1. Measures that Save The Most Energy Jackie Berger David Carroll ACI New Jersey Home Performance Conference March 5, 2010

2. Session Outline 1.Introduction 2.Measuring Energy Savings – Projections 3.Measuring Energy Savings – Billing Data 4.Average Savings by Type of Measure 5.Energy Education Savings Potential 6.Maximizing Measure Savings 7.Conclusions 2

3. Introduction - Perspective Evaluator’s Perspective Based on findings from: –Program design research –Survey research –In-field research –Energy impacts 3

4. Introduction - Scope Sources –APPRISE evaluation studies –Blasnik and Associates evaluation studies –Dalhoff and Associates evaluation studies –ECW plug load study Geographic scope –Northeast –Midwest –Mountain 4

5. MEASURING ENERGY SAVINGS Projections 5

6. Projected Savings vs. Measured Savings Value of projections Projection methodology Issues with projections Comparison of projected savings to measured savings 6

7. Projections vs. Impacts Data Needs Projections Data Driven Projections Impacts Installed measuresNoYes Pre treatment usageNoYes Post treatment usageNo Yes Degree daysNo Yes Comparison groupNo Yes 7

8. Projections vs. Impacts Basic Projection Methodology –Assumptions Measure installation rates Measure retention rates Pre installation usage Measure effectiveness 8

9. Projections vs. Impacts Basic Projection Methodology –Calculation Average household saving = Measure Installation Rate * Measure Retention Rate * (Pre Installation Usage – Post Installation Usage) 9

10. Projections vs. Impacts Basic Projection Methodology –Calculation Pre Installation Usage per bulb per hour = 60 watts *.001 =.06 kWh Post Installation Usage per bulb per hour = 13 watts *.001 =.013 kWh Change per Bulb per hour =.06 -.013 =.047 kWh 10

11. Projections vs. Impacts Basic Projection Methodology –Calculation Change per bulb per day =.047 kWh * 2.5 hours/day =.1175 kWh/day Change per bulb per year =. 1175 kWh/day * 365 days = 43 kWh/year 11

12. Projections vs. Impacts Basic Projection Methodology –Calculation Number installed per home = 43 kWh * 8 bulbs = 344 kWh Retention rate = 344 kWh *.8 = 275 kWh saved per home per year 12

13. Projections vs. Impacts So simple, what could go wrong… Incorrect assumptions –Measure installation rate –Measure retention rate Bulbs left for occupants to install Bulbs removed Bulbs broken –Existing bulb kWh –Hours of use 13

14. Projections vs. Impacts So simple, what could go wrong… Interactions –Adding up individual measure savings can overstate results –Need to account for reduced heat gain from CFLs Increase heating usage Reduce cooling usage 14

15. Projections vs. Impacts Survey Results Annual kWh Savings by Hours Used 2.5 hr/day1.5 hr/day Bulbs provided (database)15.4660396 Bulbs provided (client reported)12.0515309 Number installed by auditor or client11.6497298 Number not burned out or removed10.6455273 (clients reported that 46% are used at least 1 hour per day) 15

16. Projections vs. Impacts Impact Analysis Results High UseModerate Use ProjectedActualProjectedActual kWh Savings per Bulb76417226 kWh Savings per Home1,231677878316 Survey Results – High and Moderate Use Months After Service Delivery 4-67-910-1213-1516-19 % Burned Out6%8%9%13%17% Source: M. Blasnik and Associates. 16

17. Projections vs. Impacts How far are we off with the projections? Evaluations that measure actual usage impacts usually find 50% to 70% of projected savings –NEAT Audit – measured savings were 57% and 54% of projected savings (Sharp, 1994 and Dalhoff, 1997) –Ohio electric baseload savings were 58% to 68% of projected –NJ electric baseload savings were 60% - 69% of projected Source: M. Blasnik and Associates. 17

18. MEASURING ENERGY SAVINGS Billing Data 18

19. Average Savings by Measure Type Methodology for developing measured savings Methodology for attribution of savings to measures Evaluation findings – electric baseload Evaluation findings – space heating measures 19

20. Usage Impact Analysis Usage Impact Methodology –Obtain pre and post energy usage data for program participants –Use regression model to adjust usage for changes in weather from “normal weather year” –Construct weather normalized change in usage for treated households –Construct weather normalized change in usage for comparison households 20

21. Usage Impact Analysis Usage Impact Methodology –Run regression to determine measure specific impacts Usage change = α + β * household characteristics + γ 1 * measure 1 + γ 2 * measure 2 + γ 3 * measure 3 + μ 21

22. Measure Savings – Evaluation Findings kWh Savings Per Measure Ohio EPP 1 PPL WRAP 1 NJ CP 1 CO E$P 1 PECO LIURP 2 High Use Mod Use Baseload Full Cost CFL41264047 Refrigerator926544777532912692949 Freezer760 Air conditioner172 1 M. Blasnik and Associates. 2 APPRISE. 22

23. Measure Savings – Evaluation Findings $ Cost per kWh Savings By Measure NJ CP CFL$0.061 Refrigerator$0.069 Source: M. Blasnik and Associates. 23

24. Measure Savings – Evaluation Findings CCF Savings Per Measure NJ CPCO E$PIL WAPOH WAPIA WAP Heater Replacement 85114146117 Attic insulation851109214865 Air sealing2470-10853 Thermostat41 Source: M. Blasnik and Associates. 24

25. Measure Savings – Evaluation Findings $ Cost Per CCF Saved By Measure NJ CP 1 CO E$P 1 PECO 2 Heater Replacement$1.30$1.90 Attic insulation$0.60$0.20$1.28 Air sealing$1.23 Thermostat$0.19 1 M. Blasnik and Associates 2 APPRISE 25

26. ENERGY EDUCATION Potential Savings 26

27. Potential for Education Major opportunities Potential vs. realization Successful models 27

28. Potential Education Savings WattageReductionNumberSavings Electric MeasureskWh Turn off lights604 hrs/day4350 Turn off lights at night608 hrs/day2350 Reduce central AC3º3ºAll times250 Reduce TV usage1004 hrs/day2292 Turn off computer2508 hrs/day1730 Gas MeasuresTherms Turn down water temperature10°All times25 Turn down thermostat2°2°All times84 Use cold water for clothes washingCold wash4/week52 Set back temperature at night4º4ºNight58 AC – 72 to 75 degrees, heating 72 to 70 degrees 28

29. ECW Plug Load Study Telephone survey and mailed appliance survey 50 site visits –Household survey –Electronics inventory –Metering (5-30 appliances per home) Metered for one month 6-minute intervals Computers, televisions, audio, telephone, HVAC – space heaters, dehumidifiers, room AC, fans, humidifiers Kitchen appliances 29

30. ECW Plug Load Study Potential Education Savings Computers Potential Savings from Full Power Management % of homes Estimated Annual Savings Always on20%400 Long idle periods40%190 Off when not in use25%15 Not used much15%2 Average savings160 30

31. ECW Plug Load Study Potential Education Savings Wattage When Not in Use Annual kWh Savings if Unplugged When Not in Use Mini Stereos23200 Older TVs758 Printers643 DVD/VCR Player759 31

32. ECW Plug Load Study Potential Education Savings Saving Strategies –Power management –Unplug –Turn off –Use timer –Use power strip Assessment –Potential savings –Motivation 32

33. Education Impacts Ohio EPP Unprompted Agreed toTaken Turn off lights54%16% Turn off appliances14%3% Use CFLs10%5% Conserve energy10%2% Use double spin on clothes washer9%2% Reduce heating temperature5%1% Line dry clothes4%0% Reduce water heater temperature3%1% Wash clothes in cold water1% None19% 33

34. Education Impacts Niagara Mohawk Unprompted Actions Taken As a Result of: WorkshopVideo In-Home Education Turn off lights43%40%33% Install CFLs27%20%24% Turn down thermostat14%15%10% Reduce TV usage11%3%6% Turn off appliances11%9% Turn down water temperature10%12%10% Reduce use of AC9%3%6% Use cold water for clothes washing9%5%6% Set back temperature at night/when out5%4%2% 34

35. Behavioral Impacts 35 Reduced use of _____ as a result of participating in the program? Obs. Mean Electric Savings (kWh) All Customers233854 Electric space heater** yes1051150 no128611 Air conditioner yes136947 no97723 Electric dryer yes71995 no162792 Dehumidifier yes181058 no215837 Number of lights left on all night* yes431174 no190781

36. Recap Projected savings tend to overestimate Billing data are critical Potential for savings from education 36

37. Maximizing Savings Programs that save the most: –Target measures to the highest use households –Install measures in a way that maximizes effectiveness –With an understanding of what is going on in this house 37

38. Targeting Usage (ccf)SpendingSavings$ per ccf saved <1,000$65326 ccf$25 1,000-1,400$83680 ccf$10 1,400+$1,043171 ccf$6 38

39. Targeting kWh Savings Per Measure Ohio EPP 1 PPL WRAP 1 NJ CP 1 CO E$P 1 PECO LIURP 2 High Use Mod Use Baseload Full Cost CFL41264047 Refrigerator926544777532912692949 Freezer760 Air conditioner172 1 M. Blasnik and Associates. 2 APPRISE. 39

40. Measure Effectiveness Duct Sealing –Ducts outside envelope = High Savings –Ducts inside envelope = Low/No Savings –Ducts in basement or crawl space = It Depends Insulation –With properly sealed envelope = High Savings –Without air sealing = Low Savings 40

41. Focus on This House Example – Baseload Job in Massachusetts House –Pre-visit Information: Annual electric usage of 10,000 kWh –On-Site Measurement: 6,000 kWh for appliances / 4,000 kWh for space heater –Problem: Program only pays for baseload measures –Solution: Install cfls, encourage behavioral changes, and refer to electric heat program 41

42. Maximizing Savings Programs that save the most per dollar spent: –Spend lots more when there are more opportunities –Spend substantially less when there are fewer opportunities 42

43. Targeting Usage (ccf)SpendingSavings$ per ccf saved <1,000$65326 ccf$25 1,000-1,400$83680 ccf$10 1,400+$1,043171 ccf$6 43

44. Maximizing Savings Programs that save the most per dollar spent: –Conduct tests to focus resources and time –Use models as a guide for action 44

45. Testing Field inspections of New Jersey programs found that better testing was needed to … –Find and isolate sources of infiltration in complex structures (enclosed porch, addition, sun room) –Identify unobservable leaks in ductwork outside the thermal envelope 45

46. Testing Blasnik refrigerator study found that testing is needed, but more is not necessarily better … –Low Savings / Net Benefits Rating Protocol = $101 1 Hour Metering = $111 2 Week Metering = $135 –High Savings / Net Benefits Rating Protocol = $419 1 Hour Metering = $414 2 Week Metering = $445 46

47. Audit Tools / Modeling Benefits –Clarify decision rules on measure installation –Improve consistency across program Barriers –Data entry can be a communications barrier –Reconciliation is poorly understood 47

48. Financial Decision Rules Spending Limits –Do they focus delivery on highest saving measures or restrict delivery of cost-effective measures? Spending Goals –Do they ensure comprehensiveness or encourage a program to over-invest? Spending Targets –Do they furnish flexibility or result in over-investment in some homes and under-investment in others? 48

49. Recommendations Usage Data – Essential for good decision-making Decision Criteria - Field staff need a good tool for determining which measures to install Financial Guidelines – Should vary with energy savings potential and should be expressed as a range 49

50. Contact Information Jackie Berger, 609-252-8009, jackie- berger@appriseinc.orgjackie- berger@appriseinc.org David Carroll, 609-252-8010, david- carroll@appriseinc.orgdavid- carroll@appriseinc.org 50