ESTIM IMATIN ING M G MEASUR URE L LIF IFE F FROM BEHAVIO - - PowerPoint PPT Presentation

Opinion Dynamics

ESTIM IMATIN ING M G MEASUR URE L LIF IFE F FROM BEHAVIO IORAL P PROGR GRAMS

Hannah Arnold

BECC 2013 November 20, 2013

Opinion Dynamics

…What we know, What we need to know, How it challenges our thinking

Estimating Measure Life from Behavioral Programs 3

Wha hat i is t the he i issue a and nd w why d y does i it ma matter?

Opinion Dynamics

Why a y are w we t talki lking ng a about t thi his i issue?

Estimating Measure Life from Behavioral Programs 4

§ Program administrators (PAs) have been offering behavioral programs for a relatively short time and their fate as an effective program intervention depends on their associated costs and benefits. § As part of cost effectiveness calculations, PAs look to Effective Useful Life (EUL) an estimate of the median number of years that a measure installed under a program is still in place and operable. § However, EUL is less clear when considering actions taken by participants due to behavioral programs particularly when those actions are not equipment purchases or constitute multiple equipment installations.

Opinion Dynamics

Purcha hase a and nd no non-p n-purcha hase a actions ns a are b both a h aspects o

• f

beha havior

5

End Use Profile Percentage of Total Annual kWh Waste Profile Percentage of Annual Lighting kWh

Lighting

Waste: ¡How ¡much ¡of ¡the ¡ remaining ¡usage ¡is ¡due ¡ to ¡inefficient ¡equipment ¡

• vs. ¡wasteful ¡behavior? ¡

Current ¡Usage: ¡How ¡ much ¡electricity ¡ actually ¡goes ¡to ¡each ¡ end ¡use? ¡ Efficient ¡Usage: ¡How ¡li=le ¡ ligh>ng ¡energy ¡could ¡be ¡used ¡if ¡ all ¡customers ¡installed ¡efficient ¡ lamps ¡and ¡turned ¡lights ¡off ¡when ¡ ¡ not ¡needed? ¡

Opinion Dynamics

De Determi mini ning ng a an E n EUL UL f for b beha havioral p l programs ms i is c current ntly ly conf nfound nded b by t y two i issues

Estimating Measure Life from Behavioral Programs 6

§ Uncertainty as to what occurs with non-purchase energy efficiency actions without continued program intervention § The practical application of an EUL given how these programs are implemented

Uncertainty Application

Opinion Dynamics

Unc ncertaint nty A y Around nd t the he S Source o

• f E

Ene nergy S y Saving ngs

Estimating Measure Life from Behavioral Programs 7

§ It is unclear what specific actions are driving savings, and therefore how long those savings might persist § Once a behavior is internalized, it is known to persist without continued prompting from outside sources. However, behaviors can decay and not are habituated indefinitely § If non-purchase behaviors persist, we have no empirical evidence about the length of time they persist

Opinion Dynamics

Practical A l Appli lication o n of a an E n EUL UL i in C n Cla laime med S Saving ngs

Estimating Measure Life from Behavioral Programs 8

§ Unlike standard programs with a specific “installation date,” behavioral programs treat customers over periods of time, usually multiple program years, in which customers can take a wide range of actions. § This complicates how a program team might claim savings that extend beyond one year.

§ One example is claiming an EUL for each year of continued treatment

Opinion Dynamics

Cha halle lleng nges o

• f P

Potent ntial E l EUL UL a appli lication n

Estimating Measure Life from Behavioral Programs 9

§ Under this scenario, savings would be effectively double counted year-over-year.

Note: orange shading represents double counted savings for customers who are present the first year, as well as subsequent years.

PY1 PY2 PY3 PY4 PY5 PY1 * PY2 * PY3 * Years Savings are Claimed Actual Treatment Year

Estimating Measure Life from Behavioral Programs 10

Wha hat w we kno know

Opinion Dynamics

The he C Current nt S State o

• f E

Evidenc nce f from A m Ana nalys lyses

Estimating Measure Life from Behavioral Programs 11

§ Persistence with treatment § Persistence without treatment

Opinion Dynamics

Persistenc nce w with T h Treatme ment nt

Estimating Measure Life from Behavioral Programs 12

§ Our review of behavior-based programs demonstrates that savings continue to persist with treatment year-over-year when the program continues to provide information to participants.

Persistenc nce with T h Treatme ment nt SMUD UD (Int Integral A l Ana nalyt lytics, , 20 2012) 12) Nationa nal Gr l Grid (Opini nion Dyna n Dynami mics, 2 , 2012) Puget S Sound nd E Ene nergy (KEMA, 2 , 2010) Coho hort Wave 1 Wave 2 2009 Electric 2010 Electric 2009 Gas Electric Gas Year 1 1 1.80% 1.60% 1.61% 1.25% 0.81% 1.71% 1.17% Year 2 2 2.40% NA 2.06% 1.63% 1.25% 2.00% 1.46% Year 3 3 2.40% NA 2.21% 2.12% 1.43%

• Year 4

4 2.10% NA NA NA NA 2.60% 1.30%

Opinion Dynamics

Persistenc nce w with T h Treatme ment nt ( (Cont nt.) .)

Estimating Measure Life from Behavioral Programs 13

Lift i in U n Uptake

(Treatme ment nt % % - C

• Cont

ntrol % l %)

Measure L Life ( (Years)

Average a and nd R Rang nge Building Envelope Consumer Electronics Low-Cost Measures Appliances Light Fixtures Heating / Cooling Consumer electronics Hot water usage Lighting Space Heating / Cooling Refrigerator Maint. HVAC Maintenance

• 4.7%
• 0.7%

0.8% 0.8% 4.2% 4.2% 1.8% 2.2% 5.0% 6.2% 6.8% 7.0%

• 10%
• 5%

0% 5% 10% 15%

2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

HVAC maintenan Refrigerato r Lighting Space heating Consumer electronics Hot water usage Heating / Cooling Light Fixtures Appliances Low-Cost Measures Consumer Electronics Building Envelope

Chart o

• rder

ered ed f from h highes est t to l lowes est l lift i in uptake b e by t trea eatmen ent g group Range b e bands r rep epres esen ent m minimum a and m maxi ximum mea easure l e life o e of m mea easures es w within ea each g group

Beha haviors Me Measu asures s

Opinion Dynamics

Example le o

• f P

Potent ntial B l Beha havioral S l Saving ngs o

• ver T

Time me d due t to the he E EUL UL o

• f M

Measures

Estimating Measure Life from Behavioral Programs 14

0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Annu nnual S l Saving ngs le level a l as Percent ntage o

• f P

PY1 S Saving ngs Years f from P m Program S m Start

PY1 Poten>al ¡persistence ¡of ¡HER ¡program ¡savings ¡ electronics ¡& ¡ligh.ng ¡ measures, ¡4-­‑6 ¡yrs ¡ recycled ¡refrigerator, ¡5yrs ¡ showerheads ¡& ¡ aerators, ¡10 ¡yrs ¡ ¡

1.7% savings

Opinion Dynamics

Persistenc nce w witho hout T Treatme ment nt

Estimating Measure Life from Behavioral Programs 15

§ Program administrators and evaluators are studying the persistence of savings from behavioral programs once the treatment (i.e., the messages from the program) has been stopped several jurisdictions, but it’s still early § There are current analyses within the industry that indicate that savings persist for longer than a year, but these analyses do not help answer how long they last.

Estimating Measure Life from Behavioral Programs 16

Wha hat w we ne need t to kno know

Opinion Dynamics

At p present nt, t , the here i is i ins nsufficient nt e evidenc nce f for a a s specific E EUL UL estima mate

Estimating Measure Life from Behavioral Programs 17

§ Current studies do not provide conclusive evidence of standard, predictable actions taken as result of the programs. § Program design and implementation matters:

§ Empirical research has demonstrated that savings magnitude and persistence with treatment varies based on target population and program model (opt-in vs. opt-out). § Further, the frequency and duration of behavior interventions has a big impact on the persistence of the behavior being promoted by the intervention.

Opinion Dynamics

Ind Industry r y research o h on E n EUL UL i is ne needed

Estimating Measure Life from Behavioral Programs 18

§ Cohorts must be dropped from treatment to assess persistence and some experimentation is happening § There are two primary ways to determine the EUL of these programs:

§ Conduct a longitudinal persistence test: Remove treatment

• f reports and observe how savings change over time

0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 PY Poten>al ¡persistence ¡of ¡HER ¡program ¡savings ¡ electronics ¡& ¡ ligh.ng ¡ measures, ¡4-­‑6 ¡ yrs ¡ recycled ¡ refrigerator, ¡ 5yrs ¡ showerheads ¡ & ¡ aerators, ¡10 ¡ yrs ¡ ¡

§ Conduct annual survey research: Determine measure installations due to the program

Opinion Dynamics

Wha hat E EUL UL s sho hould ld b be u used a and nd ho how?

Estimating Measure Life from Behavioral Programs 19

§ At present, savings should be counted only in the current program year and research should be conducted to figure out what’s occurring in the home § However, in jurisdictions where there is interest in developing a revised value: § Leverage the stakeholder process § Clarify how these savings can be claimed over time § Use conservative estimates

Estimating Measure Life from Behavioral Programs 20

How t thi his c cha halle lleng nges o

• ur c

current nt thi hinki nking ng

Opinion Dynamics

A C Cha halle lleng nge f for t the he A Audienc nce

Estimating Measure Life from Behavioral Programs 21

§ Our regulatory framework was developed to deal with equipment-based programs, and the engineering-based assessments of these measures

§ The framework needs to change

§ Persistence for behavioral programs is very different than for traditional “effective useful life” concepts

§ The language needs to change

Opinion Dynamics

Tha hank Y nk You

Estimating Measure Life from Behavioral Programs 22

Hannah Arnold Project Manager 510-444-5050 ext. 118

harnold@opiniondynamics.com

Visit us at www.opiniondynamics.com to take our Energy Efficiency Industry Survey for your chance to win an iPad!