southern california edison's 2012 demand response load ... · this report summarizes the load...

Southern California Edison's 2012 Demand Response

Load Impact Evaluations Portfolio Summary

Prepared for: Southern California Edison

May 30, 2013

Prepared by:

Stephen S. George, Ph.D. Josh A. Schellenberg, M.A.

Candice A. Churchwell, M.S.

Freeman, Sullivan & Co.

101 Montgomery St., 15th Fl San Francisco, CA 94104

fscgroup.com

The FSC Group

i

Table of Contents

1 Introduction .......................................................................................................... 1

2 Overview of Demand Response Programs .................................................................. 3

2.1 Emergency Programs ....................................................................................... 3

2.1.1 Base Interruptible Program ......................................................................... 3

2.1.2 Agricultural and Pumping Interruptible Program............................................. 3

2.1.3 Summer Discount Plan - Commercial ............................................................ 4

2.2 Price-responsive Programs ............................................................................... 4

2.2.1 Summer Discount Plan - Residential ............................................................. 4

2.2.2 Critical Peak Pricing.................................................................................... 4

2.2.3 Demand Bidding Program ........................................................................... 5

2.3 Demand Response Aggregator -managed Programs ............................................. 5

2.3.1 Capacity Bidding Program ........................................................................... 5

2.3.2 Demand Response Contracts ....................................................................... 5

2.4 SmartConnect-enabled Programs ...................................................................... 6

2.5 Non-event Based Programs .............................................................................. 6

2.6 Program Enrollment ......................................................................................... 6

3 Methodology .......................................................................................................... 7

3.1 Selection of 1-in-2 and 1-in-10 Weather Years .................................................... 9

3.2 Overview of Evaluation Methods ...................................................................... 10

3.3 Program Specific Analysis Methods .................................................................. 11

4 Ex Post Load Impact Estimates .............................................................................. 16

4.1 Summary of 2012 Events ............................................................................... 16

4.2 Event Averages by Program ............................................................................ 19

5 Ex Ante Load Impact Estimates .............................................................................. 20

5.1 Projected Change in Portfolio Load Impacts from 2013–2023 ............................. 20

5.2 2015 Portfolio Aggregate Load Impacts by Month .............................................. 21

5.3 Portfolio Load Impacts by Program Type .......................................................... 22

5.4 Portfolio Load Impacts by Program .................................................................. 23

6 Recommendations ................................................................................................ 27

6.1 Emergency Programs ..................................................................................... 27

6.2 Price-responsive Programs ............................................................................. 28

6.3 Aggregator-managed Programs ....................................................................... 28

6.4 SmartConnect-enabled Programs .................................................................... 29

6.5 Non-event Based Programs ............................................................................ 29

Appendix A Regression Specifications ........................................................................ 30

ii

A.1 Base Interruptible Program .......................................................................... 30

A.2 Agricultural and Pumping Interruptible Program ............................................. 31

A.3 Summer Discount Plan – Commercial ............................................................ 32

A.4 Summer Discount Plan – Residential ............................................................. 33

A.5 Critical Peak Pricing .................................................................................... 35

A.6 Demand Bidding Program ............................................................................ 38

A.7 Capacity Bidding Program and Demand Response Contracts ............................ 39

A.8 Save Power Day ......................................................................................... 40

A.9 Real Time Pricing ........................................................................................ 41

Appendix B Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-2 System

Conditions by Month and Forecast Year ..................................................... 43

Appendix C Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-10 System


Appendix D Program Specific Aggregate Ex Ante Load Impact Estimates for

1-in-2 System Conditions by Month and Forecast Year ................................ 53

Appendix E Program Specific Ex Ante Load Impact Estimates for 1-in-10 System


1

1 Introduction

This report summarizes the load reduction capability from Southern California Edison’s (SCE) portfolio

of Demand Response (DR) programs. It details the load impacts from 2012 events (ex post impacts)

and load reduction capabilities for 2013 through 2023 under 1-in-2 and 1-in-10 system conditions (ex

ante impacts). The report adheres to the April 8, 2010 decision by the California Public Utilities

Commission (CPUC) that requires a DR portfolio summary and specifies the format and content of

the summary.1

The 13 DR resources listed in Table 1-1 are summarized in this report. Two programs listed in the

CPUC decision are not included in this report. Optional Binding Mandatory Curtailment (OBMC) is a

program of last resort, triggered immediately prior to rolling blackouts and is not considered a DR

program by SCE. The Scheduled Load Reduction Program (SLRP) is also not included because there

are no participants in the program and no enrollments are projected.

Table 1-1: Summary of Programs and Categorization

Emergency Price-responsive

Demand Response

Aggregator-managed

SmartConnect-enabled

Non-event Based

Base Interruptible Program with 15-minute advance notice (BIP-15)

Summer Discount Plan - Residential (SDP-R)

Capacity Bidding Program with Day-ahead Notification (CBP-DA)

Save Power Day (SPD)

Real Time Pricing (RTP)

Base Interruptible Program with 30-minute advance notice (BIP-30)

Default Critical Peak Pricing (CPP)

Capacity Bidding Program with Day-of Notification (CBP-DO)

Agricultural and Pumping Interruptible Program (AP-I)

Demand Bidding Program (DBP)

Aggregator Demand Response Contracts with Day-ahead Notification (DRC-DA)

Summer Discount Plan - Commercial (SDP-C)

Aggregator Demand Response Contracts with Day-of Notification (DRC-DO)

For all other DR resources, this report summarizes the 2012 program evaluations filed by SCE on April

1, 2012 in accordance with the Load Impact Protocols. Specifically, the contents of the following

reports are summarized:

George, Schellenberg, Oh and C. Hartmann. Load Impact Estimates for SCE’s Demand Response Programs: Agricultural and Pumping Interruptible Program, Real Time Pricing. Final Report. April 1, 2012.

George, Schellenberg and Savage. 2012 Load Impact Evaluation of California’s Statewide Base Interruptible Program. Final Report. April 1, 2012.

1 D.10-04-006

2

Bode, Churchwell and George. 2012 California Statewide Non-residential Critical Peak Pricing Evaluation. Final Report. April 1, 2012.

Braithwait, Hansen and Armstrong. 2012 Statewide Load Impact Evaluation of California Aggregator Demand Response Programs Volume 1: Ex Post and Ex Ante Load Impacts. Final

Report. April 1, 2012.

Hansen, Braithwait and Armstrong. 2012 Load Impact Evaluation of California Statewide Demand Bidding Programs (DBP) for Non-residential Customers: Ex Post and Ex Ante Report. Final Report. April 1, 2012.

Braithwait, Hansen and Hilbrink. 2012 Load Impact Evaluation Southern California Edison’s Peak Time Rebate Program. Final Report. April 1, 2012.

Braithwait, Hansen and Hilbrink. 2012 Load Impact Evaluation of Southern California Edison’s

Residential Summer Discount Plan (SDP) Program. Final Report. April 1, 2012.

Hanna, Elliot and Du. 2012 Impact Evaluation of Southern California Edison’s Commercial Summer Discount Plan (SDP-C). Ex Post and Ex Ante Report. April 1, 21012.

Ex post results are summarized for all programs that experienced an event in 2012. Ex post load

impacts determine what happened over some historical period, based on the conditions that were in

effect during that time. Because historical performance is tied to past conditions such as weather,

price levels and dispatch strategy (e.g., localized dispatches), ex post load impacts may not reflect

the full option value of a DR resource.

Ex ante load impacts are summarized for each program and for SCE’s DR portfolio as a whole.

Portfolio impacts summarize the load reduction that can be expected from all of SCE’s DR programs if

jointly dispatched. In other words, they avoid double counting of load impacts from dually enrolled

customers. Ex ante load impacts are forward-looking and are designed to reflect the load reduction

capability of a DR resource under a standard set of conditions that match the market and system

conditions that drive the need for investing in additional capacity – 1-in-2 and 1-in-10 system

peaking conditions.

This report begins with a description of SCE's DR programs summarized within this filing, including

current program enrollment and forecast enrollments that are linked to ex ante impacts. The program

overview section is followed by a summary of the methods employed in analyzing the ex post and ex

ante load impacts for each program. The next two sections summarize the ex post and ex ante results

for each program as well as the portfolio of programs collectively. The final section summarizes the

recommendations from the 2012 program evaluations. Appendix A describes the regression

specifications that were used in modeling customer load for each program evaluation. Appendix B

through Appendix E contain all of the ex ante results that must be included in the portfolio summary.

3

2 Overview of Demand Response Programs

SCE's current programs can be assigned to one of five categories: emergency, price-responsive,

demand response aggregator-managed, SmartConnect-enabled programs and non-event based. In

general, emergency programs are called when operating reserves are limited, either immediately prior

to or during system emergencies. Price responsive programs can be called based on market

conditions defined by market prices, generator heat rates, temperature or other indicators. In

aggregator-managed programs, aggregators contract with commercial and industrial customers and

assist them in delivering load reduction. Each aggregator forms a portfolio of individual customer

accounts and nominates specific accounts for either an existing DR program such as the Capacity

Bidding Program or for meeting contractual load reduction obligations. Non-event based programs

are not dispatchable, but provide incentives for customers to shift or reduce loads during peak periods

through either time-varying prices or explicit incentives. SmartConnect-enabled programs refer to

programs that are tied to SCE's rollout of smart meters.

2.1 Emergency Programs

Emergency programs are called when operating reserves are limited, either immediately prior to or

during system emergencies.

2.1.1 Base Interruptible Program

Each of California’s three major investor-owned utilities (IOUs), including SCE, offer the Base

Interruptible Program (BIP). BIP is a tariff-based, emergency-triggered demand response program

that CAISO can dispatch for system emergencies. The IOUs can also dispatch BIP for local

emergencies or on a test event basis to verify the program’s load reduction capability.

The program can be dispatched both for instances when electricity system demand approaches

installed generation capacity – a resource shortage – or in response to emergencies due to

transmission and generation outages. Customers enrolled in BIP receive incentive payments in

exchange for committing to reduce their electrical usage to a contractually-established level referred

to as the firm service level (FSL). Participants who fail to reduce load to the FSL are subject to a

financial penalty assessed on a kW per hour basis.

BIP at SCE differentiates payment levels based on the timing of the advance notification provided.

Customers can commit to providing load within 15 or 30 minutes of notification. The load impacts

for both options are summarized in this report.

2.1.2 Agricultural and Pumping Interruptible Program

The Agricultural and Pumping Interruptible (AP-I) program provides a monthly credit to eligible

agricultural and pumping customers for allowing SCE to temporarily interrupt electric service to their

pumping equipment during CAISO or other system emergencies. Agricultural and pumping customers

with a measured demand of 37 kW or greater, or with at least 50 horsepower of connected load per

service account, are eligible to participate in the AP-I program. Participating customers must already

be served under an agricultural and pumping rate schedule. When an interruption is deemed

necessary and is allowed under the terms of the tariff, SCE sends a signal to the load control device

installed on a customer’s pumping equipment. The signal automatically turns off the equipment for

4

the entire duration of the interruption event. The number of interruptions cannot exceed 1 per day,

4 per week and 25 per calendar year. The duration of an interruption cannot exceed 6 hours and the

total hours of interruption cannot exceed 40 per calendar month or 150 per calendar year. In

exchange for allowing SCE to interrupt pumping service during emergencies, AP-I customers receive a

monthly credit. For customers on time-of-use (TOU) rates, the credit is based on measured peak and

mid-peak electricity consumption. For customers that are not on a time-of-use rate, the credit is

based on monthly consumption.

2.1.3 Summer Discount Plan - Commercial

The Summer Discount Plan – Commercial (SDP-C) is a central air conditioning direct load control

program for commercial customers. During high system peak hours or emergency conditions, a signal

is sent to control devices that limit the operation of the compressor in the air conditioner. Participants

can elect whether or not to limit the number of maximum events and the degree of the air

conditioning control – the cycling strategy. The basic plan allows SCE to control an air conditioner up

to a maximum of 15 events per summer and up to 6 hours at a time. Participants can agree to allow

an unlimited number of events in exchange for higher incentives, known as the enhanced plan. The

load impacts and enrollment forecasts in this report are summarized across the basic and enhanced

options of the program for commercial customers.

2.2 Price-responsive Programs

The distinguishing feature of price-responsive programs is that they are dispatched based on

economic criteria rather than solely for emergency conditions. SCE has the option of dispatching

these programs when minimum conditions – defined by market prices, generation heat rates,

temperature and other market indicators – are met.

2.2.1 Summer Discount Plan - Residential

The Summer Discount Plan – Residential (SDP-R) program is a central air conditioning direct load

control program for residential customers. SCE began to operate SDP-R as a price-responsive, rather

than an emergency, program in 2012. During high system peak hours, a signal is sent to control

devices that limit the operation of the compressor in the air conditioner. The program is available

year round and for all hours of the day. It can be dispatched up to 180 hours per year per participant.

For any given day, air conditioning units can only be controlled up to six hours a day for normal

operations, but can be controlled for a longer period, if needed, under emergency conditions. Like

with the SDP Commercial program, participants can elect the degree of air conditioner control – the

cycling strategy. They also have the option of using a cycling device equipped with customer override

capability, but for a lower incentive. Customers that elect the override capability can override up to

five SDP event days per calendar year for each load control device by physically accessing the device

on the air conditioning compressor.

2.2.2 Critical Peak Pricing

Critical Peak Pricing (CPP) is a dynamic pricing program for commercial and industrial customers.

In 2010, SCE's large customers over 200 kW were defaulted onto CPP. Under the default CPP rate,

higher prices on critical peak days are offset by a reduction in off-peak prices, demand charges or

both. SCE has a 2–6 PM event window on CPP days and only calls events on non-holiday summer

5

weekdays. SCE is committed to a minimum of 9 events and a maximum of 15 events each year. In

2012, only large customers with peak demands exceeding 200 kW received service under CPP except

for some voluntary small and medium business customers.

2.2.3 Demand Bidding Program

The Demand Bidding Program (DBP) is a voluntary demand buy-back program that provides enrolled

customers with the opportunity to receive financial incentives as payment for load reductions on event

days. The program is designed to allow commercial and industrial facilities to provide load reduction

without firm commitments or participant risk. Because a firm commitment is not required,

participants can decide whether or not to bid in load reduction on an event-by-event basis. As such,

the mix of event participants (versus enrollment) and magnitude of load reduction may vary from

event-to-event.

2.3 Demand Response Aggregator-managed Programs

Technically, aggregator-managed programs are also price responsive resources, but they are

given a separate category because customers typically are not directly enrolled with the utility. In

aggregator-managed programs, aggregators contract with commercial and industrial customers and

assist them in delivering load reduction. Each aggregator forms a portfolio of individual customer

accounts and nominates specific accounts for either an existing demand response program such as the

Capacity Bidding Program (CBP) or for meeting contractual load reduction obligations. The aggregator

assumes responsibility for managing relationships with individual customers, arranging for load

reductions on event days, receiving incentive payments and paying penalties (if warranted) to

the utility. SCE currently has two aggregator managed programs: CBP and Demand Response

Contracts (DRC).

2.3.1 Capacity Bidding Program

CBP is a statewide program that provides aggregators with monthly capacity payments ($/kW) based

on load reduction commitments for each month, plus additional energy payments ($/kWh) based on

actual electricity demand reductions during events. Each month, aggregators may adjust the

nominated load reduction, the mix of customers that provide load reduction and event options (e.g.,

day-ahead or day-of events, and four-hour, six-hour or eight-hour event lengths). CBP events may

be called on non-holiday weekdays in the months of May through October, between the hours of 11

AM and 7 PM. CBP day-ahead (CBP-DA) and day-of (CBP-DO) resources are summarized separately

in this report.

2.3.2 Demand Response Contracts

DRC is very similar to the CBP program. The primary difference is that the contracts are individually

negotiated and span a longer period of time over which load reduction resources ramp up to

contractual levels. Like CBP, aggregators contract with commercial and industrial customers to act

on their behalf with respect to all aspects of the program, including receiving notices from the utility,

arranging for load reductions on event days, receiving incentive payments and paying penalties to the

utility (if warranted). Each aggregator forms a portfolio of individual customer accounts so that their

aggregated load participates in the DR programs and penalty risk is mitigated. DRC day-ahead (DRC-

DA) and (DRC-DO) day-of resources are summarized separately in this report.

6

2.4 SmartConnect-enabled Programs

Ex post and ex ante load impact estimates are provided for one program in the SmartConnect enabled

category, which is a segment of demand response programs tied to SCE's rollout of smart meters.

Save Power Day (SPD) is a peak time rebate program for residential customers. In 2012, residential

customers with smart meters were defaulted onto SPD. Customers on the program receive a rebate

for reducing load during peak periods when events are called. Customers who do not reduce load

during peak periods when events are called are neither rewarded nor penalized.

2.5 Non-event Based Programs

Non-event based programs provide load reduction or load shifting on a daily basis, but are not

dispatchable. They provide incentives for customers to shift or reduce loads during peak periods

through either time-varying prices or explicit incentives. One non-event based program, real time

pricing (RTP), is summarized in this report.

RTP is a dynamic pricing tariff that charges participants for the electricity they consume based on

hourly prices that vary according to day type and temperature. It attempts to incorporate time-

varying components of energy costs and generation capacity costs. The RTP tariff consists of nine

hourly pricing profiles that vary by season, day type and daily maximum temperature as measured by

the Los Angeles Civic Center weather station. The tariff is available to large commercial and industrial

customers. Because the rate schedules are linked to variation in weather, participants experience

higher prices on hotter days and a greater number of high-price days during extreme weather years

than in normal weather years.

2.6 Program Enrollment

Table 2-1 summarizes the SCE DR enrollment forecasts for 2013 to 2023. SDP-C is expected to

slightly decline over the forecast horizon due to limited marketing plans. AP-I is forecast to grow

modestly according to historic growth rates. There are no changes forecast for BIP enrollment.

DBP is expected to grow as it has in the past, however, there will be an effort in 2013 to remove

those customers who are enrolled but do not participate. These removals are anticipated to be slightly

offset by a 15% participant recapture rate. Additional enrollment growth is anticipated due to opening

up the program to customers with less than 20 kW in demand. The other price-responsive programs

are also expected to grow slightly: CPP should see improved future retentions due to a new CRL

offering and SDP-R should also gain enrollments due to a planned inbound call solicitation campaign.

CBP enrollments will decline due to aggregators moving customers to DRC, and DRC is anticipated to

grow as new contracts with aggregators are signed. RTP enrollment is expected to remain steady,

while SPD enrollments in event notification are forecast to top out near 1,000,000 but is expected to

decline as customers opt-out of notification.

7

Table 2-1: SCE DR Portfolio Projected Enrollments for 2013-2023 by Program (Values reflect expected enrollment in August)

Program Type Program

Forecast Year

2013 2014 2015 2016 2017-2023

Emergency

BIP-15 79 79 79 79 79

BIP-30 568 568 568 568 568

AP-I 1,157 1,209 1,264 1,282 1,282

SDP-C 9,311 9,162 9,016 8,945 8,945

Price-responsive

SDP-R 302,347 306,429 310,567 311,959 311,959

CPP 3,068 3,099 3,130 3,140 3,140

DBP 1,375 1,022 1,611 2,927 3,428

Demand Response Aggregator-managed*

CBP-DA 3 3 3 2 2

CBP-DO 255 250 245 243 243

DRC-DA 123 135 149 153 153

DRC-DO 1,468 1,616 1,778 1,835 1,835

SmartConnect-enabled SPD 950,000 941,192 850,763 817,073 817,073

Non-event Based RTP 134 136 139 140 140

Portfolio Total 1,269,888 1,264,900 1,179,312 1,148,346 1,148,847

8

3 Methodology

The 2012 evaluations address two main questions for DR programs: What demand reductions were

delivered when resources were dispatched in 2012? And, what is the load reduction capability of each

DR program?

Ex post impacts reflect the demand reductions attained during actual events, but do not necessarily

reflect the load reduction capability of the DR program. Historical ex post results are tied to specific

conditions that occurred for that given event, including weather conditions, the number of participants

who were dispatched, the mix of customers and other factors such as switch failure rates. Several

programs are dispatched strategically to address congestion in specific zones, test load response

capabilities or for economic reasons. Due to the lack of extreme weather in 2012, emergency

resources such as BIP were only dispatched to test load reduction capabilities. Other resources, such

as SDP were never dispatched in full; specific regions were dispatched for each event. In addition, the

timing and duration of the dispatch varied across event days for many programs. As a result, the

impacts for individual event days are not necessarily representative of the program.

Ex ante impacts reflect the load reduction capability of a DR program for each month under a standard

set of 1-in-2 and 1-in-10 weather conditions. They reflect the reduction that can be attained if all

enrolled participants are dispatched under the weather conditions that drive system planning.

Whenever possible, ex ante load impacts are grounded on analysis of historical load impact

performance. These estimates are used in assessing alternatives for meeting peak demand,

cost-effectiveness comparisons and long term planning.

Figure 3-1 shows the connection between ex post load impacts, ex ante impacts, cost-effectiveness

analysis and resource planning. Analysis of historical program data is then employed to produce ex

ante load impact estimates that are subsequently used for resource adequacy, cost-effectiveness

assessment and, by connection, resource planning.

9

Figure 3-1: Summary of Ex Post and Ex Ante Analysis Process and Connections

3.1 Selection of 1-in-2 and 1-in-10 Weather Years

The selection of 1-in-2 and 1-in-10 weather years was the same as in the 2011 load impact

evaluations. In order to better align the weather with the primary applications of the load impact

estimates – long term planning, resources adequacy and cost-effectiveness – the selection of the

1-in-2 and 1-in-10 monthly system peak weather conditions was based on an analysis of recent

system load data and 20 years of weather data from 25 weather stations located throughout the

SCE territory. The process consisted of the following steps:

Develop a demand model that estimates system load (using recent system load data) as a function of weather conditions, hour of day and seasonal factors;

Predict the system load using the 20 years of historical weather conditions;

Identify the days on which monthly system peak loads were estimated for each month of each year;

Rank the monthly system peak load for each month;

Identify the 50th and the 90th percentile monthly system peaks (i.e., 1-in-2 and 1-in-10 weather year conditions); and

Select the weather associated with the selected monthly peak days as the 1-in-2 and 1-in-10

year weather conditions.

The analysis relied on a demand model rather than on historical system peak data for several reasons.

Central air conditioning saturation, population centers, industry, building and appliance codes changed

substantially over the 20-year span. By relying on a system demand model, demand estimates given

-Event days

-Weather

-Participant

characteristics

Interval data

(sample or

population)

Statistical

Analysis of

historical data

Statistical

Analysis of

historical data

Evaluation

planning/ goals

Methodology

-Regression

-Day matching

- Other

Ex-post load

impacts

Ex-post load

impacts

Ex-ante

impact

estimates

Ex-ante

impact

estimates

Cost-

Effectiveness

Tests

Cost-

Effectiveness

Tests

DR costs

DR benefits

Comparison

with other

resources

Comparison

with other

resources

Generation

alternatives

Adjustments

DSM alternatives

Day Types1-in-2 weather year

1-in-10 weather year

Avg. weekday by month

Monthly system peak day

Weather

Participant characteristics

Other – e.g. switch failures

1-in-2 and 1-in-10

-Weather data

- System Load Data

- Day traits

Participation

Forecasts

Participation

ForecastsMeasurement

& Verification

Studies

Measurement

& Verification

Studies

10

current drivers of system load and known historical weather variation were produced. The time

span includes a diverse set of weather conditions, enabling the prediction of system load for extreme

weather conditions. Once developed, the demand model was applied to the historical weather for the

20-year time span in order to identify extreme and normal conditions for monthly system peaks.

3.2 Overview of Evaluation Methods

The methods used to estimate ex post and ex ante load impacts for each of the DR programs in the

SCE portfolio are conceptually similar. Each of the 2012 evaluations relied on regression analysis to

estimate a model reflecting the relationship between customer or end-use load and key determinants

of the variation in energy use over time, such as weather and time-of-day, day-of-week and seasonal

patterns that reflect the normal pattern of business or household operations. These models are based

on historical hourly or sub-hourly electricity use data for customers who have participated in the DR

programs. Each model or set of models is used to estimate the reference load for an average

customer enrolled in a program, which represents what customers would be expected to use in

the absence of an event on days in which program events either were called (for ex post impact

estimation) or have a high probability of being called (for ex ante impact estimation). For the single

non-event based program (RTP), the methods were slightly different. RTP reference loads represent

what the average customer would use on a specific day if they faced the otherwise applicable tariff,

TOU-8, rather than the RTP tariff.

In most instances, ex post impacts were estimated by comparing the reference level energy use in

each hour with the estimated load with DR in the hour on each event day. For ex ante estimation,

predicted energy use in each hour was estimated under the assumption that an event occurred and

also under the assumption that it did not occur, while everything else (e.g., weather, day-of-week

effects) was held constant at values representative of a typical event day or monthly system

peak day.

At a more technical level, two general approaches were used to estimate the regression models:

Individual Customer Time Series Regressions: This method works well for event-based programs with numerous events (e.g., CPP) and for programs with substantial variation in the drivers of load response or load shifting (e.g., RTP). This approach is also useful for programs with substantial differences in the magnitude and load patterns of customers, which is more typical among large customers. The coefficients vary at the customer level. While the regressions do not necessarily explain individual customer behavior perfectly, in aggregate, they explain most of the program level variation in loads. Importantly, individual customer

regressions can be employed to describe the distribution of customer load reductions as well as the distribution of percent load reductions. They can also be used to describe impacts for segments of the participant population. The key limitation to individual customer regressions is their inability to make use of control groups.

Aggregate Time Series Regressions: Similar to the individual customer regression approach, but rather than estimating reference loads and load impacts for individual customers,

estimates are made for groups of customers in taken in aggregate.

Panel Regressions: This method is particularly suitable when equivalent control groups are available, or sample sizes are sufficient for the territory, but inadequate for smaller segments such as local capacity areas. A key strength of panel regressions is the ability to control for certain omitted or unobservable variables.2 While panel regressions can increase the accuracy

2 Panel regressions can account for omitted variables that are unique to customers and relatively time invariant over the

analysis time frame (fixed effects) such as household income. Panel regressions can also account for omitted variables

11

of impact estimates for the average customer, they cannot be employed to describe the distribution of impacts among the participant population. Importantly, panel regressions cannot control for customer characteristics that interact with occupancy and or weather unless those variables are explicitly included.

The regression models used to predict the reference load were developed with the primary goal of

accurately predicting average customer load given the time of day, day of week, temperature and

location of each customer and predicting load reductions under different temperature conditions. The

focus was on the accuracy of the prediction and the validity of load impact estimates. The regression

equations used to model load patterns and estimate load impacts for each program are detailed in

Appendix A.

3.3 Program Specific Analysis Methods

Table 3-1 summarizes the analysis methodology for each program. It describes the general approach

used for load impact estimation and details any key assumptions required in the analysis. The specific

methodology chosen for each program was based on the available data, event dispatch patterns and

the strengths and weakness of each regression approach.

that are common across the participant population but unique to specific time periods (time effects). They cannot,

however, account for omitted variables that vary both by participant and by time period or for household characteristics

(e.g., central air conditioning) that interact with variables that vary over time, such as weather and occupancy.

12

Table 3-1: Summary of Analysis Methodologies by Program

Program Method Evaluation Description Key Assumptions

Baseline Interruptible Program

(BIP-15 and BIP-30)

Regression models - individual customer

Individual load patterns were modeled using hourly data from 2011–2012 for all participants with available data. Ex ante impacts were estimated as the reference load under 1-in-2 and 1-in-10 system peak conditions minus the firm service level, with adjustments based on historical over or under performance.

Customers will continue to perform relative to their FSL in the future as they have in the past

Participant load is expected to increase 1.5% annually through 2014.

Enrollment growth is expected to remain flat

Agricultural Pumping Interruptible Program (AP-I)

Regression models – individual customer

Agricultural pump loads were modeled as a function of time of day, day of week, temperature and other factors. Estimates of switch activation success rates were developed based on the 2012 test event and applied to reference loads in the ex ante analysis.

Pump loads are fully shut down when switch activation is successful

Switch activation success rates are assumed to improve through 2014 due to an effort to identify and fix communication and switch failures

Enrollment is projected to grow as it has in the past - 4.5% annually

Summer Discount Plan - Commercial (SDP-C)

Ex post: Regression models – individual customer

Ex post load impacts, using premise-level data for the single 2012 event, could only be estimated using a convenience sample of customers with smart meters or legacy load research meters. These impacts were estimated with hourly load models with variables controlling for weather, day-type and Flex Alerts. Ex ante load impacts were developed with end-use reference loads estimated by panel regressions in the 2010 load impact evaluation. Percentage load impacts were borrowed from the 2010 SDG&E air conditioning program ("Summer Saver") load impact evaluation.

SDP-C 2012 ex post load impacts are not representative of impacts that the program can expect in the future

SDG&E's air conditioning cycling program load impacts are representative of SCE SDP-C load impacts

Ex ante: Regression models – panel

Enrollment is expected to decline by 1.6% annually due to limited marketing and a transition to price-responsiveness programs in 2013

13


Summer Discount Plan - Residential (SDP-R)

Regression models – aggregate

Reference loads and impacts were analyzed for customer groups by A-bank and cycling strategy. Ex post reference load and impacts formed the basis for the summer ex ante load impacts. The SPD sample was used to develop a winter proxy reference load.

The program began, in 2012, to operate as a price-responsive demand response program after previously operating as a reliability program

SDP-R load impacts are assumed to be zero when the temperature is below 70 degrees Fahrenheit.

Enrollment is expected to grow 1.3% annually

Default Critical Peak Pricing (CPP)

Ex post: Regression models – panel

The CPP ex post hourly load impacts for program year 2012 were estimated with a difference-in-differences panel regression using a control group. Ex ante load impacts were estimated using individual regression models fit for ex post conditions. The ex ante load impacts factored in historic performance over the period 2010 through 2012 of all customers enrolled as of the last 2012 CPP event.

Future load impacts for each customer will be similar to historical performance over 2010 through 2012

Ex ante: Regression models –individual

Minimal growth in enrollment (1% annually) due to anticipated CRL offering

Demand Bidding Program (DBP)


Ex post hourly load impacts were estimated using regression equations applied to customer-level hourly load data. Ex ante load impacts were estimated using percentage load impacts directly calculated from 2010–2012 ex post results and applied to 1-in-2 and 1-in-10 weather reference loads. Program-level load impacts are significantly higher than portfolio-level load impacts in all forecast years due to dual enrollment in BIP or DRC.

Future bidding behavior will be similar to current bidding behavior

SCE will begin permitting customers with maximum demand less than 200 kW to enroll in DBP in 2015

SCE will begin removing non-performing customers from the program in 2014

14


Capacity Bidding Program (CBP-DA and CPB-DO)


Direct estimates of total program level ex post load impacts for each program were developed from the coefficients of individual customer regression equations for customers enrolled in CBP in 2012. The ex ante estimates factored in historical performance from 2010 and 2012 events for each customer enrolled in the program at the end of the 2012 cycle.

Future load impacts for each customer will be similar to historical performance in 2010 and 2012

Customer mix will be similar to that of the 2012 participants

0% growth for DA option as aggregators move customer to DO option; -2.1% growth in DO option as aggregators move customers to DRC

Demand Response Contracts (DRC-DA and DRC-DO)


Direct estimates of total program level ex post load impacts for each program were developed from the coefficients of individual customer regression equations for customers enrolled in DRC in 2012. The ex ante estimates factored in historical performance from 2010 and 2012 events for each customer enrolled in the program at the end of the 2012 cycle.

Future load impacts for each customer will be similar to historical performance in 2010 and 2012

Customer mix will be similar to that of the 2012 customer mix

9% annual growth in enrollment

Save Power Day (SPD) Regression models – aggregate

A sample (n = 400,000) of non-SDP-R SPD customers who had elected to receive electronic event notification was selected. Hourly load data for sampled customers was aggregated into four groups (by climate zone and notification type). Hourly load for all SDP-R was aggregated into three groups (by notification type). Regression models were estimated for all aggregate groups. Ex ante results were only produced for notified customers, and were segmented by type of notification, LCA and dual enrollment with SDP.

Enrollment forecasts reflect an SPD (default) enrollment goal of 1,000,000 customers by September 2013, but opt-offs are expected in subsequent summers

Growth in the number of customers who opt-in to receive event notification is not expected to grow substantially

15


Real Time Pricing (RTP) Regression models – individual customer

Customer load was modeled as a function of time of day, day of week, weather (for some customers) and hourly price schedules using 2011-2012 hourly data. The impacts were estimated as the difference between customer loads under RTP and estimated hourly loads under the otherwise applicable tariff prices based on individual customer price response.

Customers will continue to respond to prices as they have in the past

New participants are expected to be smaller than the average existing participant.

Enrollment is projected to grow as it has in the past - 2% annually

RTP will be available to TOU-8 customers

Future RTP and TOU-8 rates will be similar to present rates

16

4 Ex Post Load Impact Estimates

This section summarizes the load impacts in 2012 for event-based programs. Ex post load impacts

are based on modeling electricity use patterns and load impacts over a historical period. They

estimate what happened, based on the conditions that were in effect during that time. While historical

load patterns and impacts are critical to understanding the magnitude of load reduction resources,

they have limitations. Because historical performance is tied to past conditions such as weather, price

levels and dispatch strategy (e.g., localized dispatches), ex post load impacts may not reflect the full

option value of a DR resource. For example, a test event for a highly weather sensitive program such

as SDP-C may yield lower impacts than what the program can provide because future events might

occur at hotter temperatures when air conditioning loads are higher. Likewise, resources such as CBP

or DRC may be dispatched partially – one product line is called – in which case ex post events do not

necessarily reflect the program load reduction capability.

4.1 Summary of 2012 Events

In 2012, SCE DR resources were dispatched based on program rules and need. The event days and

event hours differed across programs and, sometimes, within programs. Table 4-1 summarizes the

events called in 2012 by date and program. RTP is omitted because it is not an event-based program.

CBP and CPP were dispatched most frequently of the event-based programs.

As noted earlier, several programs are dispatched strategically to address congestion in specific zones,

test load response capabilities or for economic reasons. CBP, DRC and SDP were never dispatched in

full. For SDP, SCE engaged in localized testing of resources. Given the sensitivity of air conditioning

loads to weather, the ex post events are not representative of the SDP population nor representative

of the weather during peaking conditions. For DRC and CBP, different combinations of program

products and/or aggregators (if applicable) were dispatched for each individual event. As a result, the

impacts for individual event days are not necessarily representative of the resources available should

SCE solicit demand reductions from all aggregator resources at once.

17

Table 4-1: Summary of 2012 Events by Date and Program

Date BIP AP-I SDP-C SDP-R CPP DBP CBP-DA CBP-DO DRC-DA DRC- DO SPD

29-Jun

14:00 - 18:00

10-Jul

13:00 - 16:00

12-Jul

14:00 - 18:00 12:00 - 20:00

14:00 - 18:00

20-Jul

16:00 - 19:00

23-Jul

14:00 - 18:00

14:00 - 18:00

24-Jul

14:00 - 18:00

25-Jul

15:00 - 17:00

30-Jul

14:00 - 18:00

31-Jul

14:00 - 18:00

1-Aug 15:00 - 17:00

18:00 - 19:00

3-Aug

15:00 - 18:00

7-Aug

14:00 - 18:00

13:00 - 17:00

8-Aug

15:00 - 18:00

12:00 - 20:00

9-Aug

15:00 - 18:00 14:00 - 18:00

10-Aug

12:00 - 20:00

14:00 - 18:00

13-Aug

14:00 - 18:00

13:00 - 17:00

14-Aug

15:30 - 21:23 15:50 - 21:27 15:00 - 21:27

12:00 - 20:00

13:00 - 17:00 15:00 - 17:00 15:00 - 17:00

15-Aug

15:00 - 18:00

16-Aug

12:00 - 20:00

14:00 - 18:00

17-Aug

16:00 - 18:00

20-Aug

14:00 - 18:00

21-Aug

15:00 - 18:00

18

Date BIP AP-I SDP-C SDP-R CPP DBP CBP-DA CBP-DO DRC-DA

DRC- DO

SPD

22-Aug

15:00 - 18:00

27-Aug

14:00 - 18:00

28-Aug

15:00 - 18:00

29-Aug

15:00 - 18:00 14:00 - 18:00 12:00 - 20:00

14:00 - 18:00

31-Aug

14:00 - 18:00

7-Sep

14:00 - 18:00

10-Sep

15:00 - 18:00 14:00 - 18:00

14:00 - 18:00

14-Sep

15:00 - 18:00

13:00 - 19:00

20-Sep

15:00 - 18:00 14:00 - 18:00

21-Sep

15:00 - 18:00

26-Sep 15:00 - 17:00 14:50 - 16:00

28-Sep

15:00 - 18:00 14:00 - 18:00

1-Oct

12:00 - 20:00

2-Oct

14:00 - 18:00

13:00 - 17:00 14:00 - 18:00

3-Oct

14:00 - 17:00

5-Oct

15:00 - 17:00

7-Oct

12:00 - 20:00

17-Oct

15:00 - 18:00

14:00 - 17:00

18-Oct

14:00 - 18:00

14:00 - 18:00 13:00 - 19:00

26-Oct

15:00 - 18:00

29-Oct

18:00 - 19:00

Total 2 3 2 23 13 9 12 8 2 2 8

19

4.2 Event Averages by Program

Interpreting the average event impact across events can be difficult because multiple factors can vary

across event days, including temperature, the normal pattern of energy use, enrollment, the number

of customers called, dispatch strategy and number of event hours. For programs such as large

customer DBP and CPP with stable participation, fixed event windows, less weather sensitive

customers and universal dispatch for all events, the average event impacts can provide meaningful

and insightful data about program performance. However, for resources that do not have those

characteristics, the average event impacts provide limited insight and can be misleading. In short,

ex post load impacts may not reflect the full option value of a DR resource and should be interpreted

with caution. In the case of AP-I, SDP, DRC and CBP, not only was a subset of customers called

for each event, but the customers called for each event were not necessarily representative of the

overall program.

Table 4-2 summarizes the average event impacts across all events for each of SCE's programs that

had an event in 2012. A total row at the bottom is not provided because these are different types of

programs that were dispatched at different times in 2012, as shown in Table 4-1.

Table 4-2: 2012 Ex Post Load Impacts for the Average Event by Program

Program Reference Load (kW)

Load with DR (kW)

Load Impact per Customer

(kW)

% Load Impact

Aggregate Impact (MW)

Accounts Called

Number of

Events

AP-I 51.2 12.5 38.8 76% 21 533 2

BIP 1,163.1 304.0 859.1 74% 573 667 1

SDP-C 33.3 28.6 4.7 14% 4 811 1

SDP-R 3.1 2.2 0.8 27% 83 303,5033 21

CPP 221.0 207.9 13.1 6% 33 2,508 12

DBP 750.6 690.1 60.5 8% 83 1,369 8

CBP-DA 548.7 530.4 18.3 3% 0 2 12

CBP-DO 243.0 197.1 45.9 19% 17 359 7

DRC-DA 233.4 79.9 153.5 66% 22 142 1

DRC-DO 334.1 236.9 97.2 29% 160 1,648 2

3 SDP-R customers were split into three groups for the curtailment events. A 3-hour event entailed each group being called

for one of the hours. The per-customer results in the table represent the load impacts for the hour in which the group was

called (using only the 1-hour events). The aggregate results are the total load impacts (considering only the current-hour

load reductions from the called group, excluding rebound effects from the groups called in previous hours) for an average

event hour, representing approximately one-third of the impact obtained by multiplying the per-customer load impact by

the total number of customers.

20

5 Ex Ante Load Impact Estimates

The portfolio ex ante load impact estimates summarize the load reduction that can be expected from

all of SCE’s DR programs if called simultaneously. They are based on a common event window and

the weather conditions underlying 1-in-2 and 1-in-10 monthly system peak days. The ex ante

estimates provide estimates of the resources available under conditions that are linked to the

need for investment in additional capacity. The load impact estimates for each program align with

revised resource adequacy hours, 1–6 PM in April through October and 4 PM to 9 PM in November

through March.

Portfolio-adjusted load reductions reflect the assignment of load impacts from dually enrolled accounts

to a single program in order to avoid double counting impacts. The load impacts of customers

enrolled in both an emergency program and a price-responsive program are attributed to the

emergency response program for portfolio-adjusted reporting.4 Although dual participation is allowed

for many of SCE’s DR programs, currently, overlaps are almost exclusively between customers dually

enrolled in BIP and DBP.

The remainder of this section summarizes the ex ante load impact estimates for SCE's portfolio of DR

programs. The discussion focuses on the high level, portfolio aggregate impacts by forecast year,

month and program type. The remainder of the portfolio-adjusted and program-specific estimates

that are required by the Protocols can be found in Appendices B, C, D and E.

5.1 Projected Change in Portfolio Load Impacts from

2013–2023

Figure 5-1 presents the portfolio-adjusted aggregate load impact estimates for the August system

peak day under 1-in-2 and 1-in-10 system conditions by forecast year. The estimated aggregate load

reduction is highest in 2017 and stays constant through the end of the forecast horizon. Under 1-in-2

system conditions, SCE's DR portfolio is projected to grow from 1,246 MW in 2013 to 1,311 MW in

2017 and then remain stable thereafter. Under 1-in-10 system conditions, aggregate load impacts

grow at an average rate of about 2.5% each year until the program load impacts stabilize in 2015. By

2017, SCE's DR portfolio is expected to deliver 1,335 MW for the 1-in-10 August system peak day.

4 For purposes of estimating aggregate load impacts that apply to the cap on emergency DR programs, the allocation rule is

reversed in that the load impacts for dually enrolled customers are attributed to the price responsive program.

21

Figure 5-1: Portfolio Aggregate Load Impact Estimates (MW) for the August System Peak Day By 1-in-2 and 1-in-10 System Conditions and Forecast Year

5.2 2015 Portfolio Aggregate Load Impacts by Month

Figure 5-2 shows how the 2015 portfolio load impacts vary by month under 1-in-2 and 1-in-10 system

conditions. In 2014, SCE's DR portfolio is projected to be capable of delivering up to 1,304 MW of

load reduction during the August monthly system peak day under 1-in-10 system conditions. The

July and September load impacts under 1-in-10 system conditions are quite similar, both of which are

close to 1,300 MW. The portfolio load impacts during non-summer months are substantially lower for

two main reasons:

SDP - Commercial is only available during summer months; and

CBP and DRC are only available to be called from May through October.

Although the portfolio load impacts during non-summer months are substantially lower, it is important

to note that SCE's portfolio of DR programs is maximized from July through September, which are the

months in which a system peak is most likely to occur.

0.0

200.0

400.0

600.0

800.0

1000.0

1200.0

1400.0

1600.0

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Ag

gre

gate

Lo

ad

Im

pa

ct

(MW

)

Forecast Year

1-in-10 System Conditions 1-in-2 System Conditions

22

Figure 5-2: 2015 Portfolio Aggregate Load Impact Estimates (MW) By 1-in-2 and 1-in-10 System Conditions and Monthly System Peak Day

5.3 Portfolio Load Impacts by Program Type

SCE has moved towards a more balanced DR portfolio by program type with fewer emergency

response resources. Figure 5-3 shows the distribution of portfolio aggregate load impacts by program

type at the end of the ex ante forecast horizon, years 2017 through 2023. Load impacts from

emergency response programs are forecast to comprise 59% of SCE's DR portfolio during this period,

a much smaller fraction of the total portfolio than it has been in previous years. Most of the remaining

load impacts are forecast to come from aggregator-managed programs (14%) and price-responsive

programs (25%).

0.0

200.0

400.0

600.0

800.0

1000.0

1200.0

1400.0

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Ag

gre

gate

Lo

ad

Im

pa

ct

(MW

)

Monthly System Peak Day

1-in-10 System Conditions 1-in-2 System Conditions

23

Figure 5-3: Distribution of Portfolio Aggregate Load Impacts by Program Type 2017-2023 August System Peak Day under 1-in-2 System Conditions

5.4 Portfolio Load Impacts by Program

Table 5-1 summarizes the portfolio load impacts by program for 2013-2023 under 1-in-2 system

peak conditions. The following list provides some background behind the aggregate impact results

for each program:

Although BIP enrollment does not change, aggregate load impacts increase over time because it is assumed that BIP customers experience 1.5% load growth per year from 2013 through

2014 and reach a steady state in 2015 through 2023;

AP-I aggregate load impacts increase over time because of enrollment growth and a projected improvement in switch activation success rates;

SDP-Residential aggregate load impacts increase over time because of projected enrollment growth;

CPP aggregate load impacts increase slightly due to an increase in expected participant retention due to the forthcoming Capacity Reservation Level offering;

DBP portfolio impacts are lower than the ex post estimates or ex ante program-specific impacts because of dual enrollment with BIP;

Load impacts for aggregator-managed programs stay relatively constant apart from DRC-DO where program growth is expected;

RTP load impacts reflect low enrollment growth; and

Load impacts for SPD decrease in the outer forecast years due to anticipated opt-outs.

Aggregator-managed

14%

Emergency 59%

Non-event based

1%

Price-responsive

25%


1%

2017-2023 MW = 1,311

24

Table 5-1: Portfolio Aggregate Load Impact Estimates (MW) for the August System Peak Day Under 1-in-2 System Conditions by Program and Forecast Year


Forecast Year

2013 2014 2015 2016 2017-2023

Emergency

BIP-15 146 148 149 149 149

BIP-30 467 475 478 478 478

AP-I 59 65 68 69 69

SDP-C 78 77 76 76 76

SUB-TOTAL 750 765 771 772 772

Price-responsive

SDP-R 291 294 298 298 298

CPP 18 19 19 19 19

DBP 4 4 6 9 10

SUB-TOTAL 313 317 323 326 327

Demand Response Aggregator-managed

CBP-DA 0 0 0 0 0

CBP-DO 11 11 10 10 10

DRC-DA 16 17 19 19 19

DRC-DO 129 142 156 156 156

SUB-TOTAL 156 170 186 186 186

SmartConnect-enabled SPD 20 20 18 18 18

SUB-TOTAL 20 20 18 18 18

Non-event based RTP 7 7 7 7 7

SUB-TOTAL 7 7 7 7 7

PORTFOLIO TOTAL 1,246 1,280 1,305 1,309 1,311

Tables 5-2 and 5-3 summarize the monthly variation in portfolio aggregate load impacts in 2015 for

1-in-2 and 1-in-10 system conditions. Similar tables are available in Appendix B, C, D, and E for each

forecast year from 2013-2023, for 1-in-2 and 1-in-10 system conditions and on both the portfolio-

adjusted and program-specific bases.

25

Table 5-2: 2015 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions


Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 36 37 43 53 67 71 68 68 60 51 39 36

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 629 629 679 673 702 741 758 771 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 25 24 25 24 22 21 19 19 19 21 25 24

DBP 3 2 2 3 4 4 5 6 6 6 3 4

SUB-TOTAL 27 26 27 92 165 251 284 323 317 197 28 151

Demand Response Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0

SmartConnect-enabled SPD 3 3 3 3 12 12 14 18 16 14 3 4

SUB-TOTAL 3 3 3 3 12 12 14 18 16 14 3 4

Non-event Based RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 659 658 708 767 1,043 1,171 1,232 1,305 1,284 1,085 624 811

26

Table 5-3: 2014 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions


Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 39 40 49 56 70 72 68 69 62 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 648 624 662 682 708 751 760 776 772 713 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 25 23 23 21 20 18 41 18 40 21 25 20

DBP 3 3 3 3 4 5 5 6 6 6 4 4

SUB-TOTAL 28 108 222 221 216 300 342 342 323 260 29 218

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 3 5 3 12 14 16 19 13 15 3 4

SUB-TOTAL 3 3 5 3 12 14 16 19 13 15 3 4

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 678 736 895 906 1,108 1,242 1,295 1,330 1,290 1,162 631 879

27

6 Recommendations

The 2012 DR program evaluations contain recommendations for each program. The recommendations

provide steps to improve the measurement and evaluation of DR resources and to improve program

performance. This section summarizes the recommendations for each program. More detailed

explanations of the recommendations can be found in the individual program evaluations.

6.1 Emergency Programs

Overall, emergency programs are characterized by infrequent use, but substantial load reductions are

linked to both automated control technology and contractual agreements with substantial penalties for

non-performance. Their importance and infrequent dispatch make it critical to understand the

electricity use patterns of participants, call test events and measure the extent to which

communications work well. The following summarizes the recommendations for the

emergency programs:

BIP events in 2012 improved the quality of the over/under performance analysis, which in

turn, improved the quality of the ex ante estimates. It is recommended that at least one event is continued to be called each year. When calling a test event, the event conditions that are being attempted to simulate should be taken into consideration. If a BIP test event is meant to simulate a generation supply shortage, we recommend giving at least one day notice, but not the exact timing of the event. If a BIP test event is meant to simulate a transmission or distribution outage, no day-ahead notice should be given.

Continue to improve AP-I switch success rate though the following steps:

1. Run tests or actual events during the summer, when pumps are on. Ideally, the test event would occur during peak hours and last long enough to determine whether pumps that were operating immediately before the event ramped down when the event signal was sent to the switches;

2. Analyze the 15-minute interval data to identify units that were on immediately prior to the event but were not activated. The criteria for determining activation must factor in that

some pumps ramp down over five minutes and that additional loads not controlled by switches are measured by the same meter for a small fraction of participants; and

3. Target the identified accounts for a switch activation inspection and repair, as appropriate.

Calling events facilitates the identification of pumps that are not providing load reduction and improve the switch success rates. Out of necessity, the improvement in switch success rates would be conducted over the course of two or three years. It takes time to call events, identify units that are not providing load reduction, inspect and repair. Moreover, not all units

will be on for a given event due to the variable nature of pump loads. As a result, improving switch success is an iterative process, requiring continuous adjustment to meet stated goals.

2012 SDP-C load impacts are limited by the design of the single event that was called in 2012.

Only participants from within a relatively small area were called, and the event was initiated late in the afternoon and ended fairly late in the day. It is recommended to call SDP-C test events for this reason.

The availability in 2012 of some premise-level data from SmartConnect meters indicates that

there may be SDP-C system performance issues due to control signals not reaching control switches, malfunctioning control switches or disconnected control switches. As more SmartConnect meters are deployed, analyses should be conducted to evaluate switch success.

28

6.2 Price-responsive Programs

Price responsive programs are dispatched more frequently based on economic criteria rather

than solely for emergency conditions. The following recommendations were made for price

response programs:

SDP-R ex post load impact results indicated a pattern of declining net load impacts in the second and third hours of a typical three-hour economically-dispatched SDP event, due to separate groups of participants each being called for one hour. This finding raises a question

about the resource value of each hour’s load impacts, and suggests a possible strategy for smoothing the load impacts across all event hours. The question has to do with the fact that the groups called for each one-hour piece of a three-hour event tend to be somewhat geographically dispersed. Thus, to the extent that the regional value of the load reductions is important, then post-event rebound effects in one region may have less effect on the value of the subsequent load reductions in the other regions than suggested in this study.

However, if the above types of regional effects are not that important, then one recommended

strategy for smoothing out the load impacts across hours would involve reducing the number of customers called in the first hour, increasing the number called in the second hour, and further increasing the number called in the third hour. This strategy would trade off lower load reductions in the first hour for a more constant level of load impacts across a three-hour event. This approach is somewhat analogous to the adaptive strategies that are applied in some air conditioning cycling programs in which the degree of cycling is allowed to increase

over the event period to compensate for the units attempting to increase their natural duty cycle as indoor temperatures rise, which would tend to degrade load reductions over a multi-hour curtailment.

The primary CPP recommendations are to

o Evaluate demand reductions closer to real time by using a control group with a difference-in-differences calculation; and

o Estimate the effect of program changes through research design rather than after-the-

fact analysis. A phased roll-out of program changes such as CRLs and event window

changes in combination with random assignment can lead to better estimates of the effects of those program changes.

Based on the performance of dually enrolled customers, SCE should continue to encourage customers in BIP and the aggregator programs (AMP and CBP) to enroll in DBP. They tend to be the most responsive customers in DBP and provide a means for the utilities to increase the amount of demand response that can be obtained on DBP-only event days.

In addition, the day-of adjustments to the 10-in-10 baselines appear to significantly improve the accuracy of, and reduce the bias in, program baseline performance. The improvements are not very sensitive to the level of the day-of adjustment cap, though there is some evidence that a cap of 20 to 40 percent would strike a reasonable balance between improved performance and limited risk (i.e., preventing extreme adjustments).

6.3 Aggregator-managed Programs

Recommendations for SCE’s aggregator-managed programs (CBP and DRC) include the following:

The 2011 evaluation recommended maintaining the CBP programs due to their useful role in both providing opportunities for aggregators that don’t have bilateral contracts with utilities, and providing aggregators that do have contracts with a pool of customers that they may eventually move to contract-based programs (e.g., AMP and DRC). However, it is worth asking whether programs with extremely small enrollment, such as SCE’s CBP DA, should be maintained as is, or if aggregators should be encouraged to bring in additional customers, even if to transition them to other program options.

29

6.4 SmartConnect-enabled Programs

SPD ex post load impact results suggest that greater event-day load reductions may be achieved

by educating customers, making them more aware of the program and its potential benefits and

developing strategies to encourage them to sign up to receive event notifications.

Results for customers who were defaulted into SPD notification, which indicate small and not

statistically significant estimated load reductions for these customers, suggest that expanding

default notification may have limited benefits (but no detrimental effects) on program performance.

A continuing measurement challenge will be to refine methods for estimating small load impacts

among large groups of customers. Exploring a combination of SPD customer-level load analysis and

customer surveys may provide insights into characteristics of customers most likely to respond to

PTR events.

6.5 Non-event Based Programs

An evaluation was conducted for one non-event based resource in 2012, RTP. Future RTP aggregate

load impacts are closely tied to the size of new participants relative to the existing population. If all of

the new participants come from the 200 kW to 500 kW category, the resulting aggregate load

reduction will be relatively lower. On the other hand, if SCE is able to successfully market RTP and

recruit more large customers, the resulting aggregate load reduction will be relatively higher. It is

important that SCE continues to market RTP to large customers and not just focus on the 200 kW to

500 kW segment.

RTP would also likely benefit from an analysis of how to further optimize price schedule selection. The

schedules are currently selected based on downtown Los Angeles daily maximum temperatures on the

previous day. The current rule is transparent and easy for participants to understand and track, but

may not always target load impacts to time periods when they are most needed. Based on the

evaluator’s experience in modeling SCE system and individual customer loads, the main difference

between high and extreme system loads is not daily maximum temperature, but rather overnight heat

build-up. It is recommended that an assessment be made of the incremental improvement of

different pricing schedule selection rules and the associated tradeoffs, including the effect on

transparency and clarity.

30

Appendix A Regression Specifications

A.1 Base Interruptible Program

∑∑

∑∑

∑

∑

∑

∑

∑

∑∑

Variable Description

hourly BIP customer load at time t

estimated constant term

through estimated parameters

, , and

binary variables that indicate which TOU rate block is in effect for each hour

series of binary variables for each hour, which is interacted with all of the remaining variables because each has an impact that varies by hour

series of binary variables representing five different day types (Monday, Tuesday-Thursday, Friday, Saturday, Sunday/Holiday)

series of binary variables for each month

total number of cooling degree hours (base 70) per day

total number of cooling degree hours per day squared

total number of heating degree hours (base 70) per day

total number of heating degree hours squared

binary variable for event days from other DR programs

binary variable representing each BIP event day

et error term

31

A.2 Agricultural and Pumping Interruptible Program

∑∑

∑∑

∑

∑

∑

∑

∑

∑

Variable Definition

average hourly demand (kW) for each time period

estimated constant term

through regression model parameters

series of binary variables for each hour, which account for the basic hourly load shape of the customer after other factors such as weather and prices are accounted for

series of binary variables representing three different day types (Monday, Tuesday-Thursday, Friday); weekends are excluded from the model

series of binary variables for each month designed to reflect seasonality in loads

sum of cooling degree hours (base 65) for the day

TotalCDHt squared

sum of heating degree hours (base 65) for the day

TotalHDHt squared

binary variable representing a customer’s participation in another DR event

binary variable representing an AP-I event day

error term

32

A.3 Summer Discount Plan – Commercial

24

2

24

2

4,

5

2

,3

24

2

21

t

tt

t

ttdt

dt

tdtt

t

ttCDHHrDTypeHrLoad

E

DR t t

t

t

ttDRtDRtCDDCDDSQSDPHr

1

24

1

24

1

7

24

1

6,,5

24

1

8

t

ttAlrt


tLoad vector of observations on 60-minute interval usage for a customer

tHr binary time variables that are 1 in interval t and 0 otherwise

tdtDType

, day-type variables set to 1 on a specific day type and 0 otherwise

tCDH cooling degree hour variable

tCDDSQ cooling degree hour squared variable

tCDD cooling degree day variable

tDRSDP

, set of binary variables set to 1 on an SDP event day and 0 otherwise

tAlrt binary variable indicating days on which there was a Flex Alert

1 estimated constant term

t2 through

8 coefficients to be estimated that quantify the impacts associated with the various

interactions between variables

t error term

33

A.4 Summer Discount Plan – Residential Ex post and ex ante summer months:

∑ ∑

∑( )

∑

∑

∑

∑ ∑( )

∑

∑


load in hour for the average customer in A-bank

estimated parameters

dummy variable for hour

number of events that occurred for A-bank

dummy variables for program event days, specific to A-bank

average of the day’s hourly load in hours 1 through 10 for the average customer in A-bank

dummy variable for PTR event days

weather variables during hour for customers in A-bank

dummy variable for Mondays

series of dummy variables for days of the week, Tuesday through Friday

dummy variable for the month of September

series of dummy variables for the months of July, August and October

error term

The summer month ex ante model specifications are the same as the ex post specifications, with the

following exceptions:

The ti

MornLoad,

term is not included in the ex ante model specification

34

A cooling degree hour (base 65) variable is used in the ex ante model specification rather than

the t

Weather term

Ex ante winter months:

∑

∑

∑( )

∑ ∑

∑

∑


customer demand in hour for the modeled customer group



cooling degree hours

heating degree hours5

dummy variable for Monday

dummy variable for Friday

series of dummy variables for each day of the week

series of dummy variables for each month

error term

5 Heating degree hours (HDH) is defined as MAX[0, 50 – TMP], where TMP is the hourly temperature expressed in degrees

Fahrenheit. Customer group-specific HDH values are calculated using data from the most appropriate weather station.

35

A.5 Critical Peak Pricing Ex post, average event:

Ex post, individual event:

∑

∑


indicates each individual , date and event

model constant

pre-existing difference between treatment and control customers6

difference between event and non-event days common to both CPP participants and control group members

7

net difference between CPP and control group customers during event days – this parameter represents the difference-in-differences

time effects for each date controlling for unobserved factors that are common to all treatment and control customers but unique to time period

customer fixed effects controlling for unobserved factors that are time-invariant and unique to each customer. It does not control for fixed characteristics such as air conditioning that interact with time-varying factors like weather

Treatment binary indicator of whether or not the customers is part of the treatment (CPP) or control group

Event binary indicator of whether an event occurred that day. Impacts are only observed if the customer is on CPP (Treatment = 1) and it was an event day

ε error for each individual and date

Ex ante:

Ten models were tested for each customer. The final results for each customer are based on the

model that produces the smallest errors and least bias for that customer. The 10 models vary in

how weather variables were defined, if at all, and in the inclusion of monthly or seasonal variables.

6 In practice, this term is absorbed by the fixed effects, but it is useful for representing the model logic.

7 In practice, this term is absorbed by the time effects, but it is useful for representing the model logic.

36

Model #

Specification

1

∑ ∑

∑

∑ 2 3 … 24}

2

∑ ∑

∑

∑ 2 3 … 24}

3

∑ ∑

∑

∑ 2 3 … 24}

4

∑ ∑

∑

∑ 2 3 … 24}

5

∑ ∑

∑ ∑

2 3 … 24}

6

∑ ∑

∑ 2 3 … 24}

7

∑ ∑

∑ 2 3 … 24}

8

∑ ∑

∑ 2 3 … 24}

9

∑ ∑

∑ 2 3 … 24}

10

∑ ∑

∑

2 3 … 24}

37


kW energy usage in each hourly interval t = (1,2,3 …24) for each date, d

Year binary variable for year of the hourly observation

Season binary variable indicating whether the hourly observation falls in the summer or winter season

Month binary variable indicating the month of the hourly observation

Daytype binary variable for the day type of the hourly observation (Sundays and holidays and Tuesday through Thursday are grouped together)

CDH cooling degree hour - the maximum of zero and the hourly temperature value less a base value

CDHSQR square of cooling degree hour

CDD cooling degree day - the maximum of zero and the mean temperature of the day of the hourly observation less a base value

CDDSQR square of cooling degree day

OvernightCDH average of CDH from 12 AM through 9 AM

eventday1,2,3 ... n binary variables indicating each event day, 1, 2, 3 ... n

38

A.6 Demand Bidding Program

t

i

ti

MONTH

i

i

ti

DTYPE

i

i

ti

h

i

i

tti

SUMMER

i

i

tti

FRI

i

i

tti

MON

i

i

tti

Weather

i

i i

titi

OTH

i

i

titi

MornLoad

itti

DBP

Evti

E

Evt

t

eMONTHbDTYPEbhbSUMMERhb

FRIhbMONhbWeatherhb

OtherEvthbMornLoadhbDBPhbaQ

)()()()(

)()()(

)()()(

10

6

,

5

2

,

24

2

,

24

2

,

24

2

,

24

2

,

24

1

,

24

1

24

1

,,

24

1

,,,,

1


Qt demand in hour t for a customer enrolled in DBP prior to the last event date

b's estimated parameters

hi,t dummy variable for hour i

DBPt indicator variable for program event days

Weathert Weather variables selected in the model screening process

E number of event days that occurred during the program year

MornLoadt variable equal to the average of the day’s load in hours 1 through 10

OtherEvtt equals one on the event days of other demand response programs in which the customer is enrolled

MONt dummy variable for Monday

FRIt dummy variable for Friday

SUMMERt dummy variable for the summer pricing season8

DTYPEi,t series of dummy variables for each day of the week

MONTHi,t series of dummy variables for each month

et error term

8 The SCE summer pricing season is July through September.

39

A.7 Capacity Bidding Program and Demand Response

Contracts

t

i

ti

MONTH

i

i

ti

DTYPE

i

i

ti

h

i

i

tti

SUMMER

i

i

tti

FRI

i

i

tti

MON

i

i

tti

Weather

i

i

titi

OTH

i

i i

titi

MornLoad

itti

AGG

Evti

E

Evt

t

eMONTHbDTYPEb

hbSUMMERhbFRIhb

MONhbWeatherhbOtherEvthb

MornLoadhbAGGhbaQ

)()(

)()()(

)()()(

)()(

10

6

,

5

2

,

24

2

,

24

2

,

24

2

,

24

2

,

24

1

,

24

1

,,

24

1

24

1

,,,,

1


tQ

demand in hour t for a customer enrolled in an aggregator demand response program prior to the last event date

b ’ estimated parameters

tih

, dummy variable for hour i

tAGG indicator variable for program event days

tWeather weather variables selected in the model screening process

E number of event days that occurred during the program year

tiMornLoad

, variable equal to the average of the day’s load in hours 1 through 10

tiOtherEvt

, equals one in the event hours of other demand response programs in which the customer is

enrolled

tMON dummy variable for Monday

tFRI dummy variable for Friday

tSUMMER dummy variable for the summer pricing season9

tiDTYPE

, series of dummy variables for each day of the week

tiMONTH

, series of dummy variables for each month

te error term

The ex ante model specifications are the same as the ex post specifications, with the following

exceptions:

The ti

MornLoad,

term is not included in the ex ante model specification

A cooling degree hour (base 60) variable is used in the ex ante model specification rather than

the t

Weather term

9 The SCE summer pricing season is July through September.

40

A.8 Save Power Day Ex post and ex ante summer months:

∑ ∑

∑( )

∑

∑∑

∑

∑( )

∑

∑


customer demand in hour

estimated parameter coefficients

dummy indicator for hour

indicator variable for program event days

weather conditions during hour (e.g., measured by CDD, CDH or THI)

number of event days that occurred during the program year

variable equal to the average of the day’s load in hours 1 through10

dummy variable for day type


dummy variable for the month of September

dummy variable for days on which CAISO encouraged conservation (near Flex Alert Days)

dummy variable for CAISO Flex Alert days

error term

41

Ex ante winter months:

∑

∑

∑( )

∑ ∑

∑

∑


customer demand in hour for the modeled customer group



cooling degree hours

heating degree hours10

dummy variable for Monday

dummy variable for Friday

series of dummy variables for each day of the week


error term

A.9 Real Time Pricing

∑ ∑

∑

∑

∑∑

∑∑

∑ 2 2

10 Heating degree hours (HDH) is defined as MAX[0, 50 – TMP], where TMP is the hourly temperature expressed in degrees

Fahrenheit. Customer group-specific HDH values are calculated using data from the most appropriate weather station.

42

The following weather variables were also included in the models for weather-sensitive customers:

∑

∑


estimated constant

estimated parameter coefficients

indicator variables representing the hours of the day, designed to estimate the effect of daily schedules on usage behavior and event impacts

indicator variable for the month

RTP price in effect for each hour

RTP price squared

ratio between the RTP price in effect for each hour and the maximum price for the day, which captures load shifting to hours when prices are relatively low

series of binary variables representing three different day types (Monday, Tuesday-Thursday, Friday)

2 2 binary variable for the most recent year of load data

total number of cooling degree hours (base 70) per day

total number of heating degree hours (base 70) per day

error term

43

Appendix B Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions by Month and Forecast Year

Table B-1: 2013 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions


Monthly System Peak


Emergency

BIP-15 124 134 138 138 132 144 142 146 151 143 132 145

BIP-30 450 439 479 465 486 469 476 467 477 497 412 463

AP-I 30 31 36 45 57 60 59 59 52 44 34 32

SDP-C 0 0 0 0 0 42 58 78 67 0 0 0

SUB-TOTAL 604 604 653 648 675 716 735 750 748 685 578 639

Price-responsive

SDP-R 0 0 0 63 135 220 253 291 284 166 0 120

CPP 24 23 24 23 21 21 19 18 19 20 25 23

DBP 2 2 2 3 3 4 4 4 4 3 2 3

SUB-TOTAL 26 26 26 89 160 244 275 313 307 190 27 146

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 11 11 11 10 0 0

DRC-DA 0 0 0 0 13 15 16 16 14 13 0 0

DRC-DO 0 0 0 0 114 118 120 129 125 115 0 0

SUB-TOTAL 0 0 0 0 137 142 146 156 150 137 0 0


SPD 3 2 2 2 11 12 15 20 18 17 3 4

SUB-TOTAL 3 2 2 2 11 12 15 20 18 17 3 4

Non-event Based

RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 633 632 682 739 983 1,111 1,173 1,246 1,228 1,035 608 790

44



Monthly System Peak


Emergency

BIP-15 125 136 140 140 134 146 145 148 153 145 134 147

BIP-30 458 446 487 473 494 477 484 475 485 506 419 471

AP-I 33 35 40 50 63 67 65 65 58 49 37 35

SDP-C 0 0 0 0 0 41 57 77 66 0 0 0

SUB-TOTAL 617 617 667 662 692 732 750 765 763 700 590 653

Price-responsive

SDP-R 0 0 0 64 137 223 256 294 288 168 0 122

CPP 24 24 24 24 21 21 19 19 19 20 25 24

DBP 2 2 2 3 4 4 4 4 5 4 2 3

SUB-TOTAL 27 26 27 91 162 248 279 317 311 192 27 148

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 15 16 17 17 16 14 0 0

DRC-DO 0 0 0 0 125 129 132 142 137 126 0 0

SUB-TOTAL 0 0 0 0 150 155 160 170 163 150 0 0


SPD 3 3 3 3 13 14 16 20 17 15 3 4

SUB-TOTAL 3 3 3 3 13 14 16 20 17 15 3 4

Non-event Based

RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 647 646 697 756 1,017 1,145 1,206 1,280 1,259 1,063 620 805

45



Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 36 37 43 53 67 71 68 68 60 51 39 36

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 629 629 679 673 702 741 758 771 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 25 24 25 24 22 21 19 19 19 21 25 24

DBP 3 2 2 3 4 4 5 6 6 6 3 4

SUB-TOTAL 27 26 27 92 165 251 284 323 317 197 28 151

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 12 12 14 18 16 14 3 4

SUB-TOTAL 3 3 3 3 12 12 14 18 16 14 3 4

Non-event Based

RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 659 658 708 767 1,043 1,171 1,232 1,305 1,284 1,085 624 811

46



Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 37 39 44 54 69 72 69 69 61 51 39 37

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 630 630 680 675 704 742 759 772 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 25 24 25 24 22 21 19 19 19 21 25 24

DBP 4 4 4 3 7 7 7 9 10 9 5 6

SUB-TOTAL 29 28 29 92 168 254 286 326 321 200 30 153

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 12 12 14 18 16 14 3 4

SUB-TOTAL 3 3 3 3 12 12 14 18 16 14 3 4

Non-event Based

RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 662 661 712 770 1,047 1,175 1,235 1,309 1,289 1,089 625 813

47

Table B-5: 2017-2023 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions


Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 37 39 44 54 69 72 69 69 61 51 39 37

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 630 630 680 675 704 742 759 772 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 25 24 25 24 22 21 19 19 19 21 25 24

DBP 6 6 5 4 9 8 8 10 12 10 5 6

SUB-TOTAL 31 30 30 93 169 255 287 327 322 201 30 154

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 12 12 14 18 16 14 3 4

SUB-TOTAL 3 3 3 3 12 12 14 18 16 14 3 4

Non-event Based

RTP 0 0 0 0 0 -3 1 7 5 6 0 0

SUB-TOTAL 0 0 0 0 0 -3 1 7 5 6 0 0

PORTFOLIO TOTAL 664 663 713 770 1,049 1,176 1,236 1,311 1,290 1,090 625 813

48

Appendix C Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions by Month and Forecast Year

Table C-1: 2013 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions


Monthly System Peak


Emergency

BIP-15 126 131 132 136 129 137 138 142 149 145 131 143

BIP-30 463 436 463 472 491 480 475 466 479 501 426 468

AP-I 33 33 41 48 60 62 59 60 54 47 29 30

SDP-C 0 0 0 0 0 46 65 87 71 0 0 0

SUB-TOTAL 622 600 636 655 680 726 737 755 753 693 585 641

Price-responsive

SDP-R 0 80 191 192 187 270 289 311 269 227 0 188

CPP 24 23 23 21 19 18 40 17 39 20 25 20

DBP 2 3 3 3 4 4 4 4 4 4 2 3

SUB-TOTAL 27 105 217 215 210 291 332 332 313 251 27 211

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 11 11 11 10 0 0

DRC-DA 0 0 0 0 14 15 16 16 14 13 0 0

DRC-DO 0 0 0 0 115 120 121 130 123 118 0 0

SUB-TOTAL 0 0 0 0 139 145 148 157 148 141 0 0


SPD 3 3 4 3 11 14 17 21 16 18 3 5

SUB-TOTAL 3 3 4 3 11 14 17 21 16 18 3 5

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 651 708 863 873 1,046 1,180 1,235 1,270 1,233 1,109 616 857

49



Monthly System Peak


Emergency

BIP-15 128 133 134 138 131 139 140 144 151 147 133 145

BIP-30 471 443 471 480 500 488 483 474 487 510 433 475

AP-I 36 37 46 53 66 69 65 66 60 52 31 33

SDP-C 0 0 0 0 0 46 64 86 70 0 0 0

SUB-TOTAL 635 613 650 671 697 742 753 770 767 708 597 653

Price-responsive

SDP-R 0 81 194 194 189 273 293 315 273 230 0 191

CPP 24 23 23 21 20 18 40 18 40 21 25 20

DBP 2 3 3 3 4 4 4 5 5 4 3 3

SUB-TOTAL 27 106 219 218 213 295 337 337 317 255 28 214

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 15 16 17 18 16 14 0 0

DRC-DO 0 0 0 0 127 132 133 143 136 129 0 0

SUB-TOTAL 0 0 0 0 152 158 161 172 161 153 0 0


SPD 3 4 5 3 14 15 17 20 15 16 3 5

SUB-TOTAL 3 4 5 3 14 15 17 20 15 16 3 5

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 665 723 881 892 1,081 1,215 1,269 1,304 1,265 1,139 628 872

50



Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 39 40 49 56 70 72 68 69 62 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 648 624 662 682 708 751 760 776 772 713 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 25 23 23 21 20 18 41 18 40 21 25 20

DBP 3 3 3 3 4 5 5 6 6 6 4 4

SUB-TOTAL 28 108 222 221 216 300 342 342 323 260 29 218

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 3 5 3 12 14 16 19 13 15 3 4

SUB-TOTAL 3 3 5 3 12 14 16 19 13 15 3 4

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 678 736 895 906 1,108 1,242 1,295 1,330 1,290 1,162 631 879

51



Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 40 41 50 58 72 74 70 70 63 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 649 626 664 684 709 752 761 777 773 714 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 25 23 23 21 20 18 41 18 40 21 25 20

DBP 4 4 4 4 7 7 7 9 10 9 5 6

SUB-TOTAL 29 110 224 222 219 303 345 346 327 263 30 220

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 3 5 3 12 14 16 19 13 15 3 4

SUB-TOTAL 3 3 5 3 12 14 16 19 13 15 3 4

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 681 739 899 908 1,113 1,246 1,299 1,334 1,294 1,166 633 881

52

Table C-5: 2017-2023 Portfolio Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions


Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 40 41 50 58 72 74 70 70 63 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 649 626 664 684 709 752 761 777 773 714 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 25 23 23 21 20 18 41 18 40 21 25 20

DBP 6 6 6 4 9 9 9 10 11 10 5 6

SUB-TOTAL 30 111 226 222 221 304 346 347 328 264 30 220

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 3 5 3 12 14 16 19 13 15 3 4

SUB-TOTAL 3 3 5 3 12 14 16 19 13 15 3 4

Non-event Based

RTP 0 0 6 0 6 5 1 5 5 6 0 0

SUB-TOTAL 0 0 6 0 6 5 1 5 5 6 0 0

PORTFOLIO TOTAL 683 741 900 909 1,115 1,248 1,300 1,335 1,296 1,167 633 881

53

Appendix D Program Specific Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions by Month and Forecast Year

Table D-1: 2013 Program Specific Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions


Monthly System Peak


Emergency

BIP-15 124 134 138 138 132 144 142 146 151 143 132 145

BIP-30 450 439 479 465 486 469 476 467 477 497 412 463

AP-I 30 31 36 45 57 60 59 59 52 44 34 32

SDP-C 0 0 0 0 0 42 58 78 67 0 0 0

SUB-TOTAL 604 604 653 648 675 716 735 750 748 685 578 639

Price-responsive

SDP-R 0 0 0 63 135 220 253 291 284 166 0 120

CPP 37 36 37 34 33 33 31 31 31 33 37 36

DBP 59 63 69 61 59 71 73 70 76 71 67 71

SUB-TOTAL 95 99 106 158 228 323 357 392 390 270 104 227

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 11 11 11 10 0 0

DRC-DA 0 0 0 0 13 15 16 16 14 13 0 0

DRC-DO 0 0 0 0 114 118 120 129 125 115 0 0

SUB-TOTAL 0 0 0 0 137 142 146 156 150 137 0 0


SPD 3 3 3 3 13 16 19 26 23 21 4 5

SUB-TOTAL 3 3 3 3 13 16 19 26 23 21 4 5

Non-event Based

RTP 0 0 0 0 0 -8 2 18 24 17 0 0

SUB-TOTAL 0 0 0 0 0 -8 2 18 24 17 0 0

PORTFOLIO TOTAL 702 707 761 809 1,053 1,189 1,260 1,341 1,335 1,129 686 872

54



Monthly System Peak


Emergency

BIP-15 125 136 140 140 134 146 145 148 153 145 134 147

BIP-30 458 446 487 473 494 477 484 475 485 506 419 471

AP-I 33 35 40 50 63 67 65 65 58 49 37 35

SDP-C 0 0 0 0 0 41 57 77 66 0 0 0

SUB-TOTAL 617 617 667 662 692 732 750 765 763 700 590 653

Price-responsive

SDP-R 0 0 0 64 137 223 256 294 288 168 0 122

CPP 37 36 37 35 34 33 31 32 31 33 38 36

DBP 61 69 77 69 69 83 85 82 89 83 76 82

SUB-TOTAL 98 105 114 167 240 339 373 408 408 284 114 240

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 15 16 17 17 16 14 0 0

DRC-DO 0 0 0 0 125 129 132 142 137 126 0 0

SUB-TOTAL 0 0 0 0 150 155 160 170 163 150 0 0


SPD 4 3 3 3 15 17 20 26 22 18 3 5

SUB-TOTAL 4 3 3 3 15 17 20 26 22 18 3 5

Non-event Based

RTP 0 0 0 0 0 -8 2 18 24 17 1 1

SUB-TOTAL 0 0 0 0 0 -8 2 18 24 17 1 1

PORTFOLIO TOTAL 718 726 784 833 1,097 1,235 1,306 1,387 1,380 1,169 708 898

55



Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 36 37 43 53 67 71 68 68 60 51 39 36

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 629 629 679 673 702 741 758 771 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 37 36 37 35 34 34 32 32 31 33 38 37

DBP 70 68 74 65 66 79 81 78 85 80 73 78

SUB-TOTAL 107 104 111 165 239 338 373 409 408 284 111 238

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 14 16 19 24 20 17 3 4

SUB-TOTAL 3 3 3 3 14 16 19 24 20 17 3 4

Non-event Based

RTP 1 1 0 0 0 -8 2 18 25 17 1 1

SUB-TOTAL 1 1 0 0 0 -8 2 18 25 17 1 1

PORTFOLIO TOTAL 740 736 793 842 1,119 1,256 1,326 1,407 1,399 1,186 707 899

56



Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 37 39 44 54 69 72 69 69 61 51 39 37

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 630 630 680 675 704 742 759 772 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 38 37 38 35 34 34 32 32 31 33 38 37

DBP 67 70 77 67 69 82 84 82 89 84 75 81

SUB-TOTAL 105 107 114 167 242 341 376 412 413 288 113 241

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 14 16 19 24 20 17 3 4

SUB-TOTAL 3 3 3 3 14 16 19 24 20 17 3 4

Non-event Based

RTP 1 1 0 0 0 -8 2 18 25 17 1 1

SUB-TOTAL 1 1 0 0 0 -8 2 18 25 17 1 1

PORTFOLIO TOTAL 739 741 798 845 1,124 1,261 1,331 1,412 1,404 1,190 709 902

57

Table D-5: 2017-2023 Program Specific Aggregate Ex Ante Load Impact Estimates for 1-in-2 System Conditions


Monthly System Peak


Emergency

BIP-15 127 138 142 142 136 147 146 149 154 146 134 147

BIP-30 466 453 494 479 499 482 488 478 488 508 420 472

AP-I 37 39 44 54 69 72 69 69 61 51 39 37

SDP-C 0 0 0 0 0 41 56 76 65 0 0 0

SUB-TOTAL 630 630 680 675 704 742 759 772 768 705 592 656

Price-responsive

SDP-R 0 0 0 65 139 226 260 298 292 170 0 124

CPP 38 37 38 35 34 34 32 32 31 33 38 37

DBP 69 74 80 69 72 84 86 84 91 85 75 81

SUB-TOTAL 107 111 118 169 245 344 378 414 414 289 113 241

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 9 10 10 10 10 10 0 0

DRC-DA 0 0 0 0 16 18 19 19 17 15 0 0

DRC-DO 0 0 0 0 138 142 146 156 151 139 0 0

SUB-TOTAL 0 0 0 0 164 170 175 186 178 163 0 0


SPD 3 3 3 3 14 16 19 24 20 17 3 4

SUB-TOTAL 3 3 3 3 14 16 19 24 20 17 3 4

Non-event Based

RTP 1 1 0 0 0 -8 2 18 25 17 1 1

SUB-TOTAL 1 1 0 0 0 -8 2 18 25 17 1 1

PORTFOLIO TOTAL 741 745 801 847 1,127 1,264 1,333 1,414 1,406 1,191 709 902

58

Appendix E Program Specific Ex Ante Load Impact Estimates for 1-in-10 System Conditions by Month and Forecast Year

Table E-1: 2013 Program Specific Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions


Monthly System Peak


Emergency

BIP-15 126 131 132 136 129 137 138 142 149 145 131 143

BIP-30 463 436 463 472 491 480 475 466 479 501 426 468

AP-I 33 33 41 48 60 62 59 60 54 47 29 30

SDP-C 0 0 0 0 0 46 65 87 71 0 0 0

SUB-TOTAL 622 600 636 655 680 726 737 755 753 693 585 641

Price-responsive

SDP-R 0 80 191 192 187 270 289 311 269 227 0 188

CPP 37 35 35 33 31 30 51 29 50 33 37 33

DBP 57 61 64 60 59 71 73 70 76 70 67 70

SUB-TOTAL 94 176 290 285 277 370 413 410 395 330 105 291

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 11 11 11 10 0 0

DRC-DA 0 0 0 0 14 15 16 16 14 13 0 0

DRC-DO 0 0 0 0 115 120 121 130 123 118 0 0

SUB-TOTAL 0 0 0 0 139 145 148 157 148 141 0 0


SPD 3 4 5 3 13 17 21 26 20 23 4 6

SUB-TOTAL 3 4 5 3 13 17 21 26 20 23 4 6

Non-event Based

RTP 0 0 14 0 17 24 2 24 24 17 0 0

SUB-TOTAL 0 0 14 0 17 24 2 24 24 17 0 0

PORTFOLIO TOTAL 719 780 945 944 1,126 1,282 1,321 1,373 1,340 1,203 695 938

59



Monthly System Peak


Emergency

BIP-15 128 133 134 138 131 139 140 144 151 147 133 145

BIP-30 471 443 471 480 500 488 483 474 487 510 433 475

AP-I 36 37 46 53 66 69 65 66 60 52 31 33

SDP-C 0 0 0 0 0 46 64 86 70 0 0 0

SUB-TOTAL 635 613 650 671 697 742 753 770 767 708 597 653

Price-responsive

SDP-R 0 81 194 194 189 273 293 315 273 230 0 191

CPP 37 35 35 34 32 30 52 30 51 33 38 33

DBP 59 67 71 68 69 83 85 82 89 82 78 81

SUB-TOTAL 96 183 300 296 290 386 429 427 413 346 115 305

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 15 16 17 18 16 14 0 0

DRC-DO 0 0 0 0 127 132 133 143 136 129 0 0

SUB-TOTAL 0 0 0 0 152 158 161 172 161 153 0 0


SPD 4 4 6 4 16 19 22 26 19 20 4 6

SUB-TOTAL 4 4 6 4 16 19 22 26 19 20 4 6

Non-event Based

RTP 0 0 14 0 17 24 2 24 24 17 1 1

SUB-TOTAL 0 0 14 0 17 24 2 24 24 17 1 1

PORTFOLIO TOTAL 735 800 971 971 1,170 1,329 1,367 1,419 1,384 1,244 716 964

60



Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 39 40 49 56 70 72 68 69 62 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 648 624 662 682 708 751 760 776 772 713 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 37 36 36 34 32 30 52 30 51 33 38 33

DBP 68 65 68 65 65 78 81 78 86 79 74 77

SUB-TOTAL 106 183 301 295 289 386 430 427 413 346 112 304

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 4 6 4 14 18 20 24 17 19 3 5

SUB-TOTAL 3 4 6 4 14 18 20 24 17 19 3 5

Non-event Based

RTP 1 1 14 0 17 25 2 25 25 17 1 1

SUB-TOTAL 1 1 14 0 17 25 2 25 25 17 1 1

PORTFOLIO TOTAL 757 812 983 981 1,194 1,351 1,388 1,439 1,404 1,262 715 966

61



Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 40 41 50 58 72 74 70 70 63 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 649 626 664 684 709 752 761 777 773 714 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 38 36 36 34 32 31 52 30 51 33 38 33

DBP 65 68 71 66 68 82 84 82 90 83 76 80

SUB-TOTAL 103 186 304 297 292 389 433 431 418 350 114 307

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 4 6 4 14 18 20 24 17 19 3 5

SUB-TOTAL 3 4 6 4 14 18 20 24 17 19 3 5

Non-event Based

RTP 1 1 14 0 17 25 2 25 25 17 1 1

SUB-TOTAL 1 1 14 0 17 25 2 25 25 17 1 1

PORTFOLIO TOTAL 756 816 987 985 1,199 1,356 1,393 1,444 1,409 1,266 718 969

62

Table E-5: 2017-2023 Program Specific Aggregate Ex Ante Load Impact Estimates for 1-in-10 System Conditions


Monthly System Peak


Emergency

BIP-15 130 135 135 139 132 140 141 145 152 147 133 146

BIP-30 479 450 478 486 505 493 488 477 490 512 434 477

AP-I 40 41 50 58 72 74 70 70 63 54 33 34

SDP-C 0 0 0 0 0 45 63 85 69 0 0 0

SUB-TOTAL 649 626 664 684 709 752 761 777 773 714 599 657

Price-responsive

SDP-R 0 82 197 197 192 277 296 319 276 233 0 194

CPP 38 36 36 34 32 31 52 30 51 33 38 33

DBP 68 72 75 68 71 84 86 84 92 84 76 80

SUB-TOTAL 105 190 307 299 295 392 435 433 419 351 114 307

Demand Response

Aggregator-managed

CBP-DA 0 0 0 0 0 0 0 0 0 0 0 0

CBP-DO 0 0 0 0 10 10 10 11 10 10 0 0

DRC-DA 0 0 0 0 17 18 19 19 17 16 0 0

DRC-DO 0 0 0 0 140 145 147 158 149 142 0 0

SUB-TOTAL 0 0 0 0 166 172 176 188 176 168 0 0


SPD 3 4 6 4 14 18 20 24 17 19 3 5

SUB-TOTAL 3 4 6 4 14 18 20 24 17 19 3 5

Non-event Based

RTP 1 1 14 0 17 25 2 25 25 17 1 1

SUB-TOTAL 1 1 14 0 17 25 2 25 25 17 1 1

PORTFOLIO TOTAL 759 820 991 987 1,202 1,359 1,395 1,446 1,411 1,267 718 970

southern california edison's 2012 demand response load ... · this report summarizes the load...

Documents