inefficient units. The rebates are typi- cally awarded for refrigerators that use 10 - 20% less than the current standard. These programs are often linked so that the utilities retrieve old refrigerators from those homes that received re- bates on the purchase of efficient models. The federal efficiency standards are causing significant savings to occur without any inter- vention by the utilities. For mis reason, the stipu- lated savings are meant to reward savings be- yond those resulting from federal standards. This may take the form of replacing existing refrig- erators with new, efficient models and/or retriev- ing old units. (Some utilities will just rebate new refrigerators, others will just pick up second units, and still others will do both.) Energy Savings Calculations Many field studies have demonstrated energy savings from installation of new, efficient refrig- erators and from removal of second units.18 The following table summarizes the stipulated gross savings from the different actions by the con- sumer and utility: ' Consumer/ Utility Action Gross Savings Per Unit (kWh/year) Pick-up alone (true abandoning of second unit). Pick-up of old unit and purchase of rebated replacement unit. Pick-up alone (where consumer recently purchased a non-rebated refrigerator). Efficient purchase alone. 1200 600 450 300 Note that these estimates are consistent with the average new refrigerator consuming about 750 kWh/year. Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: The name and address of each incentive recipi- ent must be collected (but not submitted). The utility must submit a certified statement that the old refrigerators were operating. For second refrigerators, the utility must verify that a primary refrigerator exists and,that the consumer does not intend to purchase another refrigerator in the next year. " W. Bos, "SMUD's GraveyardConditions of the Deceased," Home Energy, January/February 1993, pp. 18-19; A. Meier, "Field Performance of Residential Refrigerators," ASHRAE Journal, August 1993, pp. 36-40; D. S. Parker and T. C Stedman, "Mea- sured Electricity Savings of Refrigerator Replacement: Case Study and Analysis," in American Council for an Energy Efficient . Economy, 1932 ACEEE Summer Study on Energy Efficiency in Buildings in Pacific Grave, CA, 1992, pp. 199-211; ]. Proctor and T. Downey, "Pacific Gas and Electric Refrigerator Rebate Evaluation: 1992 Field Monitoring Report," Pacific Gas and Electric Company; San Francisco, CA, 1992; E. A. Rogers, "Evaluation of a Residential Appliance Rebate Program Using Billing Record Analysis," in N. E. Collins, edv Energy Program Evaluation: Uses, Methods, and Results, Argonne National Laboratory, 1989, pp. 263-269; P. M. Witte and M. G. Kushler, "An Evaluation of the Michigan Second Refrigerator Bounty Pilot Demonstration/' Public Service Commission of the Michigan Department of Commerce, Lansing, MI, 1990. 36 ------- Residential Water Heating Conservation Measures Technology Description This measure involves reducing the energy con- sumed by residential water heaters by address- ing standby losses, hot water consumption, and heating efficiency. .The measures covered here are: water heater blankets, anti-convection valves, pipe insulation, low-flow showerheads, and heat pump water heaters. Energy Savings Calculations Energy savings from insulation blankets depend primarily on the temperature of the hot water being stored, the ambient air temperature, and the amount of insulation already around the tank.19 Actual savings for individual units will vary somewhat, but average savings from retro- fits are reliable and more consistent. Most of the savings measurements were undertaken 5-10 years ago.20 Since then tank insulation has in- creased substantially, so savings from the addi- tion of the blanket have fallen somewhat.21 Laboratory measurements have demon- strated savings from both anti-convection valves and pipe insulation, especially when insulating the pipes closest to the water heater (both hot and cold).32 Stipulated savings are given in the table on the following page. Low-flow showerheads reduce the flow in residential showers. Lower consumption of hot water translates into reduced electricity con- sumed for water heating. Beginning in 1994, all showerheads sold must be rated at less than 2.5 gal/min.23. Since many old, high-flow shower- heads are still in place, however, savings can be obtained by replacing existing showerheads with new, low-flow units. Showerheads with even lower flowssome less than 2 gal/minare available and often preferred by consumers.24 For this conservation measure, it is assumed that the utility makes no special attempt to target high- flow units. Recent monitored savings demonstrated that replacement of the old showerheads with a low- flow device by utility contractors saved about 500 kWh/year per household.25 This is somewhat less than engineering estimates but reflects the presence of some low-flow showerheads and lower flows than expected for older units. In ad- dition, occupants had slightly shorter showers than typically assumed. Water-faucet aerators have been recently found by Puget Sound Power & Light through 19 Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods for Estimating the Impacts of Demand-Side Management Programs, 3 vols., EPRI TR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994. 20 M. A. Brown, D. L. White, and S. L. Purucker, "Impact of the Hood River Conservation Project on Electricity Use for Residential Water Heating," Oak Ridge National Laboratory, Oak Ridge, TN, 1987; D. H. Sumi and B. Coates, "Persistence of Energy . Savings in Seattle City Light's Residential Weatherization Program," in N. E. Collins, ed., Energy Program Evaluation Conference: Uses, Methods, and Results, 1989, pp. 311-316; A. Usibelli, "Monitored Energy Use of Residential Water Heaters," in Proceedings oftheACEEE 1984 Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy, Santa Cruz, CA, 1984., - 21 Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods for Estimating the Impacts of Demand-Side Management Programs, EPRI TR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994; R. G. Pratt, B. A. Ross, and W. F. Sandusky, "Analysis of Water Heater Standby Energy Consumption from ELCAP Homes," Energy end Buildings, Vol. 19, No. 3,1993, pp. 221-234. ' 22 A. Usibelli, "Monitored Energy Use of Residential Water Heaters," in Proceedings oftheACEEE 1984 Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy, Santa Cruz, CA, 1984. 23 US. Department of Energy, "National Appliance Energy Standards," Washington, D.C., 1990. , 24 B. Mandark, "Low-Flow Showers Save WaterWho Cares?" Home Energy, July /August 1991, pp. 27-29. KC. Hickman and W. M. Warwick, "Cleaning Up: An Efficient Approach for Estimating Showerhead Savings," in Proceedings of the ACEEE1994 Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy, Pacific Grove, CA, 1994, pp. 8.89-8.95; M. Warwick and C. Hickman, "Everything 1 know about Energy-Efficient Showerheads I Learned in the Field," Home Energy, Jan/Feb 1994,3943. 37 ------- metering studies to provide reliable energy sav- ings. The aerators were found to generate the greatest savings in kitchen faucets, with some- what lower savings occurring in bathrooms. Accounting for the rather high first-year removal rate for these measures (25%), the studies con- cluded that on average each aerator would save 25 kWh/year, or 58 kWh/year for a single-fam- ily home and 68 kWh/year for a multi-family home.26 This conservation measure consists of plac- ing aerators in the water faucets in the kitchen and/or the bathroom(s). Heat pump water heaters have been avail- able for about a decade but have not been popu- lar owing to their high initial cost. Recent mod- els are much cheaper and just as efficient, how- ever. Recent studies have found that a heat pump water heater cuts heating energy use by about 50%. . , ' . - Stipulated savings for the residential water heating conservation measures are given in the table below. Measure Stipulated Electricity Savings (kWh/year) Insulation blanket around tank. 300 Anti-convection values. 100 Pipe insulation. 150 Low-flow showertieads (savings per home. for 1 or 2 per home): - . Installed by utility representative. 500 Installed by customer. 250 Water-faucet aerators (savings per home, . for kitchens and bathrooms): 50 Heat pump water heater. 1500 Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: , Each installation must be inspected by a util- ity representative, except for customer-in- stalled low-flow showerheads. . The name and address of each participating customer and date of inspection must be re- corded (but not submitted). Insulation blankets must have a thermal resis- tance of at least R-7. Anti-convection valves must be installed on both the incoming cold water and outgoing hot water pipes. Insulation must cover the first three feet from the water heater for both hot and cold pipes. These measures apply only to homes with elec- tric water heaters. 26 M. Schuldt and R- Mazzucchi, "Energy Efficient Showerheads and Faucet Aerator Metering Studies: Single and Multifamily Residences/' SBW Consulting, Inc., Presented to the Northwest Regional Evaluation Group, Portland, OR, November 10,1994 38 ------- Ground Source Heat Pumps For Homes (New and Retrofit) Technology Description Ground source, or "geothennal," heat pumps use the earth (or solar energy absorbed by the ground) as a heat source and sink. The unit uses a buried network of pipes to transfer the heat between the earth and a heat transfer fluid (usu- ally water). The ground source heat pump can operate more efficiently than a traditional air- source heat pump because the heat reservoir is wanner in the winter and cooler in the summer. However, ground source units are more costly to install and some open-loop systems use a sig- nificant amount of energy to pump the working fluid. On the other hand, life-cycle costs will usually be lower for most systems. Ground source heat pumps are increasingly popular due to improvements in their installa- tion, efficiency, and reliability. These units, are being installed primarily in rural areas, where installation of a pipe network is particularly easy.27 Improved drilling techniques, however, are making this technology available even on small lots. Energy Savings Calculations The typical alternatives to ground source heat pumps are electric baseboard heating, air-source heat pumps, or possibly Liquid Petroleum Gas (LPG). The exact system also depends on the cli- mate, that is, the relative size of the heating and cooling requirements. Thus, energy savings must be compared to these alternatives. Many utili- ties and government agencies have monitored the performance of ground source heat pumps in a wide range of climates and conditions.2* The annual stipulated savings for each ground source heat pump installation is 2.0 kWh/square foot of gross floor area of the served building.29 Verification Requirements The address and floor area of each house built with a ground source heat pump .must be re- corded (but not submitted). The utility must in- dicate the system design that would have been used if ground source heat pumps had not been selected. v Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods far Estimating the Impacts of Demand-Side Management Programs, EPRITR-100984, Electric Power Research Institute, Palo Alto, CA, 1993; National Rural Electric Cooperative Association, The Heat Pump Manual, EM-4110-SR, Electric Power Research Institute, Palo Alto, CA, 1986. * R. Freifelder, "Ground-Source Heat Pumps: Earth as Heat Source and Heat Sink," Home Energy, November/December 1990, pp. 32-38; F. Lenarduzzi, "Monitored Results from Commercial and Residential Heat Pump Projects," in S. Sami, ed., Heat Pumps in Cold Climates in Moncton, NB (Canada), Caneta Research, Inc., 1993. * M. LTicuyer, C. Zoi and J. S. Hoffman; Space Conditioning: The Next Frontier, US. Environmental Protection Agency, EPA430-R- 93-004, Washington, D.C., April 1993. , . . 39 ------- Higher Efficiency Lights in Office Buildings Technology Description This measure involves replacing lights in office buildings with higher efficiency units. It applies only to office buildings and usually consists of replacing incandescent or old fluorescent fixtures with high-efficiency fluorescent lamps, improved fixtures, and electronic ballasts. A retrofit con- sists of identical lights replaced on a single cir- cuit. A lighting circuit is the electrical service to a group of lights serviced by its own circuit breaker or panel. In large buildings, lighting cir- cuits are often 277 Volts; in smaller buildings 120 Volts is more common. In more complex build- ings, the lighting circuit is the combination of cir- cuits servicing the lights for a distinct area, such as one or more floors. No credit is given in this measure for improved switches, occupancy sen- sors, or daylight controls (although qualified sav- ings from these measures can be earned when verified through monitoring). Acreditfor air con- ditioning savings is available for buildings with central crullers (instead of window or rooftop package units).30 Energy Savings Calculations Energy savings from commercial lighting retro- fits will depend on the number of operating hours and the energy intensities of the original and new lighting'systems.31 In addition, light- ing levels are often adjusted during the retrofit.32 The exact values for these variables are difficult to determine without monitoring. Monitored savings have varied depending on the conditions outlined above.33 The stipulated savings listed below put caps on the variables that are most likely to be overestimated. As a result, use of these values will in some cases lead to underes- timates of actual energy savings. The energy savings are determined by cal- culating the difference in power consumption be- tween the old and new fixtures and multiplying by the number of hours of operation and num- ber of fixtures replaced. Annual savings (in kWh/year) for a single retrofit are calculated as follows: Annual Energy Savings = H x L x (PM - Paew) x A where: Annual Energy Savings = stipulated energy sayings (in kWh/year) for the measure. H = number of operating hours in the year following the retrofit. L = number of identical fixtures in building on same circuit operating prior to the retrofit. 30 R. A. Rundquist, K. F. Johnson, and D. J. Aumann, "Calculating Lighting and HVAC Interactions/' ASHRAE Journal Vol. 35, No. 11,1993, pp. 28-37. 31}. Eto, £. Vine, L Shown, R. Sonnenblick, and C. Payne, "The Cost and Performance of Utility Commercial Lighting Programs," LBL-34%7, Lawrence Berkeley Laboratory, Berkeley, CA, December 1993; A. Lovins and R. Sardinsky, The State of the Art: Lighting (Contpetitek), Rocky Mountain Institute, Snowmass, CO, March 1988; S. M. Nadel and K. M. Keating, "Engineering Estimates vs. Impact Evaluation Results: How Do They Compare and Why?" in N. E. Collins, ed.. Energy Program Evaluation: Uses, Methods, and Results, Argonne National Laboratory, 1991, pp. 24-33; M. A. Piette, F. Krause, and R. Verderber, 'Technol- ogy Assessment: Energy-Efficient Commercial Lighting," LBL-27032, Lawrence Berkeley Laboratory, Berkeley, CA, 1989; Rochester Gas & Electric Company, "Basis for Compensation," Document Reference # ED.H.8 (RG&E DSM #1503), Rochester Gas & Electric Company, Rochester, NY, March 11,1991; Xenergy, "Demand-Side Management Program Evaluation Scoping Study," ESSERCO Project EP90-34, Empire State Electric Energy Research Corporation, November 1990. 92 Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods for Estimating the Impacts of Demand-Side Management Programs, 3 vols., EPR1TR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994. 33 B. A. Atkinson, et al., "Analysis of Federal Policy Options for Improving U.S. Lighting Energy Efficiency: Commercial and Residential Buildings," LBL-31469, Lawrence Berkeley Laboratory, Berkeley, CA, 1992. 40 ------- POM = power consumption of original light fixture (in KW) that was operat- - ing prior to the retrofit. P^ = power consumption of replace- ment light fixture {in kW). A=air conditioning savings credit. The number of hours, H, must be less than or equal to 3300.34 The air conditioning savings credit, A, may not exceed 1.10. The difference (PM - P^) for fluorescent lights may not exceed the values given in the following table: Original Fluorescent Lamp Type Maximum Allowable .(**,-p,J inkW 2-lamp fixture 3-lamp fixture 4-lamp fixture 0.025 0.037 0.050 Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: The name and address of the building in which the retrofit occurred must be recorded (but not submitted). The number, type, and power rating of re- placed lights must be recorded for each build- ing (but not submitted). The number of non-functioning (removed or broken) lights must be recorded for each retro- fit (but not submitted). Retrofits replacing bro- ken or missing lights cannot be counted for credit with this stipulated measure. * Rated power for the replacement must include the entire system, that is, the fluorescent light and ballast. 34 Pacific Gas and Electric Co., Commercial, Industrial and Agricultural Direct Rebate Programs Hours of Operation Study, PG&E, Measurement and Evaluation Planning Section,.San Fransisco, CA, August 199Z 41 ------- De-Lamping In Commercial Buildings Technology Description This measure involves de-commissioning oper- ating fluorescent lights and fixtures in commer- cial buildings. De-lamping generally occurs where removal- of lights does not significantly affect safety or performance. De-lamping is an effective means of reducing electricity used for lighting. Ideally, the ballasts are removed or dis- connected so as to insure mat the lights will not be replaced during the next maintenance/light- replacement cycle. Studies of actual savings from de-lamping are rare because de-lamping is often undertaken in conjunction with other lighting retrofits.35 Nev- ertheless, energy savings are reasonably certain as long as tabulations of de-activated units and assumptions regarding operating hours are ac- curate.36 Estimates of operating hours are par- ticularly susceptible to overestimation.37 Energy Savings Calculations The general procedure is to calculate the num- ber of operating lamps that are removed and tabulate their total power consumption. This power consumption is multiplied by the num- ber of operating hours. The formula is given below: Annual Energy Savings = H x P( where: xA Annual Energy Savings = stipulated energy savings (in kWh/year).for the measure. H = number of operating hours in the year following the retrofit. ^removed = P°wer consumption of oper- ating lamps (in kW) that were removed or deactivated. A * air conditioning savings credit. The number of hours, H, must be less than or equal to 3300. The air conditioning savings credit, A, may not exceed 1.10. The total reduc- tion in power demanded for any lamp cannot exceed the values given below: Lamp type Maximum Credit (tcW) 4 foot straight 4(ootU-lamp 8 fool straight 0.030 0.030 0.050 Note that this measure does not apply to incan- descent lights. Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: The name and address of the building in which the retrofit occurred must be recorded (but not submitted). The number, type, and power rating of re- placed lights must be recorded for each build- ing (but not submitted). The number of non-functioning (removed or broken) lights must be recorded for each retro- fit (but not submitted). 35 J. Eto, E Vine, L. Shown, R. Sonnenblick, and C. Payne, "The Cost and Performance of Utility Commercial Lighting Programs," LBL-34%7, Lawrence Berkeley Laboratory, Berkeley, CA, December 1993; S. M Nadel and K. M. Keating, "Engineering Estimates vs. Impact Evaluation Results: How Do They Compare and Why?" in N. E. Collins, ed., Energy Program Evaluation: Uses, Methods, and Results, Argonne National Laboratory,1991, pp. 24-33. * F. Bryant, "Office Building's Delamping, Lighting Retrofit Yields Payback in 11 Months," Energy User News, December 1993, p. 10. , 37 Architectural Energy Corporation and RCG/Hagler, Bailly, Inc., Engineering Methods far Estimating the Impacts of Demand-Side Management Programs, 3 vols., EPRITR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994. ------- Exit Sign Light Replacements Technology Description niurrunated exit signs are required by law in al- most all non-residential and many multifamily residential buildings. They are required to oper- ate 24 hours/day in most situations.38 This mea- sure involves replacing existing lights in exit signs (which are mostly incandescent) with fluo- rescent lights or light-emitting diodes.39 Energy Savings Calculations The energy savings for each fixture replacement are determined by calculating the difference in power consumption between the old and new fixtures and multiplying by the number of hours of operation. Annual savings (in kWh/year) for a single fixture retrofit are calculated as follows: Annual Energy Savings = 8760 x (PoW - PMW) where: Annual Energy Savings = stipulated energy savings (in kWh/year) for the measure. 8760 = number of hours per year. Poid = power consumption of original exit light (in kW). P^ = power consumption of replace- ment exit light (in kW). The power consumption of the original exit light may be assumed to be 0.03 kW. A larger value for P^ may be used if verified by inspection or measurement of each model replaced. Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as verified: The number of replaced signs must be recorded (but not submitted) for each address. The original exit sign must be functioning and illuminated at the time of replacement. If only one lamp is illuminated, the default is 0.015 kW. Rated power for the replacement must include the entire system, that is, fluorescent light and ballast. M International Conference of Building Officials, Uniform Fire Code, Whittier, CA, May 1991. " A. Lovins and R Saidinsky, The State of the Art: Lighting (Competitek), Rocky Mountain Institute, Snowmass, CO, March 1988. 43 ------- Higher Efficiency Street Lights Technology Description This measure involves replacing existing street lights with higher efficiency units. . Energy Savings Calculations The energy savings are determined by calculat- ing the difference in power consumption between the old and new fixtures and multiplying by the number of hours- of operation. Annual savings (in kWh/year) for a single retrofit are calculated as follows: Annual Energy Savings = 4000 x (PM - where: Annual Energy Savings - stipulated energy savings (in kWh/year) for the measure. 4000 = number of operating hours per year. ' ^oid ~ power consumption of original street light (in kW). Pnfw - power consumption of replace- ment street light (in kW). Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: The number, type, and power rating of re- placed street lights must be recorded (but not submitted) for each community. The street light must be functioning and in use prior to the retrofit; otherwise each such retro- fit cannot be counted for credit with this stipu- lated measure. : Rated power for the replacement must include the entire system, that is, high-intensity dis- charge (HID) light and ballast, etc. The street light must be controlled by a func- tioning photocell. 44 ------- Higher Efficiency Motors For Constant Load Applications Technology Description r~ This measure consists of replacing or upgrading inefficient motors being used to power a constant load. The motor must be operating continuously or controlled by a timedock. The retrofit will .typically be undertaken in motors operating fans or pumps in factories, buildings, and agricultural sites* Note that this measure does hot apply to variable speed motors. Energy Savings Calculations The energy savings calculations are determined by calculating the difference in power consump- tion between the old and new motors and multi- plying by the number of annual operating hours. Annual Energy Savings = (No. of operating hours) x (Pold - PB(W) where: Annual Energy Savings = stipulated energy savings (in kWh/year) for the measure. No. of operating hours = number of op- erating hours per year, as controlled by timeclock. A maximum of 8500 hours is permitted. '^ow = P<>wer consumption of old mo- tor (inkW). P^ = power consumption of replace- ment motor (in kW). True power measurements of P^ and Pm must be performed during typical operation. Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: , The name and address of the building in which the retrofit occurred must be recorded (but not submitted). The motor application (e-gy pumping, air han- dling, etc.) must be recorded (but hot submit- ted). The measured power of the original and re- placement (or upgraded) motors must be re- corded (but not submitted). The motor must be operating continuously or be controlled by a timeclock. If a timeclock is used, the timeclock must be operating and actually controlling the motor. * Architectural Energy Corporation and RCG/Hagier, Bailly, Inc., Engineering Methods for Estimating the Impacts of Demand-Side Management Programs, 3 vols., EPR1TR-100984, Electric Power Research Institute, Palo Alto, CA, 1992-1994; S, M. Nadel, M. Shepard, S. Greenberg, G. Katz, and A. T. de Almeida, Energy-Efficient Motor Systems; A Handbook on Technology, Programs, and Policy Opportunities (Series on Energy Conservation and Energy Policy, Carl Bhunstein, ed.}, American Council for an Energy- Efficient Economy, Washington, DC, 1991; RCG/Hagler, Bailly, Inc., Impact Evaluation of Demand-Side Management Programs, EPRICU-7179, Electric Power Research Institute, Palo Alto, CA, February 1991. 45 ------- Amorphous Metal Distribution Transformers Technology Description Amorphous metal-core transformers reduce no- load losses by 70% or more over those in con- ventional silicon-steel transformers.41 No-load losses are the power required to energize the transformer and are constant regardless of the load. These are dominated by core losses. There are also load losses, which vary depending on the load and are dominated .by winding losses. For some transformers replaced energy savings from decreased no-load losses may be partly off- set by increased load losses, while in other cases load losses will also decrease. On average, such off sets will not be significant. No credit is given for decreases in load losses (although qualified savings can be earned when verified through monitoring). Energy savings accruing from the reduction of no-load losses through this measure will occur 8760 hours each year. Energy Savings Calculations The energy savings are determined by multiply- ing the 3/4 power of the replaced transformer's rated capacity by a factor representing the de- crease in no-load losses per unit of capacity (to the 3/4 power) and multiplying by 8760 hours. Annual energy savings (in kWh/year) for a single transformer are calculated as follows: Annual Energy Savings = C9-nx 3.1 x 10* x 8760 where: Annual Energy Savings = stipulated energy savings (in kWh/year) for the measure. C = rated capacity of replaced trans- former (kVA). 3.1 x 10* = decrease in no-load losses per unit capacity3'4 (kW/kVA3/4), 8760 = number of operating hours per year. Verification Requirements The following requirements must be satisfied in order for the stipulated savings to qualify as veri- fied: . ' Amorphous metal-core transformers must be used instead of silicon steel or silicon iron trans- formers: The rated capacity and the location must be collected (but not be submitted). To obtain greater man the stipulated savings, optional testing of no-load and load losses (ANSI C57.12.90-1987; IEEE C57.120 Draft 16-91) may be done for the new and old transformers. 41 P. R. Barnes, J. W. Van Dyke, B. W. McConnell, S. M. Cohn and S. L' Purucker, The Feasibility of Replacing or Upgrading Utility Transformers During Routine Maintenance, ORML-6804, Oak Ridge National Laboratory, Oak Ridge, TN, October 1994. ------- Stipulated Lifetimes for Common Conservation Measures Technology Description The effective lifetime of conservation measures is crucial in determining the cumulative impact and cost-effectiveness of conservation measures and programs. There remains considerable un- certainty regarding lifetimes, and research con- tinues to improve estimates.42 Energy Savings Calculations The stipulated median lifetimes are given in the following two tables. They are adapted from the consensus reached by the California utilities. regulatory commissions and other interested par- ties but are based on work by many groups across the country.43 These lifetimes may be used for calculations of savings as described in the CVPs. Verification Requirements The following requirement must be satisfied in order for the stipulated lifetimes to qualify as verified: The conservation measure in the submittal must closely resemble mat described in the list. 42 L. A. Skumatz, et al. "Bonneville Measure Life Study: Effect of Commercial Building Changes on Energy Using Equipment," SRC- No. 7619-R2, Synergic Resources Corporation, Seattle, WA, December 1991; Synergic Resources Corporation, "Effective Measure Life and Other Persistence Issues in DSM Programs," Interim Report #1.2, R77291R1, SRC, Oakland, CA, February 1992; Northeast Utilities, A Survey of Other Utilities for Measure-life and Persistence-of-Savings Issues, Conservation and Load Management Department, Northeast Utilities, Hartford, CT, March 1992. u Measurement Subcommittee, Measurement Protocols far DSM Programs Eligible for Shareholder Incentives (An Energy Efficiency Blueprint for California: Report of the Statewide Collaborative Process), State of California, January 1990, Appendix A; California Public Utilities Commission, Protocols and Procedures for the Verification of Costs, Benefits, and Shareholder Earnings from Demand- Side Management Programs, San Fransisco, CA, July. 1994; M. Gardner, Memorandum on Showerhead Lifetime, Northwest Power Planning Council, Portland, OR, November 1994. 47 ------- Useful Lives of Residential Energy Conservation Measures Measure Lifetime (Years) Measure Lifetime (Years) Caulking WeathefStripping Ceiling.insulation Wall insulation Low-flow showerhaads Water faucet aerators Duct wrap/insulation Pipe Insulation Water heater blanket Fluorescent bulbs Window shade awnings High-efficiency A/C Central heat pump 10 10 25 25 8 5 15 22 10 10 10 16 18. Evaporative coolers Clock thermostat High-efficiency clothes dryer High-efficiency refrigerator High-efficiency central furnace Whole house fan Double glazing Storm windows Furnace repair Efficient electric water heater -X Heat pump water heater Point-of-use water heater . Solar water heater 15 15 18 20 20 15 25 10 15 15 13 12 1S 48 ------- Useful Lives of Commercial and Industrial Energy Conservation Measures Equipment Type & Descriptor Lifetime (Years) LIGHTING Energy-efficient fluorescent lamp 5 Same as above with built-in ballast 2 .Energy-efficient ballast 11 Electronic ballast 10 Metal hallde lamp - 15 Low-pressure sodium lamp 5 High-pressure sodium lamp 5 Parabolic fixture , 20 Dimming systems 20 On-off switching 7 Motion sensor , 10 HVAC Economizer 12 Chiller strainer cycle system 15 Air-to-air packaged heat pump 10 Water-to-air packaged heat pump 15 Ice thermal energy storage 19 Water thermal energy storage 20 Plate type heat pipe recovery system . 14 Rotary type heat recovery system 11 Heat recovery from refrigerator condenser 11 Low leakage damper . 9 Variable air volume, inlet vane controls ' 11 Variable pitch fan for cooling power 13 Make-up air unit for exhaust hood 10 Air de-stratification fanpaddle type 10 Air de-stratification fanhigh inlet/ low discharge 15 Air curtain 10 Deadband thermostat 13 Spot radiant heat 10 Equipment Type & Descriptor Lifetime (Years) CONTROLS Computer logic energy-management systems . "13 Electronic controls 11 Time clocks 10 MOTORS, DRIVES, &TRANSFORMERS Standard electric motor 15 High-efficiency electric motor 15 Variable-speed DC motor 18 Variable-speed drivesolid type 15 Variable-speed drivebelt type 10 Efficient AC electric transformer 30 REFRIGERATION Unequal parallel refrigeration . 14 Condenser float head pressure control 10 Auto cleaning system for condenser tubes 15 Hot gas bypass defrost 10 Polyethylene strip curtain ... 3 Refrigeration case cover 11 BUILDING ENVELOPE Double glazing . 20 Heat mirror . 18 Low-emissivtty coating , 14 Solar shade film (retrofit) 7 Tinted & reflective coating 14 49 ------- i.-nc : £85 Stipulated Net-to-Gross Factors for " " Stipulated Measures Description Conservation Measure Net-to-Gross Conversion Factor Conversion factors in*the following.list can be used to convert the gross annual savings calcu- lated in the stipulated measures to net, first-year savings. These values should be used when utili- ties cannot provide estimates based on their own programs or market research. Factors such as free riders that affect the net impacts of conservation programs vary widely by the type of conserva- tion program and market conditions.44 Utilities may conduct their own analyses and conclude that their conservation programs have larger net energy savings, rather than rely on the default factors given below. Several recent studies have found, for example, free-rider rates of 10% or less. Net savings are calculated with the follow- ing formula: Net Savings = Annual Energy Savings x NGCF when: NGCF = "Net-to-Gross-Conversion Factor" listed in the table opposite: Engineering estimates of energy savings - -. - 0.70 .Exit sign light replacements 0.60 Higher efficiency motors 0.60 ../ Higher efficiency lights in office buildings 0.60 De-lamping 0.80 Higher efficiency street lights 0.90 Refrigerators pick-up 0.70 high-efficiency replacement 0.90 Residential water heating Insulation blankets 0.60 Anti-convection valves 0.90 Pipe insulation 0.60 Low-flow showerheads and faucet aerators (utility-installed) 0.70 Low-flow showerheads and faucet aerators (customer-installed) 0.50 * Heat pump water heaters 0.95 Ground source heat pumps 0.95 44 W. Saxonis, "Free Riders and Other Factors that Affect Net Program Impacts/' in E. Hirst and ]. Reed, eds., Handbook of Evaluation of Utility DSM Programs, ORNL/CON-336, Oak Ridge National Laboratory, Oak Ridge, TN, December 1991, pp. 119-34; B. M. Tolkin and P. R. Rathbum, "Quantifying Free-Ridership in Four Different Customer Segments," Proceedings of the ACEEE1992 Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy, Pacific Grove, CA, 1992, pp. 7.243-7.249; M. A. Brown and P E. Mihlmester, "Summary of California DSM Impact Evaluation Studies," ORNL/CON-403, Oak Ridge National laboratory. Oak Ridge, TN, October 1994. ------- 51 ------- |