Demand Response & Smart Grid Research from the Electricity Markets and Policy Group

November 2016

Informing Decision-Makers About the Experiences With and Opportunities for Demand Response Pricing and Programs

Peter Cappers Chuck Goldman Jennifer Potter

Andy Satchwell C. Anna Spurlock Annika Todd

Core Demand Response & Smart Grid Staff:Electricity Markets and Policy Group

2

Thanks to our funders at the U.S. Department of Energy: Office of Electricity Delivery and Energy Reliability and Office of Energy Efficiency and Renewable Energy

Emphasis of Demand Response and Smart Grid Research in Electricity Markets and Policy Group

3

We conduct research on

DR/SG markets, policies, costs,

benefits & performance

Policy Analysis & Technical

Assistance

Cost, Benefit, & Market Potential Analysis

Integrating Variable

Generation Resource

Design, Implement-

ation and Evaluation

Our work in each of these areas focuses on demand response opportunities and smart grid enabled consumer data & programs

Topic 1. Primary Research on Design, Implementation and

Evaluation Experience

Research seeking to better understand key elements of demand response rate and program opportunities as well as technologies enabled by advanced metering

infrastructure4

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Enro

llmen

t Rat

e fo

r SG

IG C

onsu

mer

B

ehav

ior S

tudi

es o

f Tim

e-ba

sed

Rat

es

Opt-in Opt-out

Residential Customers’ Preferences for Time-Based Rates

5Source: Cappers et al. LBNL Report. 2016. Sources: Cappers et al. LBNL Report. 2016.

ENROLLMENT

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Ret

entio

n R

ate

of S

GIG

Con

sum

er

Beh

avio

r Stu

dies

of T

ime-

Bas

ed R

ates

(2

nd Y

ear)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Ret

entio

n R

ate

of S

GIG

Con

sum

er

Beh

avio

r Stu

dies

of T

ime-

Bas

ed R

ates

(1st

Yea

r)

Opt-in Opt-out

RETENTION

18%

95% 92% 91%

1% 3%

Opt-in Opt-out Opt-in Opt-out Opt-in Opt-out

Enrollment Rate Retention Rate Drop-Out RateSMUD Consumer Behavior Study

Experiences of Customer Subpopulations on Voluntary vs. Default TOU

6

0.00.10.20.30.40.50.60.70.80.91.0

-100 -50 0 50 100Cu

mul

ativ

e Di

strib

utio

nPredicted Summer Bill Savings ($)

Always Takers Complacents

0%

5%

10%

15%

20%

Summer 2012 Summer 2013

Perc

ent p

eak

perio

d ho

urly

sa

ving

s per

hou

seho

ld Always Takers

Complacents

Always Takers19.5%

Complacents78.5%

Never Takers2.0%

Control Group

(N=39,323)

VoluntaryEnrollmentApproach

(N=10,865)

DefaultEnrollmentApproach(N=2,064)

Enrolled Not Enrolled

Source: Cappers et al. LBNL Report. 2016.

Experiences of Vulnerable Customer Subpopulations on CPP

7Source: Cappers et al. LBNL Report. 2016.

-45%

-40%

-35%

-30%

-25%

-20%

-15%

-10%

-5%

0%

5%

Non-Vulnerable Sub-popualtion Load Response (%

of Peak Load )Vulnerable Sub-population Load Response

(% of Peak Load)

Voluntary: Low Income

Voluntary: Elderly

Voluntary: Chronic Illness

Default: Low Income

Default: Elderly

Default: Chronic Illness

Note: The markers in this graph indicate the estimated load response as a percent of average consumption. For any of the points that lie in the gray bar area, the difference between the estimated load response for the vulnerable population was not statistically significant (at a 90% confidence level) relative to the non-vulnerable counterpart population. The gray bar in and of itself is not the 90% confidence interval, but rather a graphical way of showing which estimated differences are statistically significant at the 90% confidence level and which are not.

Note: These data are limited to those who responded to the survey. The percent of vulnerable households in the general population are based on those households from the control group that responded to the survey. * indicates that the difference between the percent of study participants that are vulnerable versus the percent that are vulnerable in the general population are statistically significant at least at the 90% confidence level, all other differences are not statistically significant.

8

Behavior Analytics Machine learning clustering algorithms

to understand patterns of discretionary energy consumption.

C-tree algorithm to predict enrollment and create customer segments based only on pre-program energy use metrics.

(Top 3 of 99 clusters)

-100%

-50%

0%

50%

100%

150%

-30% -20% -10% 0% 10% 20% 30%

-100%

-50%

0%

50%

100%

150%

-30% -20% -10% 0% 10% 20% 30%

0-5%

5-10%

10-15%

15-20%

20-30%

TOU

Ctr

ee

CPP

Ctre

e

Psyc

hogr

aphi

c

Percent of Population

Peak

Per

iod

Savi

ngs o

n Ev

ent D

ays

(Per

cent

Diff

eren

ce fr

om P

opul

atio

n Av

erag

e)

Enrollment Probability(Percent Difference from Population Average)

① This analysis was better at differentiating customers based on their probability of enrolling than traditional psychographic marketing methods.

② Those that saved the most once enrolled is the program were the least likely to enroll in the first place.

③ Those same groups that were the least likely to enroll were actually structural winners.

Structural Winners③

②

①

DR Experience in ISO/RTOs Where ARCs are

allowed to directly enroll customers in ISO/RTO DR programs, they have taken the majority of the market

Where they are precluded, traditional utility programs still dictate enrollment opportunities

This has implications for future DR opportunities and enrollment approaches

9Source: Cappers and Satchwell. LBNL Report. 2015.

0%10%20%30%40%50%60%70%80%90%

100%

MISO ISO-NE NYISO

Allo

catio

n of

Pot

entia

l Pea

k Lo

ad R

educ

tion

(201

5)

Traditional Utility Load

Management Programs

ISO/RTO Emergency

and Capacity Programs

0

500

1,000

1,500

2,000

2,500

3,000

Subs

crib

ed N

YISO

EDR

P &

ICAP

SC

R M

W

Utility Non-Utility

Topic 2. Cost, Benefit and Market Potential Analysis

Research seeking to quantify the costs, benefits and market potential of demand response rate and program

opportunities as well as technologies enabled by advanced metering infrastructure

10

California Demand Response Potential

11Source: Alstone et al. LBNL Report. 2016.

• Two year study conducted to evaluate DR in bifurcated framework for the California Public Utility Commission (CPUC)

• LBNL team developed “Demand Response Futures” Model that builds supply curves of DR service types- Shape, Shift, Shed & Shimmy- that provide service to the system

• Identified significant value to the CA system from DR technologies that can “Shift” hourly loads to address duck curve

• Developed extensive database of DR enabling technology costs and performance

DR in WECC Planning

12Source: Satchwell et al. LBNL Report. 2013.

1,966

608 506 480 390 281 200 131 60 54 52 45 34 26 20 12 6 0

500

1,000

1,500

2,000

2,500

2026

WEC

C DR

Cap

acity

-Co

mm

on C

ase

(MW

; non

-coi

ncid

ent p

eak) Interruptible

Direct Load Control

Price Responsive

Load as a Capacity Resource

LBNL works with WECC staff and the State and Provincial Steering Committee (SPSC) to develop DR assumptions and modeling inputs for WECC’s regional transmission planning studies

Two types of DR modeling assumptions required for each study case: DR resource quantities: How much DR is available to be dispatched in

any given hour for each load zone? DR dispatch mechanics: When is the DR dispatched and how does it

affect hourly loads and peak demand?

Topic 3. Integrating Variable Generation Resource

Research seeking to understand the role demand response rate and program opportunities can play in

managing the integration of variable generation resources at the bulk and distribution system levels

13

Bulk Power System Service TOU CPR CPP DA-RTP RT-RTP

Spinning Reserves

Supplemental Reserves

Regulation Reserves

Imbalance Energy

Hour-ahead Energy

Multi-hour Ramping

Day-ahead Energy

Over-generation

Resource Adequacy

Opportunities for DR to Provide Bulk Power System Services & Integrate VG Resources

14Source: Cappers et al. Energy Policy. 2012.

Bulk Power System Service DLC

Emergency DR Capacity Energy

Ancillary Services

Spinning Reserves

Supplemental Reserves

Regulation Reserves

Imbalance Energy

Hour-ahead Energy

Multi-hour Ramping

Day-ahead Energy

Over-generation

Resource Adequacy

Currently not offered and unlikely to be offered in the future

Currently not offered or offered only on a very limited basis but could be offered more in the future

Currently offered on a limited basis and could be expanded in the future

Currently offered on a wide-spread basis and likely to be continued in the future

Variable GenerationIntegration Issue TOU CPR CPP DA-RTP RT-RTP

1 min. to 5 – 10 min. variability

<2 hr. forecast error

Large multiple hour ramps

>24 hr. forecast error

Variation from avg. daily energy profile

Avg. daily energy profile by season

Variable GenerationIntegration Issue DLC

Emergency DR Capacity Energy

Ancillary Services

1 Min. to 5 – 10 Min. Variability

<2 hr Forecast Error

Large Multi-hour Ramps

>24 hr Forecast Error

Variation from Avg. Daily Energy Profile

Avg. Daily Energy Profile by Season

Market and Policy Barriers for DR as Ancillary Service Provider

15Source: Cappers et al. Energy Policy. 2013.

Bulk Power System Service

Definitions

Revenue Availability

Attributes of Performance

Enabling Infrastructure

Program Providers

Revenue Capture

Change Definition

Change Require-

mentChange Process

Reduce Costs

Increase Benefits

Bulk Power System Service Definitions

Attributes of Performance

Enabling Infrastructure Investments

Revenue Availability

Revenue Capture

Program Providers

- Primary action to overcome barrier - Secondary action to overcome barrier

Reliability Council BA IOU ARC

Utility Regulator

End-use Customer

Bulk Power System Service Definitions ,

Attributes of Performance

Enabling Infrastructure Investments

Revenue Availability

Revenue Capture

Program Providers

- Entity/Organization responsible for creating the barrier - Entity/Organization affected by the barrier

Opportunities and Challenges with Using DR In Distribution System Planning and Operations

16Source: Cappers et al. LBNL Report. 2015.

Distribution System Service

Max Capacity Relief

Emergency Load Transfer

Voltage Management

Outage Recovery

Power Quality

Phase Balancing

TOU

CPP

DA-RTP

RT-RTP

Disconnectable

Configurable

Manual

Behavioral

Current DR Signals Lack Geographic Specificity

Distribution System Service

Max Capacity Relief

Emergency Load Transfer

Voltage Management

Outage Recovery

Power Quality

Phase Balancing

TOU

CPP

DA-RTP

RT-RTP

Disconnectable

Configurable

Manual

Behavioral

Add Geographic Specificity to DR Signal

Ineffective at providing distribution system

service

Reasonably effective at providing distribution

system service

Highly effective at providing distribution

system service

Topic 4. Policy Analysis and Technical Assistance

Research seeking to understand the role demand response rate and program opportunities can play in

managing the integration of variable generation resources at the bulk and distribution system levels

17

State and Federal DR Policy Assistance LBNL conducts technical analysis & advises states and the federal

government on DR rate and program design typically linked to our research, as presented earlier

Areas include: time-based rate and incentive-based program design, consumer engagement, AMI deployment and regulatory treatment of costs and benefits, and smart meter health effects

Examples: New York time-based rate pilots and consumer engagement Michigan demand response potential studies Ohio AMI and consumer engagement

Regularly brief policy-makers on our work: e.g., NGA, OMB, CEA, CEQ

18Environmental Energy Technologies Division

18

FERC-DOE National Forum on DR

19Source: Goldberg and Agnew. FERC-DOE Report. 2013.

AMI Deployment Issues

20

1. The cost effectiveness or business case for advanced meters is typically justified through reductions in system operating costs or expenses.

2. Internal utility meter reader labor issues have typically been resolved through well planned transition strategies.

3. Adverse customer acceptance issues typically result from inadequate advance customer education, notice, and data validation.

4. Advanced meters appear to be delivering reliable, accurate performance. Meter data applications are beginning to yield additional benefits not included in original business cases.

5. Smart meters with integrated home area networks (HANs) have technical and regulatory policy issues which challenge expectations.

6. Technical options are available to support demand response, pricing, and renewable integration.

Smart Metering1. Customer engagement in regulated environments is

not well understood primarily because “benefits” are typically derived from policy and technology decisions directed at the utility system, rather than the customer.

2. Customer engagement and education are inseparable long-term propositions. There is no single solution or quick fix.

3. Energy savings and educational expectations for in-home displays (IHDs) and access to near real-time meter data are not necessarily supported by existing research

Consumer Engagement

1. Cyber security is an unresolved complex, evolving multi-jurisdictional effort to protect the electric system physical assets.

2. Privacy is a complex effort to protect the information gathered in support of electric service delivery.

3. Cyber security risks can impact all grid-related technology and communication options.

Security and Privacy

Source: Levy. LBNL Presentation. 2013.

Conclusions

21

The Value of and Audiences for Our Work Are Multifaceted

Diverse product types Direct assistance to policymakers, on request Foundational data collection Rigorous analysis of underlying data Other selected research efforts where a need exists

Diverse audiences: from international regulators to local policymakers, and from utility managers to academics

Three over-riding goals Stay nimble to be responsive to emerging issues Maintain a mix of “foundational” and “intellectual” work Emphasize rigor, objectivity, and independence

22

23

Questions?

Peter Cappers315-637-0513 - PACappers@lbl.gov

To hear more about our work:

• visit our homepage: http://emp.lbl.gov

• Follow us on twitter: @BerkeleyLabEMP

• Sign up to our email list: https://emp.lbl.gov/join-our-mailing-list

Thanks to our funders at the U.S. Department of Energy: Office of Electricity Delivery and Energy Reliability and Office of Energy

Efficiency and Renewable Energy