Self -Generation Incentive Program Greenhouse Gas Staff Proposal · 2018-12-28 · 5 multiplied by the applicable efficiency (V DC x Amp-Hours x Applicable Efficiency) (see SGIP Handbook,

1

December 31, 2018 September 6, 2018

Self-Generation Incentive Program Greenhouse Gas Staff Proposal

ENERGY DIVISION, CALIFORNIA PUBLIC UTILITIES COMMISSION

DISTRIBUTED GENERATION RULEMAKING 12-11-005

2

Revised Self-Generation Incentive Program Greenhouse Gas Staff Proposal

ENERGY DIVISION, CALIFORNIA PUBLIC UTILITIES COMMISSION

DISTRIBUTED GENERATION RULEMAKING 12-11-005

3

Table of Contents

Table of Contents................................................................................................................................................... 3

Definitions and Acronyms ...................................................................................................................................... 4

Executive Summary ............................................................................................................................................... 6

Introduction ......................................................................................................................................................... 10

Proposed GHG Signal ........................................................................................................................................... 13

Proposals for New Projects .................................................................................................................................. 15

Commercial Projects ............................................................................................................................................... 15

Residential Projects ................................................................................................................................................. 23

Proposals for Legacy Projects ............................................................................................................................... 39

Appendix A: Public Utilities Code Section 379.6 ................................................................................................... 50

Appendix B: OSESMO Analysis of PG&E E6 with New TOU Periods ...................................................................... 55

Appendix C: OSESMO LADWP Residential TOU Rate Analysis .............................................................................. 56

4

Definitions and Acronyms

DEFINIT IONS

Developer: The developer for a project is, if not the individual homeowner applying for SGIP

incentives for systems located on their own property, the corporate entity registered and in

good standing with the Secretary of State of California that handles a substantial amount of the

project’s development activities (see 2017 SGIP Handbook, Section 4.1.5).

GHG impact of storage: The GHG impact of a customer’s storage device is the difference

between the customer’s emission profiles with and without the storage.

GHG signal: A digitally accessible data feed of current marginal greenhouse gas emissions rates

(in units of kg/kWh) that updates at regular intervals (e.g. every 5 minutes) combined with

additional data feeds that deliver regularly updated forecasts of grid conditions for use in the

optimization of dispatch.

Legacy projects: Any project with an application submitted before the “go live” date for new

GHG rules (includes all currently installed projects). Staff recommends making the “go live” date

foureight months after a Commission decision on this proposal to allow sufficient time for

implementation of new rules.

Legacy Fleet: A developer shall be deemed to have a “Legacy Fleet” subject to fleet compliance

requirements if they have a minimum of 10 legacy projects statewide. Projects will enter the

fleet on January 1, 2020 and exit on December 31 of their tenth full calendar year (January 1-

December 31) in operation.

New projects: Any project with an application submitted on or after the “go live” date for new

GHG rules. Staff recommends making the “go live” date foureight months after a Commission

decision on this proposal to allow sufficient time for implementation of new rules.

New Residential Fleet: A developer shall be deemed to have a “New Residential Fleet” subject

to fleet compliance requirements if they have a minimum of 10 new residential projects

statewide. Projects will enter the fleet on January 1st following their operational start date and

exit on December 31st of their tenth full calendar year (January 1-December 31) in operation.

Program Year: A project’s program year is the year its incentive application was accepted by

the Program Administrator.

Rated energy capacity (kWh): The SGIP Handbook defines the rated energy capacity (kWh) for

DC/AC energy storage technologies as the nominal voltage multiplied by the amp-hour capacity

5

multiplied by the applicable efficiency (VDC x Amp-Hours x Applicable Efficiency) (see SGIP

Handbook, Section 5.1.2).

Roundtrip efficiency (RTE): The total kWh discharge of the system divided by the total kWh

charge over some period of time or number of cycles. SGIP storage systems are currently

required to maintain an RTE equal to or greater than 69.6% in the first year of operation in

order to achieve a ten-year average RTE of 66.5%, assuming a 1% annual degradation rate (see

SGIP Handbook, Section 5.3.1).

Single-cycle roundtrip efficiency (SCRTE): The total kWh discharge of the system divided by the

total kWh charge after one complete cycle. SCRTE is often verified in the factory and specified

on a device’s technical specifications sheet.

ACRONYMS

GHG: greenhouse gas

IOU: investor owned utility

PA: program administrator

PBI: performance-based incentive

PDP: performance data provider

PY: program year (see definition above)

RTE: roundtrip efficiency (see definition above)

SCRTE: single-cycle roundtrip efficiency (see definition above)

SGIP: Self-Generation Incentive Program

WG: SGIP GHG Signal Working Group

6

Executive Summary

California statute limits eligibility for incentives under the Self-Generation Incentive Program

(SGIP) to “distributed energy resources that the commission…determines will achieve

reductions in emissions of greenhouse gases.”1 To ensure energy storage projects met this

requirement, in 2012 the California Public Utilities Commission (Commission) adopted a

roundtrip efficiency (RTE) standard,2 with the implicit assumption that retail rates would

incentivize storage to charge during low grid emission times and discharge during high grid

emission times.

Subsequent SGIP storage impact evaluations have found that SGIP storage has led to a net

increase in greenhouse gases (GHGs), in part because time of use (TOU) peak periods have not

aligned with high grid emission times, and in part because retail rates incentivized customers to

prioritize demand charge management over TOU rate arbitrage. The Commission convened a

working group and subsequently directed Energy Division staff to propose new operational

requirements based on the emissions of the electric grid to replace the RTE standard, and new

verification and enforcement mechanisms to ensure compliance.3

This paper presents staff’sOn September 6, 2018, an Assigned Commissioner’s Ruling issued an

Energy Division staff paper with recommendations for new GHG rules, based in large part on

the working group’s discussions, modeling effort, and final report. The proposals presented

here are , and designed to ensure that SGIP systems meet minimum statutory requirements to

reduce GHG emissions while continuing to support the program’s other goals of market

transformation and providing grid support. After receiving stakeholder feedback in the form of

written comments filed on September 26, 2018 and October 5, 2018, and oral comments made

at the GHG Staff Proposal Workshop (October 22, 2018), staff revised its recommendations and

is issuing a revised Staff Proposal for comment.

Revised Proposals for New and Legacy Projects

Staff proposes the following revised operational requirements and verification and

enforcement measures for new and legacy projects.

1 SB 412 (Kehoe, 2009), codified in Public Utilities Code Section 379.6(b)(1) 2 Resolution E-4519 3 Assigned Commissioner’s Rulings in Rulemaking 12-11-005, issued December 29, 2017 and July 26, 2018.

7

For new projects, defined as those submitting applications after the “go live” date for new rules

(likely Q2/Q3 2019),, staff recommends different rules for commercial4 and residential projects:

• New commercial projects: Staff proposes a PBI incentive structure for all new

commercial projects, such that 4050% of the incentive is paid upfront and the remaining

6050% is paid over five years. PAs would verify each project’s GHG reductions annually

and, if the project is found to reduce GHGs less than 25 kg5 kilograms of CO2 per

kilowatt hour (kg/kWh) or increase GHGs, the PA would reduce the project’s annual

incentive payment in proportion toby $1/kg ($1000/ton) of CO2 over the GHGs, subject

to defined exceedance bands.5 kg/kWh reduction threshold. PAs would provide projects

with quarterlysemi-annual feedback on GHG performance. The RTE requirement would

be eliminated.

• New residential projects: Staff proposes to eliminate the annual RTE requirement and

require all new residential projects to enroll in a rate with peak starting at or after 4pm,

pair with and charge at least 75% from onsite solar, and have a single-cycle RTE of at

least 85%. In addition, staff proposes to require developer fleets of 10 or more new

residential projects (“New Residential Fleets”) to reduce GHGs and meet aggregate

cycling requirements annually. Developers whose fleets were found to increase GHGs or

fall short of their aggregated cycling requirement would be temporarily suspended by

the PAs, meaning they would not be permitted to submit new applications for 90, 180,

or 360 days.on an approved time-varying rate and have a single-cycle RTE (SCRTE) of at

least 85%. Projects that meet these criteria would be deemed to reduce GHGs, and no

annual GHG verification or enforcement would be required.

For legacy projects, defined as those submitting applications before the “go live” date for new

rules, staff proposes to provide two pathways to compliance for all customer classes. Projects

may either adhere to the existing rules at the time of application, including the minimum

annual RTE standard, or may opt in to a GHG-reducing pathway, whereby projects could choose

to forego the minimum RTE requirement and instead commit to operating in a way that

achieves annual GHG reductions. Staff proposes a developer fleet approach to verify and

enforce compliance with both the RTE and GHG pathways. Developer fleets of 10 or more

legacy projects opting in to the same pathway (RTE or GHG) would be required to comply with

the requirements of their respective pathways annually. Developers whose fleets were found in

breach of the requirements of their respective pathways would be temporarily suspended by

4 For the purposes of this report, “commercial” projects include all nonresidential projects.

8

the PAs, meaning they would not be permitted to submit new applications for a defined period

in proportion to the breach in compliance.eliminate the annual RTE requirement and instead

require developer legacy fleets to reduce GHGs annually. PAs would leverage existing

verification work performed by Itron through the annual impact evaluation process to identify

non-compliant fleets, and use existing handbook language to enforce the GHG requirement for

developers whose legacy fleets are found to increase GHGs. PAs would focus their enforcement

efforts on the highest emitters and give developers the chance to set and meet compliance plan

milestones prior to issuing infractions. Residential customers with legacy systems who enroll on

an approved time-varying rate would be exempt from enforcement.

Revisions to the Original Staff Proposal

Revisions to the September 6, 2018 staff proposal include the following:

• Direct the PAs to make an interim GHG signal available within five months of a

Commission decision, and a final GHG signal available within eight months of a

Commission decision, in order to give the PAs additional time for contracting and

program participants additional time to integrate the signal into their operations

• Postpone the “go live” date for new rules from four to eight months following a

Commission decision

• For new commercial projects:

o Require projects to reduce GHGs by 5 kg/kWh rather than 25 kg/kWh to recoup

full payment

o Change the PBI split for all new commercial projects from 40/60 to 50/50

o Direct the PAs to provide projects with semi-annual rather than quarterly

feedback on year-to-date GHG performance

o For projects that have completed their 5-year PBI term, continue to recommend

a fleet compliance approach, but use existing handbook enforcement measures

rather than requiring program suspensions

• For new residential projects:

o “Deem” projects to reduce GHGs if the customer is on an approved time-varying

rate and the projects have a SCRTE of 85% or higher; do not require additional

verification and enforcement

o Direct the PAs to maintain a list of approved time-varying rates whose TOU

periods align with grid emissions; require new residential projects to enroll on an

approved rate rather than a TOU rate with peak starting at 4pm or later

o Remove the requirement to pair with solar

• For legacy projects:

9

o Eliminate the RTE pathway for all legacy projects and instead require developer

legacy fleets to reduce GHGs on an annual basis

o Direct PAs to use existing verification methods and handbook infraction language

to verify and enforce GHG reductions rather than a fleet compliance approach

with automatic program suspensions; PAs would leverage existing verification

work performed by Itron through the annual impact evaluation process to

identify non-compliant fleets, focus their enforcement efforts on the highest

emitting fleets, and give developers the chance to set and meet compliance plan

milestones prior to issuing infractions

o Exempt legacy residential customers who enroll on an approved time-varying

rate from enforcement

o For legacy commercial projects subject to a 260 cycle/year requirement, reduce

their cycling requirement to 130 cycles/year

10

Introduction

BACKGROUND

In accordance with Public Utilities Code 379.6(b)(2), Decision (D.) 15-11-027 reiterated SGIP’s

goal to promote GHG-reducing distributed energy technologies and updated the minimum RTE

standard with the aim of ensuring SGIP energy storage projects yield net zero GHG emissions

over the first 10 years of their operation. Itron’s 2016 and 2017 SGIP Advanced Energy Storage

Impact Evaluations (August 2017 and September 2018) found that regardless of RTE, SGIP

storage projects on average increase GHG emissions, which runs counter to the SGIP’s GHG

emission reduction goal and the program requirement that all SGIP technologies operate in a

way that reduces GHG emissions in order to be eligible for SGIP incentives.5,6

The Energy Division held a stakeholder workshop on November 15, 2017 to review and discuss

the Itron reports’ findings. During the workshop, participants suggested the CPUC convene a

working group tasked with developing new operational requirements to improve SGIP storage

projects’ GHG impacts. Workshop participants also indicated that the availability of a “GHG

signal” could help storage systems operate to reduce GHGs to at least zero.

On December 29, 2017, an Assigned Commissioner’s Ruling (ACR) established the Greenhouse

Gas Signal Working Group (WG) to develop new SGIP storage operational requirements based

on the GHG emissions of the electric grid, and new verification and enforcement mechanisms

to ensure compliance with the requirements. The ACR also tasked the WG with developing a

proposed GHG signal methodology, detailing a number of minimum requirements:

• The marginal GHG emissions of the grid reported for either NP15 or SP15,7 as applicable

• In 60, 30, 15, or 5-minute increments as determined by the WG

• Forecasted for the day ahead

• Automatically transmitted to the energy storage system, or the controller of the system

if systems are controlled remotely

The WG was facilitated by Alternative Energy Systems Consulting (AESC) and consisted of SGIP

PAs, the Office of Ratepayer Advocates (ORA)8, solar and energy storage companies and trade

5 Per D.16-06-055, SGIP’s overall program goal is threefold: reduce GHG emissions, provide grid support, and

encourage market transformation. 6 Reports available at: http://www.cpuc.ca.gov/WorkArea/DownloadAsset.aspx?id=6442454964 7 NP15 and SP15 are CAISO congestion zones in northern and southern California, respectively. 8 Now the California Public Advocates Office.

11

associations, energy non-profits, and Energy Division staff. From January to June 2018, the WG

met regularly to design and carry out a modeling strategy to test alternative operational

requirements to ensure SGIP projects reduce GHGs. The WG used five proprietary models

(Tesla, AMS, Stem, Customer Power Solar, and Avalon) and one newly-developed public model

to conduct over 5,000 model runs with varying parameters (including system and customer

characteristics). AESC executed nondisclosure agreements with all modelers to be able to

review all proprietary model runs and to provide aggregated results analysis that informs many

of the WG’s recommendations. The WG also developed recommendations for verification and

enforcement mechanisms. The final corrected version of the WG report was attached to the

September 6, 2018 ACR.

On July 26, 2018, the CPUC released an ACR directing Energy Division staff to prepare a

proposal for new operational requirements based on the emissions of the electric grid to

replace the RTE standard, and new verification and enforcement mechanisms to ensure

compliance. On September 6, 2018, staff issued a paper with recommendations for new GHG

rules, based in large part on the working group’s discussions, modeling effort, and final report,

and designed to ensure that SGIP systems meet minimum statutory requirements to reduce

GHG emissions while continuing to support the program’s other goals of market transformation

and providing grid support.

After receiving stakeholder feedback in the form of written comments filed on September 26,

2018 and October 5, 2018 and oral comments made at the GHG Staff Proposal Workshop

(October 22, 2018), staff revised its recommendations and is reissuing this paper for comment.

SCHEDULE

The schedule below lays out next steps for a deliberative process and stakeholder participation

once the draft Staff Proposal is released:

Date Event

September 6, 2018 ACR issuing staff proposal with Final Working Group Report attached and

requests comments.

September 26, 2018 Comments on staff proposal due

October 22, 2018 Workshop on staff proposal

November 9, 2018 ACR issuing revised staff proposal with request for comments

November 29, 2018 Comments on revised staff proposal due

Q1 2019 Issue proposed decision

Please see the ACR to which this staff proposal is attached for deadlines for party comments on

this proposal.

12

PROPOSAL ORGANIZATION

The remainder of this report is organized into three sections and one appendix:

• The Proposal for a GHG Signal section recommends requirements and a timeline for a

GHG signal to be made publicly available.

• The Proposals for New Projects section recommends operational requirements and

verification and enforcement mechanisms for projects applying after the “go-live” date

for new rules.

• The Proposals for Legacy Projects section recommends operational requirements and

verification and enforcement mechanisms for projects applying before the “go-live”

date for new rules.

• Appendix A presents the complete text of Public Utilities Code Section 379.6, the

statute governing SGIP.

CUSTOMER B I LL IMPACTS OF PROPOSALS

The primary purpose of this document is to offer proposals for ensuring compliance with the

longstanding statutory requirement that all SGIP-eligible technologies reduce GHG emissions,

thus the proposals in this report generally require projects to achieve GHG reductions

regardless of the impact on customer bill savings. Developers and other market participants

hold the responsibility to provide potential customers with the knowledge and tools that will

help them make informed decisions regarding their energy storage investments.

Staff notes that WG modeling suggests the GHG signal may make it possible to achieve GHG

reductions with minimal bill impact. The Open-Source Energy Storage Model (OSESMO) finds

that average bill savings across all “No GHG Solution” base case modeling runs is $23,482/year

and average GHG impact is an increase of 4.0 kg/kWh, while average bill savings across all “GHG

Signal Co-Optimization”9 model runs with a GHG weighting of $65/metric ton (including both

perfect-forecast and day-ahead-forecast results) is $23,352/year and average GHG impact is a

decrease of 8.1 kg/kWh.10

9 “GHG Signal Co-Optimization” is when storage is dispatched to optimize both bill savings and GHG reductions. 10 https://osesmo.shinyapps.io/osesmo_results_viewer/

13

Proposed GHG Signal

PROPOSAL

The Commission should direct the SGIP PAs to contract with WattTime or anothera qualified

entity to provide a GHG signal with the following features:

• A digitally-accessible, real-time, marginal GHG emissions factor for NP15 and SP15

California Independent System Operator (CAISO) zones, at 5-minute and 15-minute

intervals, in units of kgCO2/kWh;

•o The signal will be calculated using the same heat rate-based methodology as in

the most recent SGIP program evaluation report, but with updated parameters

and data sources more suitable for real-time use.

o This signal will provide the marginal emissions per kWh calculated based on a

natural gas-fired power plant producing energy at a price equaling the real-time

(5-minute) CAISO Locational Marginal Price with costs equal to the most recent

publicly available data on gas prices, CO2 prices, and variable operating costs

constrained by reasonable maximum and minimum efficiencies. When the

calculated heat rate is zero or below, instead it is assumed that the marginal

generator is renewable and the marginal emissions rate is zero.

• For storage operation planning purposes, a 15-minute (updated every 15 minutes), 72

hour-ahead (updated hourly), month-ahead (updated daily), and year-ahead (updated

monthly) forecast.

TheAn interim GHG signal should be made available within fourfive months of a Commission

decision to allow program administrators, and participants and the a final GHG signal provider

should be made available within eight months of a Commission decision, to allow sufficient

time for implementation. The interim GHG signal should provide program participants with

enough information to learn how to incorporate the signal into their operational algorithms.

The final signal should meet the full parameters outlined above.

NOTE: All subsequent proposals in this report assume the full GHG signal will be publicly

available by the time new rules go into effect.

D ISCUSSION

The WG modeling results show that a GHG signal is useful to storage operators looking to

safeguard customer bill savings while co-optimizing for GHG emissions reductions. Therefore,

we propose the Commission require that a marginal GHG emissions signal, including accurate,

14

relevant real time and forecasted data and reliable, secure software delivery, be made available

to SGIP projects.

To ensure timely implementation, we recommend directing the PAs may wish to contract with

WattTime, a nonprofit organization with an existing tool for reporting the real-time GHG

intensity of the grid available to clients nationwide. WattTime provided technical expertise

throughout the WG process to help develop a GHG signal tailored to SGIP and the requirements

of the December 2017 ACR. In talks with staff, WattTime has indicated they would be able to

deploy a GHG signal four to six months after being commissioned by the PAs.

15

Proposals for New Projects

This section contains proposals for “new” projects, defined as those submitting an SGIP

application on or after the “go live” date for new GHG rules. Staff recommends making the go

live date foureight months after a Commission decision setting GHG rules for SGIP-funded

energy storage systems to allow sufficient time for WattTime, the PAs, GHG signal provider,

Energy Solutions, and market participants to implement.11 The July 26, 2018 ACR proposes to

issue a Commission decision in Q1 2019, likely putting the go live date in Q2/Q3 2019.around

January 1, 2020.12

PROPOSAL SUMMARY

Staff proposes a performance-based incentive (PBI) structure for all new commercial projects,

such that 4050% of the incentive is paid upfront and the remaining 6050% is paid over five

years. PAs would verify each project’s GHG reductions annually and, if the project is found to

reduce GHGs less than 25 kg5 kilograms of CO2 per rated energy capacity (kg/kWh) or increase

GHGs, the PA would reduce the project’s annual incentive payment in proportion toby $1 per

kilogram ($1000 per ton) of CO2 over the GHG emissions, subject to defined exceedance

bands.5 kg/kWh reduction threshold. PAs would provide projects with quarterlysemi-annual

feedback on GHG performance.

The annual RTE requirement would be eliminated for all new commercial projects. The 130

cycles per year requirement would remain.

CURRENT RULES

Commercial projects are currently required to meet a ten-year average roundtrip efficiency of

66.5% and cycle 130 times per year. The SGIP Handbook does not explicitly require projects to

reduce GHGs.

11 Eight months for implementation would allow three months for PAs to hire a GHG signal provider, two months

for development of an interim signal, and three months for program participants to learn how to incorporate the

signal into their operations before new rules go into effect. The PAs and Energy Solutions would work in parallel

during this time to make changes to the handbook and database systems. 12 If the eight-month mark falls within a couple months of January 1, 2020, staff recommends the Commission

adopt a “go live” date of January 1, 2020 for administrative ease and to avoid prorating PBI payment penalties for

new commercial projects in the last few months of 2017.

16

The current methods for awarding incentives to commercial projects are based on project size.

Projects 30 kW and larger (“Performance-Based Incentive” or “PBI” projects) receive 50% of

their incentive upfront and the remaining 50% over five years based on annual kilowatt-hours

discharged, while projects smaller than 30 kW (“non-PBI” projects) receive 100% of their

incentive upfront.

DETAILED PROPOSAL

Staff proposes to make all new commercial projects 40/6050/50 PBI (4050% of the incentive is

paid upfront and the remaining 6050% is paid over five years) and subject to all PBI rules,

including the requirement to contract with a Performance Data Provider (PDP) for five years

and install revenue grade metering equipment. listed on the CEC’s list of Eligible System

Performance and Revenue Grade Meters.13

Operational Requirements

• Projects would be required to reduce GHGs a minimum of 25 kg5 kilograms of CO2 per

rated energy capacity (kg/kWh) annually to recoup full payment

• Annual RTE requirement would be eliminated

• Cycling requirement would remain at 130/year

Verification Mechanism

• PAs would verify each project’s GHG reductions annually using PBI data

• PAs would provide each project with quarterlysemi-annual feedback on GHG

performance

Enforcement Mechanism

• PAs would reduce a project’s annual PBI payment in proportion to its GHG impact:by $1

per kilogram ($1000 per ton) of CO2 over the 5 kg/kWh reduction requirement

Annual GHG Impact14 PBI Payment

Reduced By:

Decrease of more than 25 kg CO2 per rated energy capacity (kWh) 0%

Decrease of less than 25 kg CO2 per rated energy capacity (kWh)15 25%

Increase of less than 25 kg CO2 per rated energy capacity (kWh) 50%

13 See Section 5.5 of the 2017 SGIP Handbook. 14 The GHG impact of a customer’s storage device is defined as the difference between the customer’s emission

profiles with and without the storage. 15 Inclusive of a net zero impact on emissions.

17

Increase of more than 25 kg CO2 per rated energy capacity (kWh) 100%

• PBI payment deductions would be capped at 100% of annual PBI payment

• PBI payment deductions would be permanently forfeited and returned to the SGIP

incentive budget.

• PAs would have discretion to increase or decrease payment deductions or levy other

penalties only in exceptional circumstances and with written approval from Energy Division.

VERIFICATION AND ENFORCEMENT AFTER PBI TERM

Staff proposes to continue verification and enforcement of GHG performance past a project’s

five-year PBI term via a fleet compliance approach. Developer fleets of 10 or more new

commercial projects would be required to reduce GHGs and meet aggregate cycling

requirements annually. Developers whose fleets are found to increase GHGs or fall short of

their aggregated cycling requirement would be temporarily suspended by the PAs, meaning

they would not be permittedsubject to submit new applications during a defined suspension

period. If these new rules go into effect Q3 2019, the first new commercial project would likely

receive its first upfront paymentinfractions as outlined in Q1 2021 and finish its five-year PBI

term in Q1 2026, thus fleet compliance for new commercial projects would begin

implementation in 2026.the SGIP handbook.

If these new rules go into effect January 1, 2020, the first new commercial project would likely

receive its first upfront payment in Q3 2021 and finish its five-year PBI term in Q3 2026, thus

fleet compliance for new commercial projects would begin implementation in 2026. As this

date approaches, staff recommends the PAs propose handbook modifications to implement a

fleet compliance approach.

D ISCUSSION

Commercial Projects 30 kW and Larger

Staff’s proposed operational requirements and verification and enforcement mechanisms for

commercial projects 30 kW and larger are based in large part on the “New PBI Proposal 1” from

the WG report, however staff proposes to increase the portion of incentives paid out on a

performance basis from 50% and feedback received in comments and at the workshop. The WG

put forth several options for mechanisms to 60% to provide greater incentive to reduce GHGs.

Staff’s proposed enforcement mechanism is stricter than the mechanisms proposed by the WG,

which ranged fromenforce GHG reductions, including relying solely on existing handbook

18

infraction language to, reducing PBI payments by 25%.-100%, and reducing PBI payments by a

dollar per kilogram multiplier.16 Staff supports reducing the PBI payment by 25% for small net

GHG reductions and by 50% and 100% for net GHG increases$1 per kilogram ($1000 per ton) of

CO2 over a 5 kg/kWh reduction threshold to reflect the emphasis statute places on GHG

reductions and to provide systems with incentive to reduce GHGs beyond net zero. The 5

kg/kWh reduction threshold and proposed penalties are discussed in more detail below.

Although the original staff proposal recommended a 40/60 PBI split for commercial projects,

staff now recommend a 50/50 PBI split based on feedback that withholding an additional 10%

of funds would not provide meaningful incremental incentive to reduce GHGs and would

require modifications to database systems and other aspects of program administration.

Commercial Projects Under 30 kW

Staff’s proposal to make commercial projects under 30 kW PBI is based on the “New Non-

PBI/Non-Res Proposal 1” from the WG report, in which part of a project’s incentive is withheld

and paid out over several years.

The WG debated the split, with 50/50, 70/30, 80/20, and 90/10 all proposed. Staff supports a

40/6050/50 split for the following reasons:

• Alignment with proposed rules for large commercial projects

• Provides greaterstrong incentive to reduce GHGs

• Staff did not see clear evidence in the WG report or during WG meetings that a 50/50 or

40/60 split would significantly reduce the finance-ability of smaller commercial projects

compared to a 70/30 or 80/20 split

Staff recognizes that making smaller commercial projects subject to all PBI rules will require

them to contract with a PDP and install revenue grade metering equipment, listed on the CEC’s

list of Eligible System Performance and Revenue Grade Meters,17 two requirements that did not

apply to this project class before. It is staff’s understanding that most commercial developers

already contract with a PDP or are one themselves, so this should not represent a significant

16 Also, the “New PBI Proposal 1b” in the WG report proposed to reduce a project’s PBI payments by its annual net

GHG increase multiplied by “a 4x multiple of the California Cap and Trade price of carbon at the time in which the

project applied to the program”. Staff did not determine whether this would tend to come out to more or less

than 50% of the PBI payment. See GHG Signal Working Group Final Report (September 6, 2018 Corrected Version),

pages 52-58, at (http://cpuc.ca.gov/sgip/). 17 See Section 5.5 of the 2017 SGIP Handbook.

19

additional cost. Staff have also heard anecdotally that installing revenue-gradeeligible metering

costscould cost roughly $2,000 per project.

While this is not an insignificant cost, staff believes requiring such equipment is appropriate

when a project’s incentive payment relies on accurate metering, and the costs of not requiring

such equipment may be greater when one factors in time spent by PAs verifying the technical

capabilities of non-standard metering equipment. The PAs shared with staff that they currently

receive significant administrative streamlining benefits from relying on the CEC’s list to verify

that the meter meets the requirements of the program, and requiring them to independently

approve meters could double the workload of their technical working group and delay

processing of applications. Thus staff find it prudent to continue relying on pre-approved,

independently-vetted lists of metering equipment like the CEC’s. If stakeholders are aware of

other lists of metering equipment that would meet the requirements of the program, including

lists of non-revenue grade meters, they are encouraged to share that information in their

comments on the staff proposal.

Enforcement Exceedance Bands

Staff supportsRequirement to Reduce GHGs by 5 kg/kWh to Recoup Full Payment

California statute limits eligibility for SGIP incentives to “distributed energy resources that the

commission…determines will achieve reductions in emissions of greenhouse gases,”18 thus staff

believe it is appropriate to require projects to reduce GHGs some minimum amount to recoup

full payment. Staff initially proposed to set this amount at 25 kg CO2 per rated energy capacity

(kWh), equivalent to a 0.18% reduction for non-PBI projects and 3.10% reduction for PBI

projects in Itron’s 2017 evaluated sample. However, comments on the original staff proposal

and at the workshop stated that most projects will not be able to achieve 25 kg while co-

optimizing between GHGs and bill savings.

Staff confirmed this with the publicly-available Open-Source Energy Storage Model (OSESMO)

results viewer and data.19 Figure 1 below shows OSESMO results for all commercial model

runs20 performing GHG co-optimization and using the highest $65/ton carbon-adder. The model

found that:

• 5% of model runs reduced emissions by more than 25 kg/kWh

18 Public Utilities Code Section 379.6(b)(1) 19 https://osesmo.shinyapps.io/osesmo_results_viewer/ 20 Model runs include new and old rates, storage-only and solar-plus-storage, 70% & 85% SCRTE, <30 kW and >30

kW.

20

• 34% reduced emissions by more than 10 kg/kWh

• 50% reduced emissions by more than 5 kg/kWh

• 71% reduced emissions by more than 0 kg/kWh21

Figure 1: OSESMO results showing GHG impact of all commercial model runs co-optimizing

between GHGs and bill savings with $65/ton carbon-adder22

Closer analysis of model runs that reduced emissions by more than 25 kg/kWh shows that they

do not share a single “silver bullet” characteristic that could easily be adopted as a program

requirement. Prevalent rates include A-1-STORAGE (NEW), SDG&E DG-R (NEW), and other new

TOU rates, but there are also a number of projects on old Commercial & Industrial TOU rates.

There are also a number of 70%-efficient storage systems, although the majority have 85%

single-cycle RTE, and there is a fairly even mix of standalone-storage and solar-plus-storage.

To ensure projects meet statutory requirements to reduce GHGs, staff recommends requiring

projects to reduce at least 5 kg/kWh to recoup full payment. The modeling shows that 50% of

model runs were able to use the GHG signal to achieve 5 kg/kWh or more of GHG reductions

without significantly impacting bill savings. Adopting a 5 kg/kWh reduction requirement will

help ensure GHG reductions and encourage future projects to find the combination of rate,

operational algorithm, customer load profile, system efficiency, and other factors needed to

reduce GHGs.

21 Model runs increasing emissions tended to have low SCRTE. 22 https://osesmo.shinyapps.io/osesmo_results_viewer/

21

Calculating PBI Payment Reductions

The original staff proposal supported the adoption of specific exceedance bands for payment

reductions to clearly convey pre-defined penalties to program participants and to calibrate

payment reductions to the breach in compliance. In comments on the original staff proposal

and at the workshop, there was widespread support for using a $/ton multiplier in lieu of

exceedance bands, which were criticized for penalizing small and large emitters roughly the

same, removing incentives for projects that can't reach the next exceedance band to reduce at

all, and being difficult to integrate with an algorithm responding to the GHG signal. Some

parties’ support for the $/ton approach was contingent on the $/ton value providing an

equivalent incentive to reduce GHGs as the proposed exceedance bands.

SGIP currently uses exceedance bands on PBI payment reductions to penalize generation

projects that increase GHGs relative to an annual grid emissions factor. Generation projects

that increase emissions by 0-5% are not penalized (“buffer band”), projects that increase

emissions by 5-10% have their payment reduced by 50%, and projects that increase emissions

more than 10% have their payment reduced by 100%.

Staff chose not to adopt a similar buffer band for storage projects because statute requires

SGIP projects to reduce GHGs, thus incentivized projects should aim below net zero GHGs. To

ensure meaningful reductions are achieved, the Commission should require projects to reduce

GHGs some minimum amount to recoup full payment. Staff proposes to set this amount at 25

kg CO2 per rated energy capacity (kWh), equivalent to a 0.18% reduction for non-PBI projects

and 3.10% reduction for PBI projects in Itron’s 2017 evaluated sample.

For projects that increase GHGs, staff chose the 0-25 kg per kWh (0-0.05 metric tons per kW)

and 25+ kg per kWh (0.05+ metric tons per kW) values for exceedance bands based on annual

net emissions data in the 2017 SGIP Storage Impact Evaluation.23 As can be seen in the figures

below, in 2017 the first exceedance band would have captured roughly one third of small

commercial projects and three quarters of large commercial projects, or roughly half of all

projects.

Staff supports using a $/ton approach for the reasons stated above. Staff proposes a $1/kg

($1000/ton) multiplier to provide an equivalent incentive to reducing PBI payments by 25% for

GHG reductions less than the required amount, 50% for GHG increases between 0 and 25

23 Staff chooses to normalize emissions by kWh rather than kW to align with existing SGIP rules governing incentive

calculations. Because the vast majority of projects assessed in the 2017 SGIP Storage Impact Evaluation had a 2-

hour duration of discharge, we divide by 2 to convert from the evaluation’s per kW figure to a per kWh figure.

22

kg/kWh, and 100% for GHG increases greater than 25 kg/kWh. The figure below compares

example PBI payment reductions for projects of varying sizes and emission rates at $1/kg versus

using exceedance bands. Applying a $1/kg penalty, projects emitting 15 kg/kWh would have

their annual PBI payment reduced by 57%, close to the 50% proposed originally, while projects

emitting 30 kg/kWh would have their annual payment reduced by 100%, same as the 100%

proposed originally.24

Figure 2: Net CO2 emissions per rebated capacity for non-PBI and PBI projects, reprinted from

the 2017 SGIP Storage Impact Evaluation (Figures 4-59 and 4-60) (green content added)

24 PBI payment deductions would be capped at 100% of annual PBI payment.

25 kgCO2/kWh

increase

23

: Example

PBI payment reductions using $1/kg ($1000/ton) vs. exceedance bands25

PROPOSAL SUMMARY

Staff proposes to eliminate the annual RTE requirement and require all new residential projects

to enroll in aon an approved time-varying rate with peak starting at or after 4pm, pair with and

charge at least 75% from a solar system, and and have a single-cycle RTE of at least 85%.

Projects that meet these criteria would be deemed to reduce GHGs and not be subject to

additional verification and enforcement. The PAs would maintain a list of pre-approved time-

varying rates, with TOU periods aligned with grid emissions and non-trivial price differentials

between periods.

25 Note that the PAs would reduce a project’s annual PBI payment by $1 per kilogram ($1000 per ton) of CO2 over

the 5 kg/kWh reduction requirement, not over net zero emissions. This explains why a project’s annual PBI

payment reduction amount may be greater than $1/kg multiplied by the project’s annual GHG emissions.

Annual GHG Emissions

Rate (kg/kWh) -4 0 15 30 50 -4 0 15 30 50

Annual GHG Emissions

(kg) -200 0 750 1500 2500 -2000 0 7500 15000 25000

Unadjusted Annual PBI

Payment ($350/kWh) 1,750$ 1,750$ 1,750$ 1,750$ 1,750$ 17,500$ 17,500$ 17,500$ 17,500$ 17,500$

Annual PBI Payment

Reduced By… ($) 50$ 250$ 1,000$ 1,750$ 1,750$ 500$ 2,500$ 10,000$ 17,500$ 17,500$

Annual PBI Payment

Reduced By… (%) 3% 14% 57% 100% 100% 3% 14% 57% 100% 100%

% PBI Reduction from

Initial Staff Proposal 25% 25% 50% 100% 100% 25% 25% 50% 100% 100%

500 kWh Project50 kWh Project

25 kgCO2/kWh

increase

24

In addition, staff proposes to require developer fleets of 10 or more new residential projects to

reduce GHGs and meet aggregate cycling requirements annually. Developers whose residential

fleets are found to increase GHGs or fall short of their aggregated cycling requirement will be

temporarily suspended by the PAs, meaning they will not be permitted to submit new

applications during a defined suspension period of 90, 180, or 360 days.

CURRENT RULES AND IMPACT EVALUATION F INDINGS

The SGIP Handbook currently requires residential projects to meet a ten-year average roundtrip

efficiency of 66.5% and cycle 52 times per year. The Commission’s purposes in establishing the

RTE and cycling requirements were to ensure projects reduce GHGs and are used for more than

just back-up power, respectively.26

The PAs currently require host customers to sign an affidavit that they will discharge the

storage system a minimum of 52 full discharges per year. Beyond this, the PAs currently do not

have a mechanism for annual verification of the RTE or cycling requirement. Residential

projects are currently only required to provide data in the event of an audit. Assessment of

residential project performance falls to Itron, which measures RTE, GHGs, and cycling for

residential projects as part of the annual SGIP storage impact evaluation. For the 2016

evaluation, Itron was not able to evaluate the residential class due to data availability issues.

For the 2017 evaluation, Itron installed meters directly on 30 residential projects (7% of the

total residential population subject to evaluation in 2017) and produced statistically significant

results based on that sample. For the 2018 evaluation and beyond, Itron expects to receive

sufficient data directly from newer projects with more advanced data reporting abilities.

2017 Storage Impact Evaluation Findings

The 2017 SGIP Storage Impact Evaluation found that residential projects had a mean observed

RTE of 38% and a mean capacity factor of 2.2%, indicating these systems were used almost

exclusively to provide backup power.27 Additionally, Itron found that when not idle or providing

backup, systems tended to discharge during PV generating hours and charge in the early

evening during the CAISO system peak. On average, residential projects increased emissions by

0.06 metric tons of CO2 per kW, or 30 kg of CO2 per kWh, in 2017.

26 The Commission adopted the RTE requirement in Resolution E-4519 and originally prohibited SGIP funding for

backup-only systems in D.01-03-073. 27 Interestingly, most residential projects met program cycling requirements of 52 cycles per year, indicating the

cycling requirement may be set too low to achieve the Commission’s desired capacity factor. The Commission may

want to consider raising the residential cycling requirement in the future if the residential fleet’s capacity factor

stays low even after new GHG rules are implemented.

25

This is the kind of dispatch behavior that existing SGIP rules are intended to avoid. However, it

is important to note that none of the observed projects were enrolled on a time-varying rate,

and therefore did not have information or financial incentive to dispatch in a way that would

reduce GHGs or increase capacity factor.

The impact evaluation also included a simulation of storage behavior under future grid

conditions, and found that residential projects, optimizing for customer bill savings on a PG&E

TOU rate (E6) and using new TOU periods with an on-peak definition of 4-9pm, still led to an

increase in GHGs, just less than on the old rate. This is likely because under new TOU periods,

projects are still incentivized to charge during some hours of relatively high emissions (e.g. 2-

4pm).

DETAILED PROPOSAL

Staff recommends the Commission adopt upfront eligibility criteria and a fleetdeemed

approach to verification and enforcement for new residential systems. The proposal for upfront

eligibility criteria is based primarily on the WG modeling and WG’s recommendations for

operational requirements for new residential projects. Staff’s proposal for fleet verification and

enforcement is very similar to the practice used now - impact evaluation results are

communicated to the PAs, who then determine what corrective action to take, if any. Staff

proposes to enhance the current practice by establishing a mid-year feedback loop to

developers, tightening the timeframe for communicating annual performance to the PAs, and

adopting specific penalties for non-compliance.

Definition of “New Residential Fleet”: In this section, a developer shall be deemed to have a

“New Residential Fleet” subject to fleet compliance requirements if they have a minimum of 10

new28 residential projects statewide. Projects would enter the fleet on January 1st following

their operational start date and exit on December 31st of their tenth full calendar year (January

1-December 31) in operation.


• All new residential projects would be required to meet the following upfront eligibility

criteria when applying:

o Enrolled on aan approved time-varying rate with a peak starting at 4pm or later

o Paired with solar

28 This report defines “new” projects on page 10 as those submitting an SGIP application on or after the go live

date for new GHG rules, likely in Q2/Q3 2019.

26

� Approved rates would have TOU periods aligned with grid emissions and

non-trivial price differentials between periods

o Single-cycle RTE of 85% or higher

o Sign an affidavit stating they will:

� Cycle 52 times per year

� Charge at least 75% from solar29

All New Residential FleetsVerification and Enforcement Mechanism

• Projects on an approved rate with single-cycle RTE of 85% or higher would be

requireddeemed to reduce GHGs and meet aggregated cycling requirements on an

annual basis, for each compliance period January 1-December 31.


• Staff proposesnot be subject to task Itron30 with semiannual additional verification of

fleet GHG and cycling performance. Staff confirms that additional funds from the SGIP

administrative budget are available for this purpose.31

• Itron would assess performance twice each year: mid-year, to provide feedback on

performance, and end-of-year, to verify compliance for the annual compliance period

(January 1-December 31). Itron would communicate findings to the PAs by August 15

and February 15 respectively.

• Each developer with a New Residential Fleet would be required to submit complete

project-level electrical charge and discharge data to Itron upon request and no later

than July 20 of each year and January 20 following each year in which they have an

eligible fleet. For each fleet, Itron may determine the most efficient means to collect

data to assess performance. For larger fleets, Itron may choose to use sampling, in

which case they (Itron) would select the projects sampled.

• The PAs would then communicate findings and any enforcement steps to each fleet

developer according to the following schedule:

29 Charging from solar is defined as charging the storage device during a time interval (e.g. 15-minute) when the

on-site solar is generating and not exceeding the amount (kWh) of solar generation in that time interval. 30 This report uses Itron’s name in lieu of “SGIP evaluator” because most program participants are familiar with

Itron and their role. Should the PAs subsequently hire a different entity to perform this work, that entity would

assume the duties ascribed to Itron here. 31 PG&E and SCE have large overages in their administrative budgets at present, however CSE and SoCalGas may

not have sufficient funds to cover additional M&E costs. If the Commission tasks the SGIP evaluator with

semiannual verification of fleet compliance, it may want to consider funding the work solely through PG&E and

SCE’s administrative budgets, or alternatively, re-allocating administrative funds across PAs to ensure each PA has

sufficient funds to cover their share of the work.

27

o On September 1, the PAs would communicate Itron’s findings for the January 1–

June 30 period. The PAs would work with fleet developers to review progress

reports and make operational adjustments as needed.

o On March 1, the PAs would communicate Itron’s findings for the January 1–

December 31 compliance period, and issue any penalties (see “Enforcement

Mechanism” immediately below).


• Developers whose fleets are found to increase GHGs or fall short of their aggregated

cycling requirement would be temporarily suspended by the PAs, meaning they would

not be permitted to submit new applications during a defined suspension period.

Suspension periods would advance only on days when the residential budget of the

“controlling PA”32 is open and would be proportional to the magnitude of the breach in

compliance:

o 0-10 kg CO2 per energy capacity (kWh) or 80-99% of required cycling → 90-day

suspension

o 10-30 kg CO2 per rated energy capacity (kWh) or 60-79% of required cycling →

180-day suspension

o >30 kg CO2 per rated energy capacity (kWh) or <60% of required cycling → 360-

day suspension

• If a developer incurs suspension periods for non-compliance with both the GHG and

cycling requirements, the aggregate suspension period would be the sum of the two

suspension periods.

• PAs would have discretion to increase or decrease the suspension period or levy

different penalties with written approval from Energy Division.

Following each annual compliance period, Energy Solutions33 would list and rank each fleet’s

annual GHG and cycling performance on SelfGenCA.com and/or CaliforniaDGStats.ca.gov,

highlighting high-performing developers who were successful in reducing GHGs and achieving

high capacity factors in their fleets.

32 The “controlling PA” for a developer is the PA to whom the developer has submitted the most applications in the

past 12 months. 33 Energy Solutions is SGIP’s software provider and manages the SelfGenCA.com and CaliforniaDGStats.ca.gov

websites.

28

• Systematic verification and enforcement of a project’s GHG performance would end

after its tenth year in a New Residential Fleet.

D ISCUSSION

Staff is encouraged byStaff’s original proposal drew on WG modeling findings that residential

storage systems that meet certain criteria are likely to reduce GHGs, and supports adopting

those criteria. However,but did not recommend a pure “deemed compliance” approach

because staff doesdid not have sufficient confidence in the WG modeling to recommend a pure

“deemed compliance” approach, asat that time. Staff instead proposed by some in the WG.

Staff believes a fleet approach to verification and enforcement are necessary to ensure

residential storage meets statutory program requirements toprojects reduce GHGs.

TheIn comments on the original staff proposal and at the workshop, stakeholders provided

feedback and additional information about both the WG modeling and the administrative cost

of a fleet compliance approach. Based on this feedback and additional analysis, staff has revised

its recommendation to support a deemed compliance approach to verification and

enforcement for residential projects on an approved time-varying rate with single-cycle RTE of

85% or higher. The arguments in favor of this approach are discussed in more detail below.

The following discussion first addresses upfront eligibility criteria, then verification and

enforcement.

Upfront Eligibility Criteria

Refined Understanding of WG Modeling

The original staff proposal based the recommendations for upfront eligibility criteria are based

primarily on the aggregated WG modeling showing that residential projects are likely to 83% of

model runs reduce GHGs if they enroll inon a rate with peak starting at or after 4pm3pm, pair

with and charge at least 75% from solar, and have a single-cycle RTE of at least 85%. These

results are summarized in the following figure from the WG report. The red circle highlights the

percentage of model runs with GHG reductions for projects that meet the combination of

criteria recommended herein the original staff proposal.

29

Figure 3: Residential summary of GHG and cost impacts for different scenarios, reprinted from

the WG Report (page 89 of corrected version)

Staff was hesitant in the original staff proposal to deem such projects to be GHG-reducing

because of limitations in the modeling such as inconsistent assumptions and the fact that not

all modelers ran all scenarios. However, WG participants shared at the workshop that the

OSESMO model is open-source, covers all modeling scenarios requested, and uses the standard

inputs agreed to by the WG. Accordingly, staff agree that OSESMO results are the most reliable

for our purposes. When subsetting the modeling to OSESMO runs only, 100% of model runs on

a new TOU rate,34 with 85% single-cycle RTE, and under a “no GHG solution”35 case, reduced

GHGs. Staff confirmed this result using the publicly-available OSESMO results viewer.36

The original staff proposal also cited simulated findings from the 2017 SGIP Storage Impact

Evaluation, that residential projects on PG&E’s E6 rate with new TOU periods optimizing for

customer bill savings led to an increase in GHGs, as evidence of the need for additional

verification and enforcement. At the workshop, Itron provided additional details about the

assumptions made in that analysis, including that single-cycle RTE was assumed to be 78%, and

noted that the analysis was only intended to show whether there was an improvement in GHG

34 The model includes two new residential TOU rates: PG&E EV-A (NEW) and SDG&E DR-SES. 35 Equivalent to optimizing for customer bill savings. 36 https://osesmo.shinyapps.io/osesmo_results_viewer/

30

impact when moving from old to new TOU periods (there was). After the workshop, the

OSESMO model was updated to include the E6 rate with new TOU periods at staff’s request,

and showed that 100% of Tier 2 model runs with 85% single-cycle RTE reduced GHGs (see

Appendix B).37 Some of the 70% single-cycle RTE and Tier 1 standalone-storage projects

increased emissions, so Itron’s findings that systems on this rate increased emissions seem

reasonably consistent.

Finally, in reply comments on the original staff proposal (pp. 5-6), Tesla cited public modeling

from the National Renewable Energy Laboratory’s System Advisor Model tool showing that the

deployment of storage with 85% single-cycle RTE, cycling 52 times per year, and charging at

least 75% from solar, reduces the GHG footprint of a typical residential customer on a TOU rate

with peak at 3pm or later by 2-18% depending on the operational mode (with solar self-

consumption representing the 18% figure). The model can be accessed at

https://sam.nrel.gov/.

Given staff’s refined understanding of the WG and Itron modeling, and in light of the additional

public modeling flagged by Tesla, staff has sufficient confidence that residential projects on new

TOU rates with 85% single-cycle RTE or higher will reduce GHGs to support a pure “deemed

compliance” approach.

List of Approved Rates

Time-varying residential rates with peak periods starting at 4pm or later are that are aligned

with times of high grid emissions are already available in PG&E, SCE, and SDG&E service

territories.38 Staff believes these new TOU rates, with peak periods better aligned with grid

emissions, WG modeling shows that such rates will provide residential projects with some

additional the information and incentive they need to dispatch in a way that reduces GHGs. This

is supported by the WG modeling, (see discussion above).

The original staff proposal recommended requiring residential projects to go on TOU rates with

peaks starting at 4pm or later. Staff worries that this blanket requirement may become

outdated as the emissions intensity of the grid at different hours evolves over time.

37 https://osesmo.shinyapps.io/osesmo_results_viewer/ 38 Staff have not yet been able to confirm with Los Angeles Department of Water and Power and other municipal

utilities that time varying rates with peak periods starting at 4pm or later are available in their service territories,

or if not, whether there are plans to adopt such rates between now and when new rules would go into effect.

31

To address these concerns, staff recommends the PAs maintain a list of approved residential

rates for new projects in each service territory. PAs would seek approval from the Commission

for the initial list and subsequent modifications via a Tier 2 advice letter. Rates that appear on

the list would be time-varying, have TOU periods that are aligned with grid emissions, and have

non-trivial price differentials between periods. Beyond these criteria, PAs would have discretion

to propose which found that for 85% SCRTE residential projects not responding to a GHG signal,

switching from old to new rates39 increases the likelihood projects will be GHG-reducing by 67%

for stand-alone storage and 16% for storage paired with solar (see Figure 2 above).rates they

believe should be eligible based on their likelihood to incentivize storage operation that

reduces GHGs. For the initial list, staff expects PAs will include in their proposal TOU rates with

peak periods starting at 3pm or later.

However, as noted above, Itron’s simulated dispatch results in the 2017 report found that

residential projects on a new PG&E TOU rate optimizing for customer bill savings still led to an

increase in GHGs, just less than on the old rate. The figure below shows negative GHG savings

(i.e. increased emissions) for both projects on the new E6 rate with a 4-9pm peak and projects

on the old E6 rate with a 1-7pm peak.

39 The WG report (p. 30 of the corrected version) refers to “new” rates as “rate plans proposed by the utilities (and

in a few cases already in effect) that are subject to TOU schedules substantially different to the conventional TOU

rates of the last 30 years. These TOU rates shift the on-peak period much further into the evening hours, thereby

better aligning period of high cost with periods of high marginal grid emissions.”

32

Figure 3: CO2 emissions savings per kW resulting from customer dispatch approach, by

project, for PG&E residential customers (N = 15), reprinted from 2017 SGIP Storage Impact

Evaluation (Figure 5-32)

The figure below compares marginal grid GHGs to the dispatch behavior of residential

customers using the new E6 rate. Although projects did discharge during peak times, they also

charged during times when grid emissions are relatively high, producing a net increase in

emissions.

Figure 4: PG&E residential customer average net summer discharge per rebated kw as

compared to marginal emissions rate (N = 15), reprinted from 2017 SGIP Storage Impact

Evaluation (Figure 5-33)

33

Staff surveyed residential rates either available now or proposed in recent or pending rate

cases, and identified several with super off peak periods for much of the year that may provide

more granular signals about when marginal grid emissions are at their lowest, notably PG&E’s

EV-A rate (just approved in D.18-08-013) and SCE’s TOU-D-PRIME (proposed in the A.17-06-030

settlement). However, without additional modeling to clearly show that customers on these

rates would reduce GHGs, staff does not have enough information to conclude that new TOU

rates alone will produce GHG-reducing dispatch. Other eligibility requirements are needed to

achieve a minimum level of confidence that residential projects will reduce GHGs.

APPROACH FOR NON-IOU RESIDENTIAL CUSTOMERS

Some non-IOU utilities and community choice aggregators (CCAs) may not offer time-varying

rates with TOU periods well-aligned with grid GHGs for their residential customers. For

instance, LADWP currently offers only one residential TOU rate, R-1B, with a “high” peak period

at 1-4pm, “low” peak at 10am-1pm and 5-8pm, and off peak period during all other hours.40,41

Staff was unable to reach LADWP regarding whether they planned to change TOU periods in the

future.

OSESMO was updated to include R-1B and showed that residential solar-paired storage projects

on this rate with 85% SCRTE or higher were unlikely to reduce emissions. Methodology and

results are shown in Appendix C and results can be explored interactively in the OSESMO results

viewer.42 The only way to achieve a high (~90%) percentage of GHG-reducing systems was to

perform GHG Co-Optimization in addition to meeting the requirements above. Even then,

storage systems did not reduce emissions using the imperfect day-ahead GHG forecast – only

systems using the perfect-forecast GHG signal reduced emissions.

In comments on the staff proposal, some parties proposed to allow customers without access

to appropriate rates to submit cycling plans demonstrating how they would operate their

systems to reduce GHGs. Staff does not support this option because such customers would not

have financial incentive to comply with their cycling plans in the absence of an appropriate

40 LADWP webpage at https://www.ladwp.com/ladwp/faces/ladwp/residential/r-customerservices/r-cs-

understandingyourrates/r-cs-ur-electricrates. 41 The “low” peak period is priced in between the “high” and off peak periods. 42 https://osesmo.shinyapps.io/osesmo_results_viewer/. The OSESMO results viewer displays R-1B model runs

using SP15 marginal emissions data. Modeling was also performed using LADWP-specific marginal emissions data

and is available upon request. The results were significantly worse than for SP15: none of the modeled sites

reduced emissions, even when responding to a perfect-forecast GHG signal with a $65/ton carbon weighting cost.

This appears to be because the LADWP signal is much flatter than the CAISO emissions signals, offering less

opportunity to charge at low-emissions times.

34

time-varying rate and PAs would not have the ability to verify or enforce compliance in the

absence of a fleet compliance regime. Staff invites other proposals for including these

customers in SGIP while maintaining a reasonable assurance they will reduce GHGs.

Pairing with Solar

The WG modeling foundThe original staff proposal recommended requiring new residential

projects to pair with solar, citing WG modeling results that for residential projects on new rates

and not responding to a GHG signal, adding the requirement to pair with and charge at least

75% from solar increasesincreased the likelihood projects will benumber of GHG-reducing

model runs by 16%, from 67 to 83% (see Figure 3). The WG posited that the reason for this is

that requiring a project to pair with solar makes it more likely for the developer to set the

device’s operation mode to “solar self-consumption”, which forces the storage to charge during

PV-generating hours, the cleanest hours on the grid.43 Essentially, this sets charge and

discharge timing constraints that are better aligned with grid emissions than TOU periods,

which don’t generally differentiate between low and lowest grid emission times. ).44,45 In

comments, many parties opposed excluding stand-alone storage from the program, arguing it is

equally capable of reducing GHGs as storage paired with solar and some customers may be

unable to deploy solar for various reasons (e.g. affordability, shaded roof).

This theory for why pairing with solar leads to more GHG-reducing dispatch is plausible, begging

the question why not just recommend charge and discharge timing constraints. The WG

modeled timing constraints on charge and discharge as one of the GHG reduction solution

options, but mysteriously found that they increase GHGs in some instances (see Appendix G of

the WG report). Thus, staff does not have an analytical basis to consider their adoption at this

time. Staff notes that the vast majority of residential systems are paired with solar anyway, and

expect this trend to continue independently of SGIP requirements.

44 The WG posited that the reason for this is that requiring a project to pair with solar makes it more likely for the

device’s operation mode to be set to “solar self-consumption,” which forces the storage to charge during PV-

generating hours, the cleanest hours on the grid. 45 Note that when calculating the GHG impacts of a storage device paired with a solar system, the device is

assumed to charge from the grid, not the solar. GHG reductions from the solar system are not attributed to the

storage device because it is assumed that the solar would have been installed regardless of the storage. Also, it is

not technically possible to track the source of energy used to charge the device for many systems, particularly AC-

paired systems where the storage and solar are behind separate inverters. Therefore, the calculated GHG

reductions associated with solar self-consumption stem from the timing of charge and discharge, not from energy

generated by an on-site zero-emissions source.

35

OSESMO modeled stand-alone storage on only one “new” rate, PG&E EV-A. Looking at model

results for this rate, with 85% single-cycle RTE and optimizing for customer bill savings, 100% of

runs reduced GHGs by an average of 7.67 kg/kWh. Adding the requirement to pair with solar

increased the average GHG reductions to 24.96 kg/kWh.46

There is clear evidence that pairing with solar increases GHG savings, likely because the

requirement to charge from solar effectively sends a more granular signal than TOU periods

about when grid emissions are low. However, there is also evidence that stand-alone storage is

still very likely to reduce GHGs when on “new” rates. Given this, and the fact that some

customers may be unable to deploy solar, staff recommends against requiring residential

storage to pair with solar.

85% Single-Cycle RTE

Staff supports the WG proposal to include 85% single-cycle RTE as an eligibility criterion for new

residential projects. WG modeling shows that under new rates and no GHG reduction solution,

projects with 85% single-cycle RTE are 33% more likely to reduce GHGs than projects with 70%

single-cycle RTE (see the two right quadrants in the figure below). It is staff’s understanding

that almost all new residential storage systems are lithium ion batteries and easily meet this

requirement.

46 https://osesmo.shinyapps.io/osesmo_results_viewer/

36

Figure 4: Cycles and GHGs for storage with solar on new rates by GHG reduction solution and

SCRTE (residential), reprinted from the WG report (page 73 of corrected version)

GHG Signal

Although the modeling clearly shows requiring projects to co-optimize between a GHG signal

and bill savings leads to a greater likelihood of GHG reductions than employing no GHG

reduction solution, staff questions the feasibility of requiring all new residential projects,

including those developed by individual homeowners, to respond to a remote signal. The other

eligibility criteria recommended here, combined with the fleet approach to verification and

enforcement, should produce GHG reductions without the need to mandate that all projects

develop the ability to respond to a GHG signal.

Verification and Enforcement

Although staff has sufficient confidence in the accuracy of the WG modeling effort to use it as a

basis for proposals on eligibility criteria for new residential projects, it does not have sufficient

confidence in the modeling to support a pure “deemed compliance” approach. Staff believes a

belt and suspenders approach, with eligibility criteria as the belt and verification and

enforcement as the suspenders, is necessary to ensure storage is operated in a way that

reduces GHGs.

Staff believes the WG modeling may be flawed for several reasons. The final results combined

findings from six different modelers, each using different assumptions for electric rates, SCRTE,

system size, and parasitic losses in their proprietary models. Additionally, not all modelers

37

modeled all scenarios; several scenarios had over four hundred model runs, while one had only

six. Presumably, all runs for the latter scenario came from a single modeler, making it difficult

to tell whether the results for that scenario are reflective of real impact or simply unique

assumptions made by the modeler. When combined, the aggregated modeling sometimes also

produced odd results. For example, it found that for projects paired with solar, on a new rate,

and responding to a GHG signal, removing the criteria that the storage system is paired with

solar raised the number of GHG-reducing runs from 84% to 100% (see Figure 2).

Because staff does not have credible modeling results showing that projects that meet these

eligibility criteria are highly likely to reduce GHGs, staff recommends the Commission adopt a

fleet compliance approach and put the onus on developers to ensure their projects comply with

the statutory requirement to reduce GHGs. The WG modeling indicates that projects that

respond to a GHG signal are more likely to reduce GHGs compared to projects that don’t

respond to a GHG signal. We expect each developer to determine for themselves to what

degree they need to supplement TOU rates and solar self-consumption with the GHG signal to

achieve fleet-wide GHG reductions.

In a few years, the Commission should have enough data on storage dispatch under new rates

to determine whether systems that meet certain upfront criteria can be reasonably assumed to

reduce GHGs, rendering verification and enforcement unnecessary. Staff recommends the

Commission re-assess the need for systematic verification and enforcement for residential

projects at that time, and no later than the program’s sunset in 2025.

PROJECT- VS. FLEET-LEVEL ENFORCEMENT

Enforcing GHG reductions at the individual project level is infeasible given the number of

projects and lack of sophistication of smaller developers and individual homeowners. A fleet

compliance approach is less costly to implement, is targeted at larger developers with the

ability to report performance data, and has the benefit of allowing developers to determine the

most efficient way to achieve GHG reductions across its fleet. Staff chose to set the lower limit

for fleets to 10 projects because at that number, most projects are captured and the total

number of fleets is kept manageable for verification and enforcement purposes.47

ENFORCEMENT EXCEEDANCE BANDS

47 As of August 6, 2018, there were 30 developers with ten or more residential projects paid or reserved, capturing

93% of all residential projects (SGIP Weekly Statewide Report, posted August 6, 2018).

38

Staff supports the adoption of specific exceedance bands to clearly convey pre-defined

penalties to developers, and to calibrate penalties to the breach in compliance. Staff chose the

0-10 kg per kWh (0-0.02 metric tons per kW), 10-30 kg per kWh (0.02-0.06 metric tons per kW)

and 30+ kg per kWh (0.06+ metric tons per kW) values for exceedance bands based on project-

level emissions data in the 2017 SGIP Storage Impact Evaluation (see figure below). Three

exceedance bands balance the need to make enforcement rules simple with the need to not

overly penalize fleets with small net increases in GHGs.

Figure 6: Net CO2 emissions per rebated capacity for residential projects, reprinted from the

2017 Itron Storage Impact Evaluation (Figure 4-71) (green content added)

10 kgCO2/kWh

30 kgCO2/kWh

39

Proposals for Legacy Projects

This section contains proposals for “legacy” projects, defined as those submitting an SGIP

application before the “go live” date for new GHG rules, likely in Q2/Q3 2019on January 1, 2020

(see discussion on page 15).

PROPOSAL SUMMARY

Staff proposes to provide two pathways to program compliance for all legacy projects,

regardless of customer class. Projects may either adhere to the existing rules at the time of

application, including the minimum annual RTE standard, or may opt in to a GHG reduction

pathway, whereby projects could choose to forego the minimum RTE requirement and instead

commit to operating in a way that achieves annual GHG emissions reductions. Under the GHG

reduction pathway, the annual cycling requirement for projects interconnected under older

rules requiring 260 annual cycles would be stepped down to the current 130 annual cycling

requirement.

Staff proposes to verify and enforce compliance with both the RTE compliance and GHG

reduction pathways on a developer fleet basis. Developer fleets of 10 or more legacy projects

opting in to the same pathway (RTE or GHG) would be required to comply with the

requirements of their respective pathways annually. Developers whose fleets are in breach of

the requirements of their respective pathways will be suspended by the PAs, meaning they will

not be permitted to submit new applications for a defined period in proportion to the breach in

compliance.

Itron would determine developer fleet compliance on a semiannual basis, which will then be

communicated to developers by the PAs. Enforcement would take place shortly following each

program year.

Staff proposes to eliminate the RTE requirement for all legacy projects and instead require

developer fleets to reduce GHGs on an annual basis. PAs would leverage existing verification

work performed by Itron through the annual impact evaluation process to identify non-

compliant fleets. PAs would use existing handbook language to enforce GHG reductions, grant

developers opportunities to bring their fleets into compliance prior to issuing infractions, and

focus their enforcement efforts on higher-emitting fleets. Residential customers with legacy

systems who enroll on an approved time-varying rate would be exempt from enforcement.

CURRENT RULES

40

The SGIP Handbook does not explicitly require energy storage projects to reduce GHGs at

present. All projects are instead required to meet a roundtrip efficiency of 66.5% averaged over

ten years. Commercial systems are required to discharge a minimum of either 260 or 130 full

discharges per year. Residential systems are required to discharge a minimum of 52 full

discharges per year.

DETAILED PROPOSAL

Staff proposes to offer two alternative compliance pathways for legacy projects:

1. Meet the SGIP minimum annual RTE requirement (no change from old rules), or

2. Opt in to a GHG reduction pathway that would exempt participants from the minimum

RTE requirement and establish a 130 cycle annual requirement for older projects

operating under the 260 annual cycles rule.


• The annual RTE requirement would be eliminated for all legacy projects and replaced

with a requirement to reduce GHGs at the developer fleet level on an annual basis

• At the time of implementation, PAs would email host customers/applicants48/

developers for all legacy projects and offer them the choice to opt in to one of the two

available pathways. Projects that don’t respond would be defaulted to the RTE

compliance pathway.

• Through this selection process, the PAs would develop two fleets per

developer/applicant: a GHG reduction fleet and an RTE compliance fleet. A developer

shall be deemed to have a “Legacy Fleet” (GHG or RTE pathway) subject to fleet

compliance requirements if they have a minimum of 10 legacy projects statewide in any

one pathway.

Existing projects will enter the Legacy Fleet on January 1, 2020 and exit on December 31

of their fifth full calendar year (January 1-December 31) in operation. Energy Solutions

will make developer/applicant fleet enrollment lists available online.


48 The term “applicant” precedes the developer decision established in PY 2017. Per the February 2016 SGIP

Handbook an “Applicant is the entity that is responsible for completing and submitting the SGIP application and

serves as the main point of contact for the SGIP Program Administrator throughout the application process. Host

Customers may act as the Applicant, or they may designate a third party (e.g. a party other than the Program

Administrator or the utility customer) to act as the Applicant on their behalf. Applicants may be third parties such

as, but not limited to, engineering firms, installation contractors, equipment distributors, Energy Service

Companies (ESCO), equipment lessors, etc. Host Customers may elect to change the Applicant at their discretion.

41

• Staff proposes to task Itron with semiannual verification of GHG and RTE performance,

as relevant to each Legacy Fleet. Staff confirms that additional funds from the SGIP

administrative budget are available for this purpose.49

• Each developer with a Legacy Fleet (GHG or RTE pathway) would be required to submit

complete project-level electrical charge and discharge data to Itron upon request and no

later than July 20 and January 20 of each year in which they have an eligible fleet. For

each fleet, Itron may determine the most efficient means to collect data to assess

performance. For larger fleets, Itron may choose to use sampling, in which case

developers would not have discretion to choose which projects are sampled.

• Itron would assess performance twice each year: mid-year, to provide feedback on

performance, and end-of-year, to verify compliance for the annual compliance period

(January 1-December 31). Itron would communicate findings to the PAs by August 15

and February 15 respectively.

• The PAs would communicate findings and any enforcement steps to each fleet

developer as follows:

o On September 1, the PAs would communicate Itron’s findings for the January 1-

June 30 period. The PAs would work with fleet developers to review progress

reports and make operational adjustments as needed.

o On March 1, the PAs would communicate Itron’s findings for the January 1 –

December 31 compliance period, and issue any penalties (see “Enforcement

Mechanism” immediately below).

• For all legacy fleets (RTE and GHG pathways), Itron would submit an initial report to the

PAs on February 15, 2020 for the 2019 period. This report would provide developers

with information about their fleet performance in 2019 but would not be used to verify

compliance.


• Developers whose fleets are found to be in breach of the rules under the relevant legacy

pathway would be temporarily suspended by the PAs, meaning they would not be

permitted to submit new applications during a defined suspension period. Suspension

periods would advance only on days when the residential budget of the “controlling

PA”50 is open and would be proportional to the magnitude of the breach in compliance.

49 See Footnote 14. 50 See Footnote 17.

42

• Legacy commercial projects subject to the 260 cycle/year requirement would have the

requirement reduced to 130 cycles/year. Cycling requirements for all other projects

would remain the same (130/year for commercial, 52/year for residential).

Verification and Enforcement Mechanism

• PAs would use existing verification methods and handbook infraction language to

enforce GHG reductions and would grant developers opportunities to bring their fleets

into compliance prior to issuing infractions

• PAs would focus their enforcement on higher-emitting fleets

• Residential customers with legacy systems who enroll on an approved time-varying rate

would be exempt from enforcement

Additionally, Energy Solutions would list and rank developer fleets’ annual GHG performances

on SelfGenCA.com and CaliforniaDGStats.ca.gov, highlighting high-performing developers who

were successful in achieving GHG emissions reductions associated with their fleets.

D ISCUSSION

Strengthening SGIP Minimum RTE Verification and Enforcement

SGIP’s overall legacy fleet poses a special policy challenge to upholdthe goal of upholding SGIP’s

long-standinglongstanding statutory GHG emissions reductions goals without retroactively

modifying the rules under which older projects were installed and operating.. For this reason,

staff proposesinitially proposed to maintain the current RTE requirement for the overall SGIP

legacy fleet. Projects with application dates before the new SGIP rules go into effect would

default into the existing RTE compliance pathway, unless they chooseprovide two pathways to

opt in to the GHG reduction pathway.

While the existing minimum RTE requirement would remain unchanged under our proposal,

staff is concerned about the efficacy of verification and enforcement measures currently in

place to ensure compliance. Based on the findings of past program evaluations, SGIP non-PBI

and residential compliance: legacy projects consistently fail to meet this requirement. The 2017

SGIP Storage Impact Evaluation shows that the mean RTE for non-residential non-PBI (systems

less than 30 kW) was 51% and only 38% for residential projects.51 Regardless of how effective a

tool the minimum RTE could either adhere to the existing RTE standard may at the time of

application or may not be for predicting energy storage system’s GHG performance, SGIP

51 See 2017 SGIP Advance Energy Storage Impact Evaluation (Itron) at 1-6.

43

applicants receiving incentives are bound by all program rules in effect at the time they entered

into contract with and received incentives from SGIP. Therefore, Staff proposes more stringent

RTE verification and enforcement processes to improve compliance.

To encourage developers to choose to uphold the statutory requirement for all SGIP

systemselect to reduce GHGs, staff proposes stricter non-compliance penalties for legacy . Each

pathway would be verified and enforced via a “fleet compliance” structure, in which developers

with ten or more projects in a given pathway would form fleets that opt in to the RTE

compliance pathway. Stricter penalties for RTE fleets are also warranted because SGIP legacy

projects have had several years of experience in energy storage operations, and developers

have had ample time since the annual RTE requirement was introduced to adjust their storage

operations to follow SGIP rules.

Staff proposes engaging Itron’s expertise to develop a more robust methodology for verifying

RTE compliance. In addition, we propose semiannual progress check-ins with developers to

determine whether operational adjustments will and be needed to achieve the annual

requirement. Furthermore, while Itron may employ a sampling strategy to assess whether a

developer’s fleet of projects is operating in accordance with SGIP program rules, once a

performance report is filed with the PAs, a large sample of projects in breach of the RTE

requirement will necessitate a request for project-level performance data. The latter will be

used to determine the magnitude of the developer’s breach in compliance and the appropriate

penalties.

Staff proposes to calculate RTE compliance as a percentage of the total capacity (MW) of a

developer’s fleet. As an illustrative example, a developer with two projects, one with 1 kW and

one with 1 MW capacity, would be evaluated based on the actual RTE achieved for 1.001 MW

of capacity. In this example, two extremely divergent system sizes are used to demonstrate the

weight given to larger projects when evaluating a legacy fleet’s RTE compliance. By basing the

developer’s overall RTE compliance on the sum of all projects’ capacities, larger projects, with

more potential for causing GHG and grid impacts, take on a size-proportionate obligation to

maintain RTEs at the required level.

Legacy Projectsto report charge and discharge data, Itron and the GHG Signal

n staff’s view, although the RTE requirement proved to be an imperfect proxy for a storage GHG

emission reduction standard, compliancePAs would track each fleet’s performance, and

44

developers with longstanding statutory requirements52fleets that all SGIP technologies reduce

GHG emissions should remain a program priority.

When considering possible solutions to meet the requirements of the December 2017 ACR, the

WG determined early on that, to ensure long-term market sustainability, customer value

proposition also merits consideration when optimizing storage operations for GHG savings. As a

result, the GHG co-optimization model was designed to, where possible, maximize both GHG

and customer bill savings.

As can be seen in the “GHG Signal Benefit Comparisons by Scenario” table taken from the WG

report (see figure below), compared to the “no GHG reduction model” for all scenarios–

including systems operating under old rates, the GHG signal co-optimization model nearly

always led to bill savings (“cost reductions”) while reducingincrease GHGs. This holds especially

true for commercial projects, which even under old rates53, were able to reduce GHGs and

achieve customer bill savings for about 24% of model-runs under a stand-alone storage

scenario and 40% of model-runs under a solar plus storage scenario. While these results seem

less than ideal given SGIP’s GHG reduction goals, the alternative “no GHG reduction” scenario

yielded 0% for standalone storage projects and 24% for paired-solar commercial projects

operating under old rates would be suspended from the program.

52 Senate Bill (SB) 412 (Kehoe, Stats. 2009, ch. 182) directed the California Public Utilities Commission

(Commission) to adopt rules that limit eligibility for the program to “distributed energy resources that the

commission, in consultation with the State Air Resources Board (ARB), determines will achieve reductions of

greenhouse gas emissions…” (Emphasis added.) SB 412 is codified in Public Utilities Code Section 379.6, as well as

other code sections. In response, SGIP adopted a minimum annual RTE, later updated in D.15-11-027, to ensure

energy storage projects receiving incentives reduce GHG emissions. 53 Legacy projects are most likely to be operating under old rates.

45

Figure 7: GHG signal benefit comparisons by scenario, reprinted from WG report (page 31 of

corrected version)

With residential solar-paired storage systems, the outcome is slightly better, as around 59% of

projects reduced GHGs when using the GHG signal under old rates.54 As discussed in the new

residential projects section, however, enforcing GHG reduction at this level is difficult given the

number of projects in this category, as is requiring these small individual projects to receive and

operate according to a GHG signal. Furthermore, the WG modeling results show that under the

old rates, the GHG signal co-optimization approach does not lead to customer bill savings in the

majority of cases.

Given the discussion above, staff proposes the GHG signal be made available to all legacy

systems as an opt-in pathway to reducing GHGs. This will allow developers the flexibility to

determine when their system profiles (onsite load, rate tariff, etc.) are conducive to meeting

the SGIP GHG reduction goal. Once this pathway is selected, however, the system will be

entered into a developer’s GHG emission reduction fleet, which will be subject to the relevant

enforcement and verification processes to ensure compliance.

SGIP’s legacy fleet poses a special policy challenge to the goal of upholding SGIP’s longstanding

statutory GHG goals without retroactively modifying the rules under which older projects were

54 Ninety-seven percent of all existing SGIP residential projects are paired with solar: SGIP Weekly Statewide

Report, August 10, 2018.

46

installed. For this reason, staff initially proposed to provide two pathways to program

compliance: legacy projects could either adhere to the existing RTE standard at the time of

application or elect to reduce GHGs. Each pathway would be verified and enforced via a “fleet

compliance” structure, in which developers with ten or more projects in a given pathway would

form fleets and be required to report charge and discharge data, Itron and the PAs would track

each fleet’s performance, and developers with fleets that increase GHGs would be suspended

from the program.

In comments on the original staff proposal and at the workshop, some stakeholders argued that

the RTE pathway does not ensure GHG reductions and that additional incentive is needed to

encourage projects to elect the GHG pathway. With regards to enforcement, some stakeholders

argued that the Commission should recognize the evolution and learning of industry

participants, and the relatively small amount of GHG emissions reductions and involved legacy

incentives at stake, and generally exempt legacy systems from the new, retroactive fleet

enforcement measures envisioned in the staff proposal, leaving them instead subject to existing

SGIP handbook infraction language at the time of their application submission. Others

supported staff’s proposal for fleet verification because it clearly defined verification and

enforcement measures and held projects accountable for achieving already existing GHG

reduction program rules. Additionally, the PAs shared that the administrative cost of

implementing the fleet compliance approach would be roughly one half to one additional staff

person per PA territory and $300,000 in consultant costs for initial implementation, and that

ongoing costs would be lower.

Staff revised its recommendations for legacy projects based on this feedback and additional

analysis. Below, staff discusses the rationale for using existing handbook language to enforce

GHG reductions rather than a comprehensive fleet compliance approach with automatic

suspensions, and for eliminating the RTE pathway for all legacy projects and instead requiring

them to reduce GHGs at the developer fleet level.

Using Existing Handbook Language for Verification and Enforcement

Staff recommends using existing handbook language to enforce rules on legacy projects rather

than implementing a comprehensive fleet compliance approach. Our rationale for this is the

large administrative cost of implementing the previous approach relative to the GHG impact of

legacy projects and recent successes in using existing handbook language to enforce program

rules.

In time, the GHG impact of the legacy program is likely to be small compared to the GHG impact

of new SGIP storage systems. Assuming 50-100 MW annual growth in the storage program and

47

5 kg/kWh GHG average reduction rate for new projects, GHG reductions from new projects will

outweigh GHG increases from legacy projects within 1.5-3 years.

Based on information shared by the PAs, staff conservatively estimates that the comprehensive

fleet compliance approach would cost ratepayers roughly $1.85 M to implement over five

years: $600,00055 to implement initially, and $250,00056 annually on an ongoing basis. The

following tasks would be required:

• Build a mechanism to automatically track developers with 10 or more new projects

statewide and communicate that back to Itron in order to develop participants in fleet

compliance, including a mechanism to track which projects elect which pathway

• Collect data twice a year on 200+ developers projects

• Develop data formatting requirements

• Determine how to treat developers that provide sub-par data or do not provide data at

all

• Create and execute a communication plan

• Develop a process for Itron to communicate who is out of compliance and to what

magnitude

• Create a database solution to implement suspension rules for affected developers

Dividing the administrative cost of a fleet compliance approach by the legacy program’s 2017

GHG impact, ratepayers would pay a minimum of roughly $250 per avoided ton annually (for

context, the IRP price of carbon is $150/ton.) The cost per avoided ton would likely be higher in

reality because fleet compliance would not result in all emissions from legacy projects being

avoided. Also, the figure does not include costs incurred by developers to semi-annually report

charge and discharge data or by Energy Division to oversee implementation. Given the large

administrative cost of fleet compliance relative to the GHG impact of legacy projects, staff

instead supports leveraging existing verification work performed by Itron through the annual

impact evaluation process to identify non-compliant fleets, as is done currently for RTE.

With regards to enforcement, the PAs have recently demonstrated that they can successfully

use existing handbook infraction language to bring legacy fleets into compliance with program

rules. Using existing handbook language would optimize use of ratepayer funds by giving the

PAs discretion to focus their enforcement efforts on high-emitting developers, and has the

55 $300,000 for Energy Solutions and Itron work, plus 4 x ($75,000 = 50% of salary and benefits for 1 new hire) 56 $100,000 for ongoing Energy Solutions and Itron work, plus 4 x ($37,500 = 25% of salary and benefits for 1 new

hire)

48

added benefit of not making retroactive changes to enforcement rules. For these reasons, staff

recommends using existing handbook language rather than adopting a fleet compliance

approach.

Because projects would be asked to comply with a requirement to reduce GHG that was not in

the handbook at the time they joined the program, staff believes that PAs should grant

developers opportunities to bring their fleets into compliance prior to issuing infractions. We

recommend the PAs propose changes to existing handbook infraction language to explicitly

adopt this interim step in the enforcement process for legacy projects.

Requiring Legacy Projects to Reduce GHGs

The task of bringing legacy projects into compliance with longstanding statutory requirements

to reduce GHGs is challenging due to the fact that the SGIP program rules in place at the time

the projects entered the program did not guarantee GHG reductions if followed. We know now

that RTE is an imperfect proxy for GHGs and that some projects increased GHGs even when

they complied with RTE rules. Unfortunately, we do not have data on what the GHG impact of

enforcing RTE rules would be. Legacy projects seeking to meet existing RTE standards would

have to cycle more. If projects are enrolled on a time-varying rate with TOU periods aligned

with grid emissions, increased cycling could produce GHG reductions. If projects are enrolled on

an old TOU or tiered rate, increased cycling could produce GHG increases.

The original staff proposal recommended allowing projects to elect either the RTE or GHG

pathway, and relied on slightly more punitive fleet enforcement of the RTE pathway to

encourage projects to elect the GHG pathway. This “stick” no longer exists if the Commission

adopts the recommendation above to use existing handbook language to enforce rules. Thus, if

given the option, projects are likely to stay under the compliance regime they know: the RTE

pathway.

Given the unknown GHG impact of enforcing RTE, and the likelihood of projects to elect the RTE

pathway, staff recommends the Commission eliminate the RTE requirement and instead

require all legacy projects to reduce GHGs. This would bring the legacy program into

compliance with the statutory requirement to reduce GHGs that was in place at the time the

projects applied. It would also eliminate the RTE metric from the program as an imperfect proxy

for GHG reductions but would maintain the incentive for projects to minimize parasitic losses.

49

Although requiring legacy projects to reduce GHGs would represent a change to operational

rules in the handbook, the requirement to reduce GHGs has existed in statute since 2009.57

Furthermore, WG modeling found that projects using the GHG signal are able to achieve GHG

reductions without sacrificing bill savings, so the financial impact of the rule change could be

minimal for many customers with legacy systems (for more discussion on the customer

financial impact of new GHG rules, see “Customer Bill Impacts of Proposals” on page 12). For

those customers who would be severely impacted, developers would have the opportunity to

continue operating those systems as before and counter their emissions with GHG reductions

from elsewhere in the fleet.

57 SB 412 (2009) limits the eligibility for SGIP incentives pursuant to the program to distributed energy resources

that the commission, in consultation with the state board, determines will achieve reduction of greenhouse gas

emissions pursuant to the California Global Warming Solutions Act of 2006.

50

Appendix A: Public Utilities Code Section 379.6

PUBLIC UTILITIES CODE - PUC

DIVISION 1. REGULATION OF PUBLIC UTILITIES [201 - 3260]

( Division 1 enacted by Stats. 1951, Ch. 764. )

PART 1. PUBLIC UTILITIES ACT [201 - 2120]

( Part 1 enacted by Stats. 1951, Ch. 764. )

CHAPTER 2.3. Electrical Restructuring [330 - 400]

( Chapter 2.3 added by Stats. 1996, Ch. 854, Sec. 10. )

ARTICLE 6. Requirements for the Public Utilities Commission [360 - 380.5]

( Article 6 added by Stats. 1996, Ch. 854, Sec. 10. )

379.6.

(a) (1) It is the intent of the Legislature that the self-generation incentive program increase

deployment of distributed generation and energy storage systems to facilitate the integration

of those resources into the electrical grid, improve efficiency and reliability of the distribution

and transmission system, and reduce emissions of greenhouse gases, peak demand, and

ratepayer costs. It is the further intent of the Legislature that the commission, in future

proceedings, provide for an equitable distribution of the costs and benefits of the program.

(2) The commission, in consultation with the Energy Commission, may authorize the annual

collection of not more than double the amount authorized for the self-generation incentive

program in the 2008 calendar year, through December 31, 20192024. The commission shall

require the administration of the program for distributed energy resources originally

established pursuant to Chapter 329 of the Statutes of 2000 until January 1, 20212026. On

January 1, 20212026, the commission shall provide repayment of all unallocated funds

collected pursuant to this section to reduce ratepayer costs.

(3) The commission shall administer solar technologies separately, pursuant to the California

Solar Initiative adopted by the commission in Decisions 05-12-044 and 06-01-024, as modified

by Article 1 (commencing with Section 2851) of Chapter 9 of Part 2 of Division 1 of this code

and Chapter 8.8 (commencing with Section 25780) of Division 15 of the Public Resources Code.

(b) (1) Eligibility for incentives under the self-generation incentive program shall be limited to

distributed energy resources that the commission, in consultation with the State Air Resources

Board, determines will achieve reductions in emissions of greenhouse gases pursuant to the

51

California Global Warming Solutions Act of 2006 (Division 25.5 (commencing with Section

38500) of the Health and Safety Code).

(2) On or before July 1, 2015, the commission shall update the factor for avoided greenhouse

gas emissions based on both the most recent data available to the State Air Resources Board

for greenhouse gas emissions from electricity sales in the self-generation incentive program

administrators’ service areas as well asand current estimates of greenhouse gas emissions over

the useful life of the distributed energy resource, including consideration of the effects of the

California Renewables Portfolio Standard.

(3) The commission shall adopt requirements for energy storage systems to ensure that eligible

energy storage systems reduce the emissions of greenhouse gases.

(c) Eligibility for the funding of any combustion-operated distributed generation projects using

fossil fuel is subject to all of the following conditions:

(1) An oxides of nitrogen (NOx) emissions rate standard of 0.07 pounds per megawatthour and

a minimum efficiency of 60 percent, or any other NOx emissions rate and minimum efficiency

standard adopted by the State Air Resources Board. A minimum efficiency of 60 percent shall

be measured as useful energy output divided by fuel input. The efficiency determination shall

be based on 100 -percent load.

(2) Combined heat and power units that meet the 60-percent efficiency standard may take a

credit to meet the applicable NOx emissions standard of 0.07 pounds per megawatthour. Credit

shall be at the rate of one megawatthour for each 3,400,000 British thermal units (Btus) of heat

recovered.

(3) The customer receiving incentives shall adequately maintain and service the combined heat

and power units so that during operation the system continues to meet or exceed the efficiency

and emissions standards established pursuant to paragraphs (1) and (2).

(4) Notwithstanding paragraph (1), a project that does not meet the applicable NOx emissions

standard is eligible if it meets both of the following requirements:

(A) The project operates solely on waste gas. The commission shall require a customer that

applies for an incentive pursuant to this paragraph to provide an affidavit or other form of

proof that specifies that the project shall be operated solely on waste gas. Incentives awarded

pursuant to this paragraph shall be subject to refund and shall be refunded by the recipient to

the extent the project does not operate on waste gas. As used in this paragraph, “waste gas”

52

means natural gas that is generated as a byproduct of petroleum production operations and is

not eligible for delivery to the utility pipeline system.

(B) The air quality management district or air pollution control district, in issuing a permit to

operate the project, determines that operation of the project will produce an onsite net air

emissions benefit compared to permitted onsite emissions if the project does not operate. The

commission shall require the customer to secure the permit prior to receiving incentives.

(d) In determining the eligibility for the self-generation incentive program, minimum system

efficiency shall be determined either by calculating electrical and process heat efficiency as set

forth in Section 216.6, or by calculating overall electrical efficiency.

(e) Eligibility for incentives under the program shall be limited to distributed energy resource

technologies that the commission determines meet all of the following requirements:

(1) The distributed energy resource technology shifts onsite energy use to off-peak time

periods or reduces demand from the grid by offsetting some or all of the customer’s onsite

energy load, including, but not limited to, peak electric load.

(2) The distributed energy resource technology is commercially available.

(3) The distributed energy resource technology safely utilizes the existing transmission and

distribution system.

(4) The distributed energy resource technology improves air quality by reducing criteria air

pollutants.

(f) Recipients of the self-generation incentive program funds shall provide relevant data to the

commission and the State Air Resources Board, upon request, and shall be subject to onsite

inspection to verify equipment operation and performance, including capacity, thermal output,

and usage to verify criteria air pollutant and greenhouse gas emissions performance.

(g) In administering the self-generation incentive program, the commission shall determine a

capacity factor for each distributed generation system energy resource technology in the

program.

(h) (1) In administering the self-generation incentive program, the commission may adjust the

amount of rebates and evaluate other public policy interests, including, but not limited to,

ratepayers, energy efficiency, peak load reduction, load management, and environmental

interests.

53

(2) The commission shall consider the relative amount and the cost of greenhouse gas

emissions reductions, peak demand reductions, system reliability benefits, and other

measurable factors when allocating program funds between eligible technologies.

(i) The commission shall ensure that distributed generation resources are made available in the

program for all ratepayers.

(j) In administering the self-generation incentive program, the commission shall provide an

additional incentive of 20 percent from existing program funds for the installation of eligible

distributed generation resources manufactured in California.

(k) The costs of the program adopted and implemented pursuant to this section shall not be

recovered from customers participating in the California Alternate Rates for Energy (CARE)

program.

(l) The commission shall evaluate the overall success and impact of the self-generation

incentive program based on the following performance measures:

(1) The amount of reductions of emissions of greenhouse gases.

(2) The amount of reductions of emissions of criteria air pollutants measured in terms of

avoided emissions and reductions of criteria air pollutants represented by emissions credits

secured for project approval.

(3) The amount of energy reductions measured in energy value.

(4) The amount of reductions of customer peak demand.

(5) The ratio of the electricity generated by distributed energy resource generation projects

receiving incentives from the program to the electricity capable of being produced by those

projects, commonly known as a capacity factor.

(6) The value to the electrical transmission and distribution system measured in avoided costs

of transmission and distribution upgrades and replacement.

(7) The ability to improve onsite electricity reliability as compared to onsite electricity reliability

before the self-generation incentive program technology was placed in service.

(Amended by Stats. 2016, Ch. 658, Sec. 1. (AB 1637) Effective January 1, 2017.)

(m) On and after January 1, 2020, generation technologies using nonrenewable fuels shall not

be eligible for incentives under the self-generation incentive program.

54

SEC. 2. No reimbursement is required by this act pursuant to Section 6 of Article XIII B of the

California Constitution because the only costs that may be incurred by a local agency or school

district will be incurred because this act creates a new crime or infraction, eliminates a crime or

infraction, or changes the penalty for a crime or infraction, within the meaning of Section 17556

of the Government Code, or changes the definition of a crime within the meaning of Section 6

of Article XIII B of the California Constitution.

55

Appendix B: OSESMO Analysis of PG&E E6 with New TOU Periods

https://osesmo.shinyapps.io/osesmo_results_viewer/

Appendix C: OSESMO LADWP Residential TOU Rate Analysis

57

58

59

60

61

Documents

Self -Generation Incentive Program Greenhouse Gas Staff Proposal · 2018-12-28 · 5 multiplied by the applicable efficiency (V DC x Amp-Hours x Applicable Efficiency) (see SGIP Handbook,