Volume II Market Redesign & Technology Upgrade

Transcription

1 Application No.: Exhibit No.: Witnesses: A SCE- L. Oliva (U -E) Volume II Market Redesign & Technology Upgrade Before the Public Utilities Commission of the State of California Law-# Rosemead, California September 1, 00

2 Market Redesign & Technology Upgrade Table Of Contents Section Page Witness I. Market Redesign & Technology Upgrade...1 L. Oliva A. Summary Readiness for MRTU Release MRTU Release 1A (MAP) Readiness.... Mechanism for Resolving Outstanding Technical and Operational Issues... B. Current Status of MRTU and Overview of MRTU Readiness MRTU Release 1 Expectations.... MRTU Release 1A (MAP) Expectations... C. Summary of CAISO s Planned Demand Response Products for MRTU Non-Participating Load.... Participating Load (Dispatchable Demand Response, or DDR).... Proxy Demand Resource... D. Immediate Readiness for Serving the CAISO s MRTU Markets MRTU Release MRTU Release 1A (MAP)...1. Summary of What Can and Cannot Be Accomplished Immediately...1 a) What Can be Done Immediately...1 b) What Cannot be Done Immediately and Why...1 c) Schedule for Making Necessary Corrections...1 d) Limits to Estimating Transition Costs...1 -i-

3 Market Redesign & Technology Upgrade Table Of Contents (Continued) Section Page Witness E. Issues and Barriers Issues and Barriers to DR in MRTU Release a) SC-to-SC Trades...1 b) Day-Ahead/Real-Time Tradeoffs...1 c) Changing Role for DR Aggregators...1. Issues and Barriers to DR in MRTU Release 1A (MAP)...1 a) CAGS and Associated Costs...0 b) Operations Changes...0 c) Unprecedented Detailed Mapping Requirements...0. Technical and Operational Issues Pertaining to Both PDR and PL...1 a) DR LAP and CAG...1 b) Accounting Issues...1 c) DR Certification... d) Direct Access... (1) Self-Scheduling of Direct Access Load... () Limits of DR Aggregator Bids... () DA Customer Accounting... () DA LAP and CAG... e) IT Systems Impact... f) Notification... g) Incentives Alignment.... Technical and Operational Issues Affecting PDR... ii

4 Market Redesign & Technology Upgrade Table Of Contents (Continued) Section Page Witness a) Adapting Capacity Bidding Program (CBP)... b) Single Product Portfolio... c) DR Values Mismatches.... Issues and Barriers Associated with Participating Load... a) PL and RUC Value... b) Forecasting... c) Nodal Load Forecasting... d) Metering and Telemetry... e) Geographic Customer Differentiation... f) DR Value Misalignments... g) Cost of Required Information Granularity... h) Accounting for RUC... i) Detailing Functional Requirements... j) Edison SmartConnect.... A Shared Effort to Overcome Deployment Challenges for MRTU DR Products... a) Working Group... (1) User Guide... () DR Program Consolidation... () Map Process Changes... () Map IT Requirements... () WG Meeting Schedule & Process Output... () Final WG Output Package...0 iii

5 Market Redesign & Technology Upgrade Table Of Contents (Continued) Section Page Witness () IOU Deployment Planning and Programming...0. Institutional Timing Considerations Affecting IOU and CAISO Efforts...1 a) Grid Operations Change Durations...1 b) Information System Change Durations...1 c) Asset Change Durations...1 d) Billing and Settlements Change Durations...1 e) Difficulties in Compressing Institutional Time Requirements...1 f) FERC...1 F. Proposed SCE Actions to prepare for MRTU Working Group.... Transitioning existing Spinning Reserve Pilot for PL... a) Background... b) Spinning Reserves Pilot Accomplishments... c) Transitioning to support Participating Load... (1) Objectives of the Participating Load Pilot... () Scope of Work... () Proposed Schedule... d) Program Proposal... e) Pilot Project Learning... f) Budget to Implement and Administer Program.... DR Resource Portal... iv

6 Market Redesign & Technology Upgrade Table Of Contents (Continued) Section Page Witness a) Background... b) Program Proposal... c) Budget to Implement and Administer Program...0. MRTU and SmartConnect Deployment...0 G. Other Pilots Considered Demand Response and Resource Management...1. General Definitions Pertaining to Intermittent Resources and Demand Response.... Intermittent Resources (IR)... a) Resource Characteristics... b) Challenges with Intermittent Resources... c) Solutions: Principles and Objectives... d) Issues and Barriers to Solving IR Challenges... e) Proceeding Forward to Address IR Challenges... f) Summary of What Can and Cannot Be Done Now and Why...0. Small Load Aggregation...1 a) SLA Resource Characteristics...1 b) Challenges with SLA Resources... c) Solutions: Principles and Objectives... d) Issues and Barriers to Solving SLA Challenges... e) Proceeding Forward to Address SLA Challenges... H. Concluding Points... v

7 I. Market Redesign & Technology Upgrade A. Summary This document outlines actions Southern California Edison Company (SCE) has taken and plans to take to be ready for the California Independent System Operator (CAISO) Market Redesign and Technology Upgrade (MRTU) Release 1and for contributing demand response (DR) into the CAISO s wholesale markets. This document also outlines what SCE plans to do to be ready to support full participation of DR in MRTU Release 1A, which is now referred to by the CAISO as Markets and Performance (MAP). 1. Readiness for MRTU Release 1 SCE is prepared to use DR in MRTU Release 1 markets as these markets are activated. The CAISO s estimate for deployment of MRTU Release 1 is the first or second quarter of 00. SCE s readiness for MRTU Release 1 primarily involves utilizing existing manual work-around solutions leveraging Excel spreadsheets for posting and using available DR in CAISO wholesale markets. SCE also proposes to work with CAISO and attempt to convert its existing Demand Response Spinning Reserve pilot into a pilot project for bidding DR Participating Load (PL) into CAISO s wholesale markets. A summary description of the PL pilot can be found in Section I.F. of this Volume and in Appendix J. SCE is hopeful that the PL pilot project will be ready for use in CAISO s wholesale markets by Summer 00, assuming that both the CAISO s present estimate for MRTU Release 1 deployment holds, and the proposed PL requirements are updated as the CAISO has indicated. A summary description of the technical and operating issues affecting Non Participating Load (NPL), PL and Proxy Demand Response (PDR) is included in Section I.E.1 of this Volume. Ensuring pilot project readiness in time for the Summer of 00 requires timely action and participation by the CAISO and the California Public Utilities Commission (Commission). 1

8 MRTU Release 1A (MAP) Readiness The CAISO expects to enhance its initial MRTU Release 1 approximately one year following deployment of MRTU Release 1. Accordingly, MRTU MAP expected to be launched in the first or second quarter of 0. SCE expects to be ready to bring its DR programs into alignment with one of the CAISO s three DR product categories under MAP NPL, PL and PDR. A summary description of the technical and operating issues affecting NPL, PL and PDR is included in Section I.E. of this Volume.. Mechanism for Resolving Outstanding Technical and Operational Issues SCE encourages the CAISO and the Commission to support the reactivation of the DR market design process that was begun in late 00, and that Working Group (WG), which was responsible for DR market design, be used as the focal point for working through questions, issues and barriers that may reduce the effectiveness of MAP if not addressed before its activation. This document discusses in detail what issues and barriers SCE considers critical for WG to address. The assumption is that the WG process should deliver detailed user guides, and/or other technical specification documents, that the Investor Owned Utilities (IOUs) will need in order to adapt retail DR programs to MAP requirements. B. Current Status of MRTU and Overview of MRTU Readiness The CAISO advocates the transition of retail DR programs into DR that bids into CAISO wholesale markets. The DR that bids into CAISO wholesale markets would receive a price cleared in CAISO wholesale markets, and MRTU would be used as a catalyst for this transition. The Commission has begun efforts to determine the reasonableness of transitioning DR programs to the three CAISO market products: NPL, PL and PDR. As noted in the Administrative Law Judge s Ruling Providing Guidance on Content and Format of 00-0 Demand Response Activity Applications, issued February, 00 ( February ALJ Ruling or Guidance Document ), MRTU Release 1 has adjusted its activation date several times due to technical issues identified through testing and other means, such as stakeholder processes. As of this

9 filing, MRTU Release 1 has no specific release date, although the CAISO indicates that it should be sometime in the first or second quarter of MRTU Release 1 Expectations As a result of MRTU Release 1 delays, it is unclear whether current technical requirements for Release 1 are the final specifications, or if more changes are likely before deployment. For planning purposes, SCE assumes that any further changes in MRTU Release 1 specifications will be minor. Based on information provided by the CAISO in various venues, SCE has the following expectations for how the year 00 will function when MRTU Release 1 is activated. MRTU Release 1 will be active before the Summer of 00 PDR may be available in time for the Summer of 00, but the ability of IOUs to make full use of it may be limited. CAISO staff indicated that it may have PDR ready for IOUs to use by the Summer of 00. However, while the product may be ready for use, there may be issues and barriers that minimize or, in some cases, prohibit IOU use of the product by the Summer of 00. Section I.E. and I.E. below detail issues that require resolution before PDR can be used. For example, the self schedule requirement within PDR conflicts with direct access load since the IOU is not the Load Serving Entity (LSE).. MRTU Release 1A (MAP) Expectations The CAISO estimates that MAP will be activated one year after the deployment of MRTU Release 1. Based on information provided by the CAISO in various venues, SCE has the following expectations for how the year 0 will function when MAP is activated. MAP will be active before the Summer of 0 PDR is available, and the IOUs can fully use it as a product. SCE s expectations are that for 0, all necessary processes, requirements and technical specifications of the CAISO for executing DR transactions as NPL, PDR, and/or PL are available and fully tested.

10 It is important to note that the CAISO changed technical specifications several times for MRTU Release 1, which resulted in costs to SCE for making adjustments to conform to the modifications. There is no reason to conclude that a similar, necessary, iterative process of technical specification refinements, testing of protocols, and associated recoding requirements that result there from will not be part of the deployment of MAP a year following the activation of MRTU Release C. Summary of CAISO s Planned Demand Response Products for MRTU The CAISO runs wholesale markets in which it hopes to see DR participate: Day-Ahead Hour Ahead Scheduling Protocol (HASP) Real-Time The CAISO also values resources and compensates market participants in its Residual Unit Commitment (RUC) process that is part of the overall integrated market system of the CAISO. In this Section, the characteristics of each of the CAISO s three DR products are described to ensure a shared understanding of what SCE considers the basis for its proposed deliverables. 1. Non-Participating Load Definition. Under MRTU Release 1, NPL is load that does not submit bids for participation in the CAISO HASP or Real-Time markets, and submits bids for energy only. Bid or not, NPL constitutes most of the currently available DR resources in California. NPL also may be characterized as bundled IOU customers and direct access customers where DR is captured through existing Commission-approved IOU retail programs that are targeted for specific markets (e.g., industrial, commercial, residential, agricultural), customer segments (e.g., large and small industrial and commercial), and energy efficiency needs (e.g., agricultural pumping efficiency and DR, residential energy efficiency programs). NPL Limitations. NPL is demand responsive even though it generally does not bid into the CAISO s wholesale capacity markets. Historically, DR has operated as NPL because its ability to participate has been limited by the CAISO and, more importantly, because the DR operates within the framework of existing Commission-approved retail DR programs.

11 Value of Retail DR. California electricity consumers are benefiting from the effects of DR programs and are being compensated for their participation through IOU tariffs. Typically these IOU tariffs are based on fixed incentive mechanisms updated and filed with the Commission. There may be opportunity to modify or reassess the compensation for this DR by bidding it into the CAISO wholesale markets to capture its market value. NPL Bidding. NPL can bid energy into the CAISO s Day-Ahead market, but is not allowed to participate in Ancillary Services or RUC; nor can it participate in Real-Time Energy and Ancillary Services. NPL has no interval metering or telemetry requirement for participation in the Day- Ahead-Market (DAM) Energy market. Settlements. NPL is the most aggregated in geographic terms. NPL granularity is limited to Default LAPs and therefore settles based on pricing at Default LAPs. Price Responsive NPL. As MRTU Release 1 has evolved, the CAISO s interests in incorporating NPL into its wholesale markets have led to enhancements to the core definition of NPL. For instance, the CAISO is interested in price-responsive NPL, which manifests in IOU programs such as Critical Peak Pricing, Demand Bidding Programs, Capacity Bidding Programs, and Demand Response Contracts. The CAISO characterizes price-responsive NPL as demand submitted as price-responsive energy bids at LAPs in the DAM. It seeks to adjust its RUC targets after the DAM executes, based on determined available Real-Time DR. Real-Time deviations from the DAM schedules are settled at the Real-Time Imbalance Energy price for LAPs. NPL Accounting. IOUs, or Load Serving Entities (LSEs) are responsible for DR by NPL in Real-Time. NPL may be counted as meeting Resource Adequacy Resources (RAR) requirements based on criteria established by the Commission and other relevant regulatory entities. Treatment of NPL for RAR does not require a PL agreement with the CAISO. 1 1 The description of NPL was based on the following documents: MRTU Release 1 & Functionality Related to Demand Response, by Jim Price, Lead Engineering Specialist, Market and Product Development, CAISO, November, 00; (Continued)

12 Participating Load (Dispatchable Demand Response, or DDR) Definition. PL is defined by the CAISO as an entity (including an entity with Pumping Load, providing Curtailable Demand) which has executed a Participating Load Agreement (PLA) to comply with all applicable provisions of the CAISO Tariff, as they may be amended from time to time. It should be noted that the CAISO has renamed PL to be Dispatchable Demand Response (DDR). This is an attempt to reduce confusion over the differences between NPL and PL. Here, however, the term PL will be used; since the adoption of the new terminology is early stage, thus not yet widely used by market participants and stakeholders. PL Characteristics. PL provides Curtailment Demand under the PLA with the CAISO. The CAISO s view is that the PLA is sufficiently broad that its existing FERC approved tariff is sufficient and no additional FERC filings are necessary prior to using the PLA for adoption of both PL and PDR products. PL requires CAISO technical certification and is expected to respond to CAISO dispatch instructions; PL can participate in the DAM and Ancillary Services markets, as well as Real Time Energy Imbalance Market (RTM). MRTU Release 1 includes PL of pumping load associated with pumped storage, single pumping or non-pumping load, and aggregated pumping and non-pumping load. Deployment Timing and Pilot Projects. PL engagement in the CAISO s wholesale markets is planned for MRTU Release 1 but the full scope of PL may not be achieved until MAP. Continued from the previous page Demand Response Working Group : Demand Response in MRTU Release 1 and Design Details for Post-Release 1, by Jim Price, Lead Engineering Specialist, CAISO, June 1, 00; CAISO-IOU Technical Design Session, presentation by Farrok Rahimi, Vice President, Market Design and Consulting of Open Access Technology International, Inc., and John Goodin, Lead, Demand Response, CAISO, July 0, 00, Slide. CAISO Demand Response Resource User Guide: Guide to Participation in MRTU Release 1, November, 00, Version.0, p. 0 MRTU Release 1 Provisions to Support Demand Response Programs, by Farrokh Rahimi, Vice President Market Design and Consulting, Open Access Technology International, Inc. and John Goodin, Lead, Demand Response, California ISO, CAISO-IOU Demand Response Technical Design Session, July 0, 00, Slide.

13 Accordingly, SCE is proposing to adapt the existing DR Spinning Reserves pilot as a Participating Load pilot by the Summer of 00, aligned with the expected deployment of MRTU Release 1. See Section I.F. as well as Appendix J for a description of SCE s PL pilot project. Markets Served with PL. PL as a DR resource may bid into the DAM and its RTM under MRTU Release 1. PL is eligible to provide Non-Spinning Reserves as well as Imbalance Energy. PL is excluded from participation in the CAISO s RUC process, although PL will be able to offer RUC capacity and Ancillary Services in Real-Time with the enhanced functionality of MAP. Settlements. PL geographic specificity (and related information granularity) is anticipated to disaggregate to the individual network nodal point and is settled using Locational Marginal Pricing, which is at the core of the MRTU. Interval Meter and Telemetry Requirements. For PL to bid into the DAM and RTM for Energy and/or Ancillary Services, it must have an interval meter, which must be less than or equal to 1 minute intervals. Also, for PL to participate in the DAM and RTM for Energy and/or Ancillary Services, it must have telemetry capability. For PL serving Ancillary Services markets for Non-Spinning Reserve ( Non-Spin ), the telemetry requirements are as follows: One-minute polling capability between the Load and an Aggregated Load Meter Data Server (ALMDS). Scanning of the in-use Data Processing Gateway (DPG), or Energy Communication Network (ECN), which may be Internet based, every four seconds by the CAISO Energy Management System (EMS). WECC Considerations. PL for Ancillary Services must conform to WECC standards pertaining to resources providing Non-Spin. WECC, and its parent, NERC, are in the process of CAISO-IOU Technical Design Session, presentation by Farrok Rahimi, Vice President, Market Design and Consulting of Open Access Technology International, Inc., and John Goodin, Lead, Demand Response, CAISO, July 0, 00., Slide. Ibid., Slide 1.

14 1 1 1 reviewing and most likely revising the technical specifications pertaining to resources for Non-Spin in order to enable more participation from DR resources. It is unknown how long this review/revisions process may take. PL and Location Specific DR. A distinctive feature of PL is the CAISO s focus on aligning DR with locational requirements, such as helping mitigate load pocket congestion and/or instabilities. Where NPL is generally limited to the highest level of nodal aggregation the Default LAP PL can be disaggregated into several more detailed configurations Local Capacity Areas (LCAs), Custom Load Aggregation Points (CAGs), and Nodes, which are essentially aligned with the core LMP topology of the CAISO s network. These disaggregation characteristics are important to appreciate because the PDR product, which is discussed in the next Section, adopts a specific level of geographic granularity for DR bid purposes and uses another level of aggregation for settlement purposes. PL as Generator Proxy. PL can be viewed as an equivalent generation resource as a DR resource. The CAISO defines the comparability of PL to generation in Table I-1. Source for the Table is: Demand Response Working Group : Demand Response in MRTU Release 1 and Design Details for Post-Release 1, by Jim Price, Lead Engineering Specialist, CAISO, June 1, 00, unnumbered page.

15 Line No. Participating Load Resource Table I-1 Generator Resource 1. Base Load Schedule Base Load. Minimum Load Reduction Minimum Generator Output. Minimum Load Maximum Generator Output. Load Reduction Initiation Time Start-Up Time. Minimum Base Load Reduction Time Minimum Time Between Curtailments. Minimum Load Reduction Time Minimum Up Time. Maximum Load Reduction Time Maximum Daily Energy Limit. Minimum & Maximum Daily Energy Limit Maximum Daily Energy Limit. Maximum Daily Curtailments Maximum Number of Curtailments. Load Drop Rate Ramp Up Rate. Load Pickup Rate Ramp Down Rate 1. Load Reduction Initiation Cost Start-Up Cost 1. Minimum Load Reduction Cost Minimum Load Cost 1. Proxy Demand Resource The CAISO s new DR related product, Proxy Demand Resource (PDR), emanates from the recognition that NPL and PL functionality may be insufficient for capturing and integrating priceresponsive DR programs into MRTU Release 1. Purpose of PDR. PDR is designed to capture and integrate DR programs that are embedded in IOU loads (as NPL) where it can be difficult to anticipate the MW quantities of DR from price-responsive DR programs before the DAM executes, and before submission of RUC adjustments to the CAISO. PDR as Proxy Generator. PDR-based DR is load that is scheduled using the Default LAP as NPL. The DR MW quantity is included in the Default LAP and bid as a price taker, but unlike NPL

16 the DR bids are segmented into required time differentiated stages and are provided to the market using a separate proxy generator resource identifier. Scheduling of PDR. To support the scheduling of price-responsive DR, the NPL DR Provider will register, under a separate unique Resource ID, the portion of its demand that is priceresponsive. This separate component of the demand portfolio is registered as a generator resource which is the PDR. Bidding PDR. The PDR bid must be submitted for aggregations of DR loads within a Local Capacity Area (LCA), which is a step down in granularity from the Default LAP level. The options for aggregating load as PDR include loads within an LCA; or for the remaining area in a Default LAP that is not within a LCA; or as DR resources for smaller areas within a LCA, such as a Node. It is important to appreciate the differentiation of PDR within an NPL bid, because such differentiation is a process and systems challenge as discussed later in this Volume. The CAISO notes that a PDR s energy bid curve may include segments from different DR programs operated by the DR Provider, at the DR Provider s discretion, provided that the associated DR programs apply to the same LCA. The segment identification among DR programs packaged for an energy bid curve is irrelevant to the CAISO, but may be of significance to the DR Provider for internal accounting purposes (and associated allocation of revenues from the energy bid clearing the market). Limits to Market Participation. The proxy generator that derives from the process is limited with respect to what markets it can be bid into. Specifically, a PDR proxy generator cannot be used for provisioning Ancillary Services because the principal focus of the PDR model is to facilitate the integration of dispatchable DR programs into the DAM. It is important to understand that a proxy generator in the PDR model is essentially a virtual generator because it can be created at the nodal level or as an aggregation of nodes within a LCA, even though the DR itself is scheduled at the IOU s Default LAP. PDR Scheduling, Dispatch and Settlement. The CAISO recognizes that PDRs can impact Locational Marginal Prices based on its scheduling and dispatch. But because the PDR is only a proxy

17 generator and not a true generation resource, the PDR is not a metered output resource, and hence it is cannot be cleared in the market as a true generator. The CAISO s PDR model gives the CAISO responsibility for determining how to define and assign a PDR schedule. The principal choices the CAISO will make are whether to treat the PDR schedule as essentially a non-schedule by omitting it from the data received for settlement, or in the alternative, to treat the PDR, proxy generator, final output as zero. This mechanism helps ensure that the PDR is not subject to settlement because the CAISO subtracts the schedule and dispatch of the PDR from the schedule of the underlying demand at the Default LAP. Accordingly, the CAISO will settle the PDR at the Default LAP price. Effectively, the DR quantity gets netted out of the IOU s Scheduling Coordinator demand quantity for the day-ahead settlement. The CAISO s objective in applying this mechanism is to ensure that PDR functionality can work within the limits of MRTU Release 1, prior to full functionality for DR being available in MAP. PDR and RUC. PDR will be used to reduce the CAISO s RUC procurement targets, but PDR is not a resource the CAISO will allow to be reserved in the RUC process. The rationale for this is that the PDR, as a proxy generator, is already scheduled in the CAISO s Integrated Forward Markets (IFM) and not in RUC so there is no basis for a RUC payment to the PDR. Why PDR. The actual DR value consequences need to be understood more fully before it can be determined whether PDR is a truly viable product. Given how recently the PDR model was broached with IOUs, it is understandable that many questions remain to be answered. However, SCE considers the PDR model to have promise and is committed to making its design workable. The preceding description of PDR was based on an undated CAISO document Guidance Document on MRTU Release 1 Provisions to Support Demand Response Programs, CAISO Draft, for Discussion Purposes Only.

18 D. Immediate Readiness for Serving the CAISO s MRTU Markets 1. MRTU Release 1 SCE is ready to participate in MRTU Release 1 using the NPL manual work-around tools that have been developed already and are in limited use now for SCE s day-ahead retail DR programs. The CAISO seeks to have NPL register as PL, if possible. As stated previously, SCE has proposed to transform its existing DR Spinning Reserves pilot into a PL pilot by Summer of 00, assuming that SCE receives approval to do so by the Commission, and the CAISO s current estimate for MRTU Release 1 is accurate and the PL requirements are changed as the CAISO has communicated. SCE has not identified any significant technical or operational issues or barriers that would preclude or otherwise undermine its commitment to be ready for providing DR into the CAISO s wholesale markets upon the deployment of MRTU Release 1. However, this does not mean that there are no technical or operational issues concerning the manual work-around tools. These are noted in Section I.E.1 below.. MRTU Release 1A (MAP) SCE is not immediately ready for integrating its retail DR resources into MAP. There are important outstanding technical and operational considerations that require joint efforts on the part of the CAISO and the IOUs to work through in order to be ready by the expected release date of first or second quarter 0. The subject matter SCE considers important to jointly work through is discussed in Section I.E below.. Summary of What Can and Cannot Be Accomplished Immediately This Section summarizes what can and cannot be accomplished immediately with respect to readiness for MRTU Release 1. As noted, readiness for MAP, is dependent on working through technical and operational issues and barriers jointly with the CAISO and the IOUs. 1

19 a) What Can be Done Immediately Bid NPL. SCE can immediately bid NPL into MRTU Release 1 markets upon activation of the markets. Prepare Pilot Project. Upon approval by the Commission, SCE can immediately begin work preparing its existing Demand Response Spinning Reserves pilot project to extend into functioning as a pilot for PL for MRTU Release 1 markets. Resolve Technical and Operational Issues. SCE, along with the other IOUs and the CAISO, can begin working through the technical and operational issues that will impact the efficiency and effectiveness of MAP; and through this process, increase the likelihood that MAP will be successful from its start. Detail DR Program Transitions. SCE can immediately begin to detail the program and project task requirements for migrating its retail DR programs into wholesale products that can bid into the CAISO s MAP system so as to be ready to maximize DR participation upon the activation of MAP. b) What Cannot be Done Immediately and Why No Immediate Investments. The business process, systems, and work flow changes required to transition existing DR programs into products that can be bid into CAISO wholesale markets as PL or as PDR in MRTU Release 1 cannot be immediately invested in. The constraints on immediate action in these areas stem from: The need for Commission approval to spend what may be significant ratepayer funds to ensure readiness and effective transitioning of DR programs to PL or PDR products. Internal work required to detail the tasks and associated costs, and budgets that are necessary to place a request for approval before the Commission. Research the alignment of customer accounts to network nodes and detailed analysis of internal accounting, billing, and customer information system alignments to enable efficient organizing of DR for bidding from existing 1

20 retail DR programs; this research being an important feeder into the mapping of work requirements and associated budgets. Limits to Use of Existing Funds. There are limited funds absent approvals for the proposed 00-0 DR budgets, with which to work up required information. SCE is looking into appropriate means of allocating funds for the above efforts, but until such mechanisms are identified and approved by the Commission, funding constraints are real. Resolve Technical and Operational Issues. Planning, programming, and project tasking for MAP readiness cannot be immediately undertaken because of the limited funding available as noted above. There are also significant technical and operational issues that require joint efforts to work through between the IOUs and the CAISO, which must be done before work on MAP can make much more progress than has already been achieved. c) Schedule for Making Necessary Corrections The limitations on how quickly SCE and the other IOUs can affect DR retail program changes to maximize participation in MRTU Release 1 and MAP can be worked through and necessary corrections completed in a timely manner, so long as realistic expectations are maintained. The general schedule depicted in the timeline below gives the Commission SCE s best current estimates of how quickly it can move from the present state to actions (post commission decision on 00- DR applications) that will jump start SCE s commitment to maximize DR program participation in CAISO s MAP in early 0. 1

21 Figure I-1 1 The figure below summarizes the main activities associated with the transition of DR programs to the capability to bid into CAISO wholesale markets. The deployment of DR for wholesale markets is dependent upon the actual deployment schedule for MRTU Release 1 and MAP. The timeline below assumes the deployment schedule the CAISO is currently discussing. 1

22 Figure I- 1 1 d) Limits to Estimating Transition Costs SCE cannot at this juncture estimate the time and costs associated with transitioning existing DR programs into the CAISO s three product models. SCE expects to have better information once it has completed the pilot. But SCE is committed to doing what it reasonably can in a timely manner to support the CAISO s MRTU Release 1 and MAP deployments to maximize the contribution of DR. E. Issues and Barriers 1. Issues and Barriers to DR in MRTU Release 1 This Section discusses technical and operational issues affecting execution of the manual work-around for NPL. Correcting issues identified will contribute to a more efficient and effective use of the manual work-around. If unresolved, the manual work-around will still function satisfactorily. Also in this Section is a brief discussion of the changing role of DR Aggregators. 1

23 a) SC-to-SC Trades There is an important technical issue regarding Direct Access that should be worked out as soon as possible if conversion of NPL to PL is to be a part of MRTU Release 1 in 00. During execution of NPL participating in CAISO wholesale markets under MRTU Release 1 in 00, trade requirements for Scheduling Coordinator (SC) to Scheduling Coordinator transactions must be implemented, or a workable substitute solution to support direct access customers. APX has presented information to SCE on the complexities of functioning as an aggregator and as the settlement service for the Capacity Bidding Program (CBP). The central issue is whether, or how, IOUs can schedule direct access DR. Presently, constraints are driven by having to present a balanced schedule to the CAISO. While the balanced schedule requirement is eliminated in MRTU Release 1, this may not wholly eliminate problems associated with SC-to-SC DR related trades. An SC is required for an Energy Service Provider (ESP) because only certified SCs can schedule load in CAISO markets. An SC may submit a balanced schedule to the CAISO and find, upon being called, that the schedule is no longer balanced. To rebalance the schedule, an SC will execute a trade with another SC. Enforcement Issues. The complications associated with the SC-to-SC trade structure include limits on the ability of SCs to reflect ESPs delivery of what has been proffered. They are disconnected, sometimes leading to issues between the SC and its ESPs. Direct Access DR transactions execute through an SC, so there is two-way interaction between the SC and the CAISO and two-way interaction between the SC and the ESP, but there is no direct interaction between the ESP and the CAISO. IOUs cannot intercede in any of these Direct Access DR transactions, which, in turn, can affect LSE operations, leaving IOUs with only a reactive posture. In other words, IOUs have no preemptive or proactive ability to deal with variances stemming from the interactions of SCs and the CAISO, or SCs and ESPs. Portfolio Management Challenges. APX is the only scheduler where DR Aggregators are direct clients. But according to APX, scheduling with direct access is more costly, so 1

24 Aggregators are driven to focus on bundled customer bids. To achieve the kind of direct access that CAISO seeks requires managing co-mingled portfolios of both bundled and direct access customers. To execute in this environment, it is necessary to know if a specific DR source is a direct access case. If a DR source is a direct access customer, what information or transactions would require interaction with their ESP in order to enable the customer to participate in PL, NPL or PDR? At present, the processes identifying the specific DR source are insufficient because more clarity is needed regarding who gets what information about the DR source, and how and where DR information flows. APX points out that it is unclear who can help resolve these issues that require clarification in order to ensure a successful release of MAP. There are related issues to consider as well, including measurement, settlement, bid structure, and event management for direct access demand response. b) Day-Ahead/Real-Time Tradeoffs The CAISO seeks DR available for the Day-Ahead and Real-time markets, and as soon as possible, for Ancillary Services. However, there is an operational issue concerning the manual work around environment for 00 that should be addressed. As currently structured and envisaged going forward, DR must be bid into the Day-Ahead market, and thus is not be available to participate in the Real-time market. However, Day-of DR resources may be more useful, and more valuable, in the Real-time market. But if the Day-of DR is held back from the Day-Ahead market in anticipation of need in the Real-time market, the DR resources are ignored by the CAISO and their use RUC. Market Power Matters. A collateral consideration related to the tradeoffs grid operators and procurement management must make concerning what DR is bid and what DR is held back for local reliability assurance is how the FERC, CAISO and the Commission will evaluate these decisions with respect to how such decisions impact market prices. While the general assumption is that DR contributes to the mitigation of market power, what drives this assumption is the underlying assumption that DR will be bid. In some circumstances, LSEs may find it necessary to maintain realtime DR resources for local reliability mitigation. 1

25 WG Agenda Item. Protocols for clarifying how tradeoffs can and should be made should be taken up in the WG process, as this consideration affects all three CAISO DR products and their performance. Understanding these dynamics will contribute to solving problems that might otherwise reduce the full effectiveness of both MRTU Release 1 and MAP when deployed. c) Changing Role for DR Aggregators Presently, third-party DR Aggregators (sometimes referred to as third-party DR Resource Providers) conduct most of their business through retail resource and program contracts with IOUs. It is possible that DR Aggregators as future DR market players may begin to capture dispersed DR resources and bid them as aggregated products directly into CAISO markets; in effect disintermediating the IOUs. SCE is unable at the present time to fully identify the scope of the technical and operational issues that will need to be addressed as the role of the DR Aggregator changes. However, if the CAISO and the Commission want to examine DR Aggregators direct participation in CAISO markets, it will be important for SCE and the other IOUs to evaluate the effects on their IOU retail program operations of DR Aggregators running non-iou DR programs independently and in parallel. LSEs including SCE and the other IOUs will face new challenges in forecasting load, especially on a day-ahead and day-of basis. As controls on DR programs are reduced, load pattern uncertainty may rise, risks may change, and very specific issues concerning metering, meter data management, billing, realignments of liability assignments, and many other issues will need to be resolved.. Issues and Barriers to DR in MRTU Release 1A (MAP) There are numerous operational and technical challenges facing activation of the CAISO s DR products for MAP. Only when the issues discussed are jointly addressed by the IOUs and the CAISO will the confidence in the efficient and effective transitioning of DR programs into CAISO wholesale markets be elevated. 1

26 a) CAGS and Associated Costs The level of detail in functional requirements definitions of Custom Aggregation Groups (CAGs) and how customers are aligned with circuits and substations needs to be much more specific if SCE is to be ready to deliver DR when MAP is activated. The costs of mapping the entire LSE system to the level of detail required should be detailed and weighed against the benefits of PDR and PL in CAISO markets. These are open questions that each IOU should examine internally, and it may be subject matter for WG as its work unfolds. b) Operations Changes As more granular, locational DR is expected by CAISO, there may be managerial changes required in the way SCE manages its LSE operations. Currently, all cost accounting and system management execute through a geographically based district system. As MAP deploys, system management will be a more network-focused effort, raising questions about how district operations will align with new requirements for network operations under MAP. c) Unprecedented Detailed Mapping Requirements As more granular, locational, DR is expected by CAISO, there may be structural changes required in the way SCE manages its LSE operations. Currently, all cost accounting and system management execute through a geographically based district system. As MRTU deploys, system management will be a more network-focused effort, raising questions about how district operations will align with new requirements for network operations under MAP. The shift to a network focus will take time, and money, which are presently not incorporated in any of SCE s budgets or regulatory filings. Specific requirements associated with these changes include mapping meters to poles and individual customer accounts to specific circuits. If the CAISO wants granular locational DR, this level of specificity will be necessary to ensure that the DR that is proposed is actually the DR that is delivered, accounted for, and appropriately compensated. 0

27 . Technical and Operational Issues Pertaining to Both PDR and PL There are several issues that must be addressed prior to the adoption of PL and PDR for DR a) DR LAP and CAG DR programs are structured at the Load Aggregation Point (LAP) level. Under PDR the expectation is that DR would be aggregated locally and defined as a CAG. None of SCE s DR related products are currently in CAGs. At the present time, configuring them for CAGs is an unknown cost that will take an unknown amount of time given how recently CAISO introduced the PDR model for DR. Conceptually, CAGs are the equivalent of a virtual LAP. The absence of assured alignment with the physical structure of the local grid may have implications for what it will require of SCE with respect to increasing the granularity of LAP and sub-lap level information to enable appropriate configuring of CAGs. The configuration of CAGs and how CAGs may be layered in bid structures to maximize the value of PDR requires consideration of the workflows and protocols that may be necessary within SCE as well as how SCE s systems interface with MAP. b) Accounting Issues For both bid structures involving CAGs and DR Aggregator portfolios and possibly single customer direct access, there may be accounting considerations to sort through at the meter interface. For instance, a CAG may not align with meters, which may result in a portion of a meter being allocated for DR while the remainder of the meter is not. How IOUs measure, track, and validate the split for revenue accounting purposes may have impacts on both IOU information systems, including meter data management systems, and CAISO s MAP systems. Addressing these issues up front before MAP is released may save costs and help CAISO s interest in maximizing DR participation in its markets as soon as possible. 1

28 c) DR Certification For both PDR and PL, DR provisioning has to be real. That is, DR that is offered must be validated. This will require developing a robust certification process that is shared by direct access participants, DR Aggregators, IOUs and the CAISO. The structure, processes, metrics and assurance mechanisms for a robust certification system should be a central consideration of WG. d) Direct Access A critical technical issue concerns integrating direct access PL or PDR at the utility LSE. For customers to uncouple DR from LSE (IOU & ESP) grid control may produce an unintended consequence of increasing the instability in LSE grid control. If direct access is uncoupled from LSE operations, it introduces a new source of uncertainty, and associated risk, when forecasting expected loads and resources on a daily basis. The uncertainty stems from DR Aggregators offering resources directly into CAISO markets but having no requirements to provide information to LSEs trying to regulate grid operations. If the load reductions do not show up as forecasted by direct access providers, it increases the risks to LSEs in assuring reliable service and may increase sourcing costs as the CAISO allocates its sourcing costs for filling in for what was forecast but did not show as scheduled. (1) Self-Scheduling of Direct Access Load As presently structured, LSEs cannot self-schedule direct access load. This issue needs to be reviewed to determine its significance, as well as a proposed solution to any associated problems, assuming that LSE direct access load self-scheduling is important to the efficient and effective use of the PDR product. The self-scheduling of direct access load may be a challenge because of the potential scope of information disaggregation and account restructuring that will be required by IOUs. It will be important prior to the activation of MAP to determine how significant the technical challenge may be in terms of costs, timing, and benefits derived from account restructuring. () Limits of DR Aggregator Bids DR Aggregators present their customers as a single portfolio to IOUs. To shift to a direct access link to the CAISO for DR Aggregators there are technical issues to sort

29 concerning how an Aggregator portfolio might be structured for an IOU as a customer and how it might be structured for direct access. () DA Customer Accounting There are accounting challenges for rate purposes that are similar to those noted above concerning DR Aggregators presenting their customers as a single portfolio. While these are likely not insurmountable, these issues do carry with them possible tariff change requirements that would necessitate a request for changes filed with the Commission. From a revenue accounting perspective, direct access customers will require different accounting processes from existing utility account processes. Direct access accounting and account management processes are likely to differ from utility accounting and account management processes. Differentiation of direct access DR will require changes in IT, data management, and possibly customer relationship management/customer care programs. The scope of IT and process changes associated with accommodating this differentiation remain to be determined. Information systems adjustments may be necessary, and these carry not only cost implications but may require changes in specific interfaces with DR Aggregators and in related internal systems of SCE. () DA LAP and CAG There are operational challenges to MAP related to mapping of the details of how direct access is included in a PL bid. To configure DR in a locational context as the CAISO prefers, DR resources must be organized as more granular CAGs because the LAP level may not ensure needed precision with regard to locationally targeted DR that CAISO may need. e) IT Systems Impact To determine the extent of effort and requirements, SCE needs a better understanding of the overall IT challenges associated with PL and PDR. For instance, there may be requirements associated with modifying SCE s Customer Information Systems and its overall customer care programs. There may be adjustments to meter data management systems required. Settlement systems may require modification, and LSE grid control protocols also may require adjusting.

30 The IT infrastructure requirements to support programs at the more granular CAG level, including direct access, are significant and will necessitate significant changes in systems and associated systems integration. Understanding the scope in advance of MRTU Release 1 and MAP deployments will be important for costing required changes, and for achieving realistic change management schedules. Estimates of both the cost and schedule required to affect the IT changes must sequentially follow resolution of issues noted above. Presently, SCE cannot state with confidence what data requirements and associated costs will be necessary in order to achieve the level of granularity necessary to enable the configuration of CAGs for PDR because the PDR model was introduced to IOUs for the first time July 0, 00. This is a surmountable problem, but it will take time and thorough technical/engineering analysis of SCE s system in relation to CAISO s network models and associated locational DR requirements to sort out. Beyond the gathering and analyzing of the kind of granular information the CAISO expects to be available and to drive PDR bidding, there may be information technology changes necessary to process and make use of more granular data on an ongoing basis. The specific location of IT adjustments may include data storage, data analysis tools, information flow, access, and security assurances, and certain compliance requirements driven by FERC, NERC, and the Department of Homeland Security, which will take time to implement. Again, SCE reiterates that these appear to be wholly surmountable problems, assuming the costs and time required to address them are authorized. f) Notification There may be adjustments in notices required to ensure effectiveness of PL. For instance, current day-of programs require 1 minutes notice, and the RTEM market provides only minutes notice. Day-of programs would no longer serve the current Hour-Ahead market or the future equivalent of the Hour Ahead Scheduling Protocol (HASP). Day-of programs would be required to manifest themselves into PL and serve the real-time market. Thus, the notice provision would be shortened from 1 minutes to minutes for a day-of energy based DR resource.

31 g) Incentives Alignment In preparing for PDR and PL use in 0 and beyond, the structuring of how PDR and PL DR bids align with incentives should be examined. In some cases, for instance, provisioning of DR to mitigate certain local congestion constraints may yield sufficient to greater than required DR participation. The positive consequence may be a reduction in the cost of DR, and possibly a reduction in overall CAISO sourcing costs. These are only possibilities, but the point of the example is to show that more analysis of the consequences of PDR and PL is needed to ensure that MAP is structured to both minimize problems and maximize benefits of PDR and PL.. Technical and Operational Issues Affecting PDR The CAISO expects to have PDR ready for use shortly after the activation of MRTU Release 1. The general estimate is that three to six months following MRTU Release 1 activation, PDR will be ready. Given the technical and operational issues associated with PDR for SCE, and possibly the other IOUs, at the present time, with somewhat limited information about PDR, SCE is uncertain whether it will be ready to use PDR as a product if it is ready as the CAISO expects. It is SCE s understanding that the CAISO intends to continue the availability of PDR after MAP is deployed. If the CAISO intends to maintain PDR into MAP and is indifferent to whether DR is bid as PDR or as PL, it may be worth analyzing how these products will function to maximize DR for the CAISO. SCE is impressed with the PDR model and can see its value on a continuing basis. It may be a lower cost and more efficient DR product than PL. Before investments in enabling both PDR and PL are made by SCE, the IOUs and the CAISO should consider the optimal mix of DR products offered by the CAISO. Technical and operational issues affecting PDR apply both to MRTU Release 1 and to MAP. There are technical and operational issues that are specific to applying the CAISO s PDR model that should be addressed prior to the activation of MAP and SCE recommends that key barrier challenges should be resolved jointly with the CAISO.

32 a) Adapting Capacity Bidding Program (CBP) There will be technical difficulties adapting CBP, which is currently focused at LAPs, to PDR, which requires more granular CAG level configurations. b) Single Product Portfolio DR Aggregator resources are presented as a single product portfolio, but PDR is intended to enable direct access bidding. For DR Aggregators to bid directly, there will be protocol and process requirements specifications in MAP that may not now be sufficiently detailed. c) DR Values Mismatches The pricing of DR at the granular CAG level may lead to mismatches between where DR is needed and where DR is offered. Analysis of likelihood, variances, and associated risks, and related costs is important to undertake.. Issues and Barriers Associated with Participating Load In this Section, with respect to technical issues, issues identified for PDR for Day- Ahead markets are the same for PL. a) PL and RUC Value It remains unclear how RUC value is treated under MAP. Receiving RUC credits with DR bid into the MAP model must be sorted out in advance of MAP deployment. How RUC is valued will have potentially significant effects on how DR providers bid into CAISO markets. b) Forecasting During recent technical design sessions between the IOUs, the CAISO, and other parties, issues related to forecasting and forecasting accuracy were raised. While forecasting concerns may apply to all DR venues the CAISO is advancing, it is especially important when dealing with PL. c) Nodal Load Forecasting While forecasting may be manageable at the aggregate level on a daily basis year around, it is unclear what LSEs will face as granularity reduces the focus to locational points within the grid. Absent experience pertinent to an unprecedented change, it will be important to work through implications as best they can be articulated.

33 d) Metering and Telemetry Certain PL sources of DR may require metering and telemetry. For load participating in Ancillary Services markets, this is necessarily so. It is unclear if other PL sources of DR would have similar requirements. e) Geographic Customer Differentiation There may be limitations to how customers can be differentiated for DR purposes within a specific geographic territory. Based on discussions in the previously referenced technical design sessions, the significance of this barrier is unclear. SCE s interests are in assuring fair treatment to all its customers, so obtaining a clear answer to the question is important. f) DR Value Misalignments As discussed in other sections of this Volume, triggering DR activation on a CAG price, which is not cleared in CAISO markets and is settled on LAP prices, may result in misalignments of DR locational availability relative to locational need. It will be important to detail the business rules associated with PDR to understand whether this is a serious barrier to effective use of PDR. g) Cost of Required Information Granularity Reaching the level of granularity of information necessary to configure CAGs in the PDR model will take a serious effort. SCE estimates that the time required to achieve this is at least a year. Therefore, consideration should be given to when PDR can become a useful DR product and what the underlying costs of PDR will be given the costs of achieving the necessary level of granularity of information within SCE s service territory. h) Accounting for RUC As noted previously, clarification of accounting for RUC has been a topic since the beginning of MRTU and consideration of DR in MRTU. It will serve all participants in CAISO markets if ambiguities regarding how DR in RUC is accounted for are cleared up as deployment of MAP occurs.

34 i) Detailing Functional Requirements The level of detail in functional requirements definitions of CAGs, how customers are aligned with circuits and substations needs to be much more specific if SCE is to be ready to deliver DR when MAP activates. The costs of mapping the entire LSE system to the level of detail required should be detailed and weighed against the benefits of PDR and PL in CAISO markets. j) Edison SmartConnect For those PL sources of DR requiring metering and telemetry, it is in the interests of customers that no unnecessary investments in metering and telemetry be made if the implementation of advanced metering provides equivalent or comparable information sources and availability. As considerations are given to PL contributing to Ancillary Services, issues of cost effectiveness and timing should be explored. All California IOUs are in the process of seeking authorization to deploy or deploying advanced meter infrastructures. Each IOU has adopted a somewhat different meter technology which has implications for related information system changes within its operations, from meter data management systems to overall process flows associated with billing cycles, maintenance workflows, and many other functions, processes, and work flows. Variances in AMI solutions may impact each IOU s ability to package DR under the PDR and PL models.. A Shared Effort to Overcome Deployment Challenges for MRTU DR Products a) Working Group SCE believes that the technical and operational issues, barriers, questions, requirements and unknowns described in this Section pertaining to the CAISO s DR products NPL, PL, and PDR - will require a joint IOU/CAISO effort to resolve the technical and operational issues. The CAISO s existing working group structure for DR market design is a venue that can be used to facilitate resolution of these items since all stakeholders affected by DR for MRTU Release 1 and MAP are aware of the working group and the working group process is more or less in place. SCE recommends WG as the venue for addressing these challenges for MRTU Release 1, Map, NPL, PL and PDR. None of the issues identified are expected to be insurmountable. A working group process,

35 which the CAISO has already initiated, can be reactivated and used as the focal point for joint IOU/CAISO considerations of not only issues, but certain barrier challenges. SCE suggests and encourages a process to implement DR market changes in both PDR and PL, as necessary, which involves at least the following steps and associated deliverables. (1) User Guide A consolidated, comprehensive user guide for CAISO s DR products is needed. This can be developed in WG. The issues raised in preceding sections of this document should be taken up in preparing the User Guide. () DR Program Consolidation A joint review of IOU DR program designs to identify opportunities for consolidation and elimination based on overlaps and changing market circumstances should be done in WG. () Map Process Changes IOUs must outline the internal process changes required to implement the consolidated user guide with the CAISO so there is a shared understanding of scope, timing, and costs. Based on the outcome of this effort, subsequent funding requests for process and systems changes may be appropriate. () Map IT Requirements IT requirements should be specified for PDR and PL products that cover DR daily bidding, planning, settlements, metering, and billing. Internal processes related to DR daily bidding, planning, settlements, metering, and billing also should be mapped. () WG Meeting Schedule & Process Output SCE believes WG will require several meetings of the principals (IOUs, the CAISO, DR Aggregators, and customers, as well as other stakeholders) to work through the agenda items described. Given the compression of time as the CAISO seeks to achieve maximum DR participation in its wholesale markets as soon as possible, the WG working sessions should be scheduled with the goal of being finished with WG work by the middle of November, or sooner.

36 SCE considers a structured engagement processes focused on producing essential documents to be the appropriate mode of achieving deliverables. The CAISO should be responsible for drafting documents based on discussions at WG meetings and WG participants should be responsible for written responses that are discussed in subsequent meetings. () Final WG Output Package The deliverables of WG should be consolidated and packaged for access by all participants, the user guide, in particular, posted for easy access by all affected parties. (a) Tariff Changes Following completion of WG work, if there are tariff changes needed, the IOUs will make those changes through the appropriate channels. 1 (b) Contract Changes Once the level of specificity with respect to PDR and PL products is achieved through WG, IOUs may need to make changes with DR contracts, DR Aggregators, and may require procurement process modifications. () IOU Deployment Planning and Programming Once these steps have been taken, attention can focus completely on expeditiously moving forward with process and systems changes that will enable full participation in MAP consistent with the currently estimated deployment timing for MAP. Time for testing of processes and systems, simulation of how PDR and PL products work involving all affected parties, and final certification and tuning will be required. This general work flow is a significant task for each IOU and must be managed in the context of continuing to operate reliable, high quality service for customers. After resolution of the issues described above, SCE will have enough information to more accurately estimate the costs and timeline associated with tariff development implementation, systems development, nodal mapping, telemetry metering and DR forecasting. 0

37 Institutional Timing Considerations Affecting IOU and CAISO Efforts There is a central, overriding issue and barrier that affects how quickly IOUs can transition programs. IOUs must file certain changes with the Commission, and the processing of requests takes time. For modest changes to tariffs, for instance, the Commission process may take between three and six months. For extensive tariff changes, the Commission process may take upwards of a year. a) Grid Operations Change Durations When program changes impact grid control operations at the LSE level, it typically takes six to nine months from start to finish to implement changes. b) Information System Change Durations Where SCE information systems changes are required, as would be the case with most of its DR related program changes, it takes a minimum of six months, and for substantial changes to information systems, it can take a year or more to implement them. c) Asset Change Durations If there are physical asset changes required, such as modifications in circuits or substations, changes in meters or bypass solutions that do not require interfering with meters, it is typically a one-year or longer process. d) Billing and Settlements Change Durations If there are changes to billing and settlements, the change management requirements typically result in a six-month to more than a year time period for affecting changes. e) Difficulties in Compressing Institutional Time Requirements The systems that must be changed to enable the kind of DR participation CAISO seeks are substantial. Regardless of how expedited transitions are sought, the above timing considerations are difficult to shrink. f) FERC Finally, CAISO assures the IOUs that no FERC tariff changes are required to affect the DR changes that it seeks. A PLA, which has been filed with FERC and approved by FERC, 1

38 can be modified without seeking approval from FERC for amendments. If this conclusion turns out to be in error and FERC filings are required, it should be noted that this could add three months or more to the implementation of DR related programs. These timing constraints are summarized on the following graph. Figure I- 1 F. Proposed SCE Actions to prepare for MRTU 1. Working Group The change from existing programs to NPL and PL is significant, but the degree of significance is still to be specified. SCE will look to the CAISO to produce technical specifications for PDR and PL as the CAISO expects them to work with the MRTU Release 1 and MAP. SCE encourages WG to be the forum through which technical specifications are developed and completed. WG should undertake whatever analysis is necessary to identify potential suboptimal impacts associated with MAP and correct them in advance of the release of MAP in 0.

39 Transitioning existing Spinning Reserve Pilot for PL a) Background The DR Spinning Reserve Demonstration Project (also known as the DR Spinning Reserves Pilot) is a pioneering demonstration of how radio-dispatched direct load control can act as a demand-side replacement for electric system reliability supply-side resources known as spinning reserves. With the proven availability of demand-side resources providing peak demand reduction, this project demonstrates a potentially powerful new tool for grid operators at the CAISO and SCE to help prevent rolling blackouts and improve system reliability by using these resources. In 00, in addition to the widespread targeted marketing of DR, SCE implemented the Pilot to study the potential of using direct load control devices (such as those used for SDP) as a form of demand-side proxy for spinning reserves. The objective of the Pilot is to demonstrate the visibility of real-time observable load curtailments in a manner comparable to that relied on by operators when they call on generators to provide spinning reserve. In addition, the Pilot is intended to measure and characterize the speed and magnitude of the activation signal and load response. This is done so that load curtailments are compared directly to the historic performance of generators and are assessed with the impact on customers during and after curtailments in order to evaluate customer acceptance issues. It was determined during 00 program activities that leveraging the existing SDP air conditioner control technology as a spinning reserve resource can be valuable as a potential supplemental replacement for relying on supply side generators for this service. As stated in the 00 DR Spinning Reserve Demonstration Report, [t]he natural response capabilities of these loads match the response speed, duration, and frequency required to support spinning reserve. Collaborative funding for the program was supplied by the California Energy Commission Public Interest Energy Research (PIER) program through the Consortium for Electric Reliability Technology Solutions (CERTS). Other parties collaborated in the program under contract to SCE and CERTS. See SCE Advice Letters 1-E, 0-E, and 1-E. Demand Response Spinning Reserve Demonstration Report, 00, p.

40 In 00, SCE recorded approximately $,000 in expenses on the Pilot, primarily on program administration, enrollment, and testing of enhanced load control switches with telemetry functionality. SCE used direct mail and telemarketing to encourage enrollment in the Pilot and SDP to 1,00 eligible customers located on one constrained distribution circuit. Nearly 00 devices participated in the Pilot, which represented approximately a 1 percent saturation of the distribution circuit in one DR program. SCE was authorized additional funding in 00 for the continuation of the Pilot and for the increase in the number of distribution circuits in the Pilot to four. SCE recorded approximately $,000 in expenses primarily for marketing and customer incentives in 00. By August 00, SCE had installed more than 00 devices for customers participating in the Pilot. Table I- Spinning Reserves Recorded Expenses and Forecast 00 Line Expense Recorded ($) Forecast ($) No. Type SCE Labor Non-Labor,1 1,1 0,000. Incentives,00 0,000,000. Capital Total,0,1 0, b) Spinning Reserves Pilot Accomplishments Since 00, SCE and the Consortium for Electric Reliability Technology Solutions (CERTS) have worked closely together to demonstrate the use of DR from air-conditioning direct load control (DLC) switches as a resource for supplying spinning reserves. This work has led to the creation of a significant demonstration project, The Demand Response Spinning Reserves Pilot, that includes: Enhanced switches are for customers receiving service under Schedule APS-E. Percentage is for all active customer service accounts on the circuit.

41 Initiating targeted deployment of DLC switches (approximately 1,00) across four () geographically distinct distribution feeders within SCE s service territory; Providing real-time telemetry from the DLC switches (approximately 0 are configured with this capability); and Developing a web-based portal that displays near real-time information on the total MW load of the four feeders and operational status from the switches. c) Transitioning to support Participating Load In response to the CAISO s urging that some PL be ready when MRTU Release 1 is deployed, SCE proposes to modify its current Demand Response Spinning Reserves pilot to evaluate its capability as PL. SCE proposes leveraging the existing Demand Response Spinning Reserves pilot infrastructure, including the customer base, tariffs, systems, processes, and CERTS partnership, to evaluate its capability as PL. (1) Objectives of the Participating Load Pilot The objectives of the Spinning Reserves Pilot for PL (SPR-PL) include facilitating the articulation of CAISO requirements and technical specifications for DR to act as PL, and to uncover and resolve policy, process and system issues associated with using DR as PL. SCE expects to deliver the following, associated with execution of the SPR-PL pilot: First, launching the pilot for use by Summer 00, wherein the CAISO can request this resource as PL on a day-ahead basis beginning June 1, 00. In the event systems or processes are not ready by this date, or the Commission does not approve the pilot in time for a Summer 00 launch, the CAISO and SCE will work together to determine appropriate manual work-arounds.

42 Second, an actionable set of requirements and technical specifications agreed to by stakeholders as applicable to any SCE DR resource that is bid into CAISO wholesale markets as PL. Third, pending resolution of the issues to be addressed by WG, the SPR-PL pilot will help develop DR programs utilizing the markets facilitated by MAP. () Scope of Work To successfully extend the existing Demand Response Spinning Reserves pilot for consideration as a PL, the scope of work required to affect this transition includes at least these principal steps. Developing processes, procedures and systems internal to SCE to ready this resource for bidding into CAISO wholesale markets as PL. Further systems integration between SCE operations and CAISO operations, leveraging the extensive work done on the existing Spinning Reserve pilot project in this area. Developing the processes, procedures and systems for settlement both from the CAISO to SCE and from SCE to its pilot project customers. Gathering information on customer preferences in order to design future programs that will maximize participation. It will be particularly challenging to determine how (and if) SCE s SPR- PL pilot will implement the CAISO s telemetry requirement for PL. At the present time, these customers do not have interval metering. SCE will determine if the Edison SmartConnect deployment aligns with the chosen circuit areas for the SPR-PL pilot, and will examine how the interval metering capability of Edison SmartConnect can be utilized to facilitate SPR-PL pilot telemetry or settlement. There will likely be a need to consider a telemetry proxy and even a metering proxy for settlement in

43 lieu of actual metering at each customer site. SCE will work with CAISO in the SPR-PL pilot development to come to a mutually acceptable metering solution which will likely include a combination of circuit metering and sample customer metering. () Proposed Schedule Given the, objectives, deliverables and associated scope of work, the achievement of a Summer 00 pilot ready to function, as PL in CAISO wholesale markets is a significant undertaking. The chart below provides a first draft, high level overview of the general work flows associated with transforming the existing pilot into the SPR-PL pilot, as discussed above. Figure I- Proposed Schedule d) Program Proposal Please see Appendix J, the SPR-PL pilot description. e) Pilot Project Learning What is learned through the SPR-PL pilot should inform both SCE and the CAISO of practical challenges in making DR for PL to work effectively. The pilot effort will serve to increase SCE readiness for MRTU MAP by helping SCE to understand the technical and operational issues associated with actual, ongoing use of DR for Ancillary Services.

44 SCE has identified many issues of concern to the cost-effective and efficient deployment of PL in MAP. It is expected that the importance of these issues may shift based on the results of the PL pilot project. Since the pilot project cycle is still in its early stages of development, the issues concerning PL as a DR product for CAISO wholesale markets must be taken up now. Doing so will help SCE in its program planning for transitioning DR programs to participate in CAISO wholesale markets, and should strengthen the PL product for all who may seek to use it. f) Budget to Implement and Administer Program SCE forecasts $,01,01for The costs will be primarily for metering, system development and/or enhancements (for telemetry and settlement), pilot analysis, and customer recruitment and incentives. Table I- Participating Load Pilot Forecast Expenses 00-0 Line Expense Forecast ($) No. Type SCE Labor 0,0 0,0 0,0. Non-Labor,0 00,00,00. Marketing & Outreach,000,000 0,000. Incentives 0,000 0,000 0,000. Capital Total 1,, 1,0,, DR Resource Portal a) Background In the DR application, SCE requested funding for Demand Response Systems Integration for consolidation of the independent systems supporting the thirteen load management and DR programs active at that time. However, due to SCE s ERP initiative and potential development of the Edison SmartConnect back office systems, much of the proposed integration work could have isolated these load management systems from the more sophisticated systems being implemented at SCE through 01. SCE will instead work to integrate load management systems into the ERP and Edison SmartConnect systems to the extent possible.

45 One area of potential integration that does exist is with the CAISO, which has expressed the desire to have visibility of the IOUs available DR resources on a near-real time basis. This functionality is necessary to support DR Participating Load and Ancillary Services and potentially Non-Participating Load in the MRTU market. The DR Resource Portal program will develop an interface which will provide near-real time visibility into SCE s available DR resources (both reliability and price responsive resources). The DR Resource Portal is a new program in development for 00 and did not exist in 00 or 00. The DR Resource Portal architecture will be designed with flexibility to accommodate future interface/integration with the developing MRTU market rules for DR Participating Load and Ancillary Services and potentially Non-Participating Load enabling DR events and resources dispatched by either SCE, DR Aggregators or CAISO. In 00, SCE s overall objectives are the establishment of the program definition, scope, schedule budget, risk analysis, change control, quality control and labor resources. SCE is working with DOJ Consulting to define and scope this project. The 00 expenses for this effort are anticipated to be approximately $,000 and will be funded from the DR System Integration program authorized in D b) Program Proposal Providing more information about DR availability during the 00 period, as preparation for MAP in 0, may contribute to keeping the above trade-offs clear. DR Portal may also help avoid concerns related to market power, noted above, that could be raised should less DR than expected show up in CAISO markets. For 00 0, SCE is proposing to execute the plan developed in 00 which will result in a DR Resource Portal that provides the CAISO near-real time visibility into SCE s available DR resources (both reliability and price responsive resources). SCE s primary objective will be to establish a real-time data sharing infrastructure for DR events and performance to be used by both SCE and CAISO.

46 c) Budget to Implement and Administer Program Funding is being requested to develop business requirements, conduct pilots and/or trials of one or more software applications, and develop, integrate, and test the system. SCE is forecasting 00-0 total expenses for the DR Resource Portal program of $,,000. Table I- DR Resource Portal Forecast Expenses 00-0 Line Expense Forecast ($) No. Type SCE Labor Non-Labor 1,,000,000 0,000. Incentives Capital Total 1,,000,000 0, MRTU and SmartConnect Deployment During the 00-0 period covered by SCE s Amended Application, SCE s Advanced Metering Infrastructure (AMI) program, Edison SmartConnect, will be in the process of deployment. Assuming that the deployment remains on schedule, SCE s meter infrastructure for customers with <00kW of load will be fully transitioned to SmartConnect meters by the end of 01. SCE filed a detailed program and budget request with the Commission in a proceeding that is currently underway (A.0-0-0). As SmartConnect deployment is implemented, several new programs and rates are expected to take affect as well, including dynamic pricing, and demand response options for residential and small to medium commercial and industrial customers. Edison SmartConnect will also enhance SCE s customers ability to control their energy use and costs. Transitioning DR programs to PL resources may be enhanced by the completion of Edison SmartConnect deployment. Edison SmartConnect will provide interval metering with 1 minute usage information for small and medium commercial and industrial customers and 1 hour usage information for residential customers. While the current requirements for measuring PL has more stringent specifications for data frequency and telemetry, it remains to be seen to what extent Edison SmartConnect can be leveraged to facilitate PL, NPL or PDR in MRTU. 0

47 SCE and the Commission should give consideration to striking the appropriate balance between incurring costs to respond to CAISO s sense of urgency with immediate changes to DR programs and waiting to realize the potential benefits of Edison SmartConnect, which is expected to be authorized in September 00. G. Other Pilots Considered 1. Demand Response and Resource Management The CAISO, in numerous venues, has advocated for the use of DR matched to, or coupled with, an Intermittent Resource (IR) to help mitigate IR variability. Also, the CAISO, and the Commission through its guidance documents, has expressed interest in Small Load Aggregation (SLA) as a potentially significant source of DR for the CAISO s wholesale markets. In this Section, SCE addresses the Commission s guidance to IOUs on considering IR and SLA. Conceptually, linking DR to variable IRs for purposes of reducing variability, or smoothing as it will be referred to herein, is worth considering. However, SCE could find no research or case studies of sustained practical experience from other markets that provide much information or insight regarding identifying the best types of DR (and related best practices) to be matched with specific types of IR. 1 SCE s view is that reducing IR variability and increasing control of IR for assuring grid reliability is a resource management challenge that may be more significant in complexity, scope, and requirements than adapting existing retail DR programs to serve the CAISO s wholesale markets. The reasons for this perspective are developed in this Section. 1 There have been several studies of wind variability and impacts on grid operations. A study in 00 by the Alberta Electric System Operators profiled characteristics and associated issues consistent with the CAISO s concerns. A major study by the New York State Energy Research and Development Authority, THE EFFECTS OF INTEGRATING WIND POWER ON TRANSMISSION SYSTEM PLANNING, RELIABILITY, AND OPERATIONS, March 0, 00 prepared by GE Energy Consulting chronicled critical issues and impacts of wind on grid operations, and the National Renewable Energy Lab published a study in July 00, GRID IMPACTS OF WIND POWER VARIABILITY: RECENT ASSESSMENTS FROM A VARIETY OF UTILITIES IN THE UNITED STATES. In the NREL study, wind effects on several utilities were case studied. Cases included Xcel Energy, WE Energy, PacifiCorp, and CA RPS Phase III. All of these studies are consistent with the CAISO s concerns, but none of them consider how the matching of DR with IR might smooth IR variability, nor were any of these studies time series studies of any significant length. 1

48 Given how new the concept of matching IR with DR is, at best this particular category of DR services might be explored through one or more forms of pilot project. Possibilities will be discussed in this Section after discussing how IR smoothing fits within the general framework of DR and after issues and barriers concerning the concept are generally outlined. In recognition of the novelty of the concept and the limits of precedents, SCE declines to propose a pilot for IR/DR at this time, but does present its evaluation of the concept. Conceptually, Small Load Aggregation is more straightforward and consistent with other forms of DR that may align with the CAISO s three DR products NPL, PDR, and PL. But there are distinctive challenges associated with efforts to harvest SLA potential. SCE also declines to propose a Small Load Aggregation pilot at this time but offers its evaluation of the concept. This concept will be discussed following consideration of Intermittent Resources.. General Definitions Pertaining to Intermittent Resources and Demand Response Defining Demand Response. In general terms, DR refers to mechanisms to manage the demand from customers in response to supply conditions. Specific to California, DR includes programs that reduce consumption at critical times (staged alerts as precursors to emergency actions required to ensure grid reliability and stability) or in response to price signals (provided by market clearing prices or other means of generating price signals, or through pre-arranged, and pre-priced call options exercisable under specific conditions by IOUs, DR Aggregators, or the CAISO through such instruments as Interruptible Contracts). IR Related Assumptions. For consideration of IR being coupled to DR in California in service to the CAISO s objectives regarding ensuring grid reliability, certain assumptions must be made with respect to the relationships between DR, energy efficiency (EE), permanent load shifting (PLS), and distributed generation (DG). Being clear about these distinctions is necessary because if the objective is to smooth IR variability (so that it can contribute more reliably to meeting power needs), this can be done in many ways that fall outside the present standard operational definition of DR used by the CAISO and the Commission.

49 DR and Energy Efficiency. Where DR is focused on mechanisms that manage the demand from customers in response to supply conditions, as noted, energy efficiency is focused on productivity per unit of energy, i.e., performing the same work but using less power to accomplish the same work. DR involves dedicated control systems for load shedding in response to various forms of signals from the CAISO and/or LSEs, e.g., reducing or cycling air conditioning, interrupting production processes, adjusting lighting levels. In other words, energy efficiency involves changing assets, processes, and control solutions on existing work flows to continue the same work output for lower energy inputs. Permanent Load Shifting. Permanent load shifting helps mitigate grid reliability stresses by permanently altering energy consumption patterns to reduce peak demand volumes. SCE operates a PLS program focused on businesses, providing enabling technologies that assist consumers in achieving more efficient use of energy while altering the pattern of consumption. The SCE program includes such solutions as: Chilled Water Storage ("CWS") The CWS system includes the design, installation, and operation of storage systems for chilled water so that the building s air conditioning compressors can manufacture and store chilled water during the off-peak hours (generally at night) for use later during the day. During on-peak periods equipment such as smaller pumps and fans operate to circulate the chilled water to condition the spaces with the main compressors turned off, thereby permanently reducing a major portion of air conditioning energy usage and demand from the onpeak period. Ice-on-Coil ("Ice Bank") Storage This is a modular TES system that utilizes a chilled water-glycol technology to manufacture ice in an enclosed tank for central air conditioner systems. Using the same TES principle as in chilled water storage, the ice-on-coil system will "charge" the modular tank during the off-peak hours with the electric chillers that would normally operate during the day. The Ice Bank system will

50 manufacture and store ice at night, which will be used during the day to chill the water for air conditioning. Ice Storage Air Conditioning ("ISAC") ISAC uses an external ice storage tank to provide the thermal storage for a packaged air conditioner that uses refrigerant, not water, as the cooling medium. The system operates the compressors at night to freeze water in a self-contained storage system to create ice. The ice is then used during the day to cool the refrigerant, which in turn is used to air condition the building. The system may be retrofitted in existing packaged central air conditioner systems (usually to 1 tons) or installed as new systems. 1 These PLS solutions are noteworthy because when CAISO staff uses examples of how IR may be smoothed by matching it with DR, ice storage is used as a case in point. Distributed Generation. Distributed Generation (DG) may be an attractive candidate for IR smoothing, so it is useful to ensure a clear understanding of the differences between DR, EE, PLS, and DG. DG includes on-site generation of electricity as a primary source of supply or as a supplement to sourcing of electricity from the power grid. Also, DG sometimes includes small generation that is located near (but outside the fence of one or more) end-users, and generation solutions (which include small generation plants, batteries and renewables) that are located within a distribution network at, for instance, substations, dedicated to supporting grid reliability. The use of backup generation, assuming a zero impact profile could be achieved, might be an attractive resource for matching with IR. But the alignment of IR with this counter-resource, which might offset for IR variability, would not easily fit into the existing operational definition of DR used by the CAISO and the Commission. DG and Net Metering. Increasingly, DG is considered a supplement to grid-based power supplies where the resource can be used to reduce demand for grid-based power or it can be used to contribute to grid reliability assurance requirements under various grid conditions. 1

51 Federally mandated mechanisms such as Net Metering enhance the two-way flexibility of DG on the one hand, and complicate its deployment on the other hand; the latter a function of an array of problems that must be solved to enable DG to so function. DG/Net Metering related issues and barriers include, but are not limited to, meter interface considerations, billing restructuring challenges, accounting and verification of net meter flows (inbound and outbound), and operational grid control assurance requirements. As the Commissions deliberations on definitions evolve, it is SCE s understanding that storage technologies might end up being included in the general category of DG, or they may be treated as a separate resource category. If integrated into a Net Meter driven solution, the problems noted above with DG may apply to storage as well. Smoothing IR Variability Using More than Just DR. IR variability may be smoothed through DR, EE, PLS, and/or DG, or various permutations of these categories of overall demand resource management. In considering how IR may match-up with DR, this broader framework is useful to keep in mind because pilot testing might include experiments that do not easily fit into the specifically defined DR category that is under consideration with the CAISO or as defined by the Commission.. Intermittent Resources (IR) State Policy re: Renewable Resources and the Electric Power Supply Resource Stack. California state policy, supported by the Commission, is to achieve at least twenty percent of California s power supply from renewable energy supplies. These renewable resources come principally from wind and solar as fuels, but the category might include tidal-derived power, possibly run-of-the river hydro, and bio-based clean fuels driving more traditional power plants producing electric power. For example, biodiesel and/or ethanol are sometimes considered fuel inputs for certain types of cogeneration; and certain fuel cell technologies can run on ethanol. However, for the foreseeable future, the principal sources of renewable energy most likely will come from wind and solar power, with wind dominating in the nearer term. Consequently, for the remainder of this Section, attention will be focused mostly on wind.

52 1 a) Resource Characteristics Characteristics of Wind. The principal characteristics of wind (and to a somewhat lesser degree, solar) IR sources of power supply are: (a) unstable supply patterns, (b) inconsistent (always changing) variability of supply (sometimes on a minute to minute basis, and certainly on a daily basis where wind is chronically, unpredictably variable), and (c) unwieldy operations control problems, including unreliable ramping characteristics, limited load following and regulation capabilities, and over-generation mitigation challenges. Core Problems with Wind Resources. The CAISO points out why these resource characteristics are particularly vexing from an operator s viewpoint using a series of graphs. Inverse Correlation of IR Supply and CAISO Load. First, the CAISO points out that wind generation patterns tend to be inversely correlated to daily load curves, and this creates ramping problems, among other things.

53 Figure I- 1 Wind/Solar Interdependencies and their Implications. Second, wind and solar sourced IR patterns are quite different, so the specific mix of IR contributions from wind and solar potentially impacts grid reliability very significantly. The graph below is a typical pattern of wind variability against solar variability.

54 Figure I- 1 Ideally, a greater focus on solar sources may achieve a more predictable IR resource delivery profile, especially when strategically bundled with wind; but wind volumes will continue to be much greater in the near-term (that is, until solar cell conversion rates and storage technologies reach a point where higher volumes and longer, sustained delivery capabilities are possible). The tradeoffs between wind and solar may only slightly offset for IR variability in the near future, which helps explain the CAISO s amplified concerns over the impacts on grid reliability of a substantial increase (to 0 percent or more) in wind derived IR. This point is remarkably driven home in the following graph.

55 Figure I- 1 The CAISO s concerns about the impact of wind-dominated IR in the resource stack appear to include the question of how daily loads can be reliability served when IR grows as a portion of the resource stack if mechanisms for smoothing variability are not aligned with the actual flow patterns of IRs. The point is made clearer by the following graph.

56 Figure I- 1 Operator Issues Regarding Wind Resources. At a level of granularity much more important to grid operators, the daily reliability assurance question is made up of highly time sensitive reliability challenges. As IR becomes more important to the overall resource mix, these operations-level intra-day and intra-hour IR supply variances increase the complexity of ensuring a continuously reliable, high quality, uninterrupted flow of power to consumers This is the core problem that the CAISO is motivated to fix now and for upcoming challenges as IR volumes increase; the CAISO focus being on finding non-ir mechanisms for smoothing IR variability. The next graph shows an actual day in the life of wind energy supply, which is a deeper dive into data granularity to reinforce the above noted key point. 0

57 1 1 The graph below requires a brief introductory explanation. When reproduced as a grayscale rendering, it is difficult to read because the color version uses twenty-nine different line colors, one for each day of the month, for each hour of the day. The visual density is further intensified with an average wind pattern line that can be seen running across the graph between the 00 MW and 00 MW horizontal ticks. In grayscale, twenty-nine graph lines, each with a different pattern, adds up to looking like static, or sine wave noise. This is the essential point of the graph; that with such remarkable daily variability, where that variability can be different for each day of a month, controlling the potentially disruptive effects of wind-derived IR on overall grid stability, and hence grid reliability, is a uniquely new problem. To repeat the key point: the challenge is real today, where wind power is still a relatively small share of the resource mix and only will amplify as the state s renewable energy objective is achieved in the coming years. Figure I- 1 1 While the above graph is visually compelling for understanding the CAISO s motivation, concerns, and objectives regarding IR, it is equally compelling as a visual tool for 1

58 introducing the discussion of challenges and barriers that make use of DR as a counter-resource for offsetting IR variability a complicated problem. 1 b) Challenges with Intermittent Resources Forecasting Variances. One consequence of the above noted IR characteristics is that the forecasting variances of CAISO and LSE providers for meeting load requirements may increase. Studies noted previously indicate this to be the case in other market spaces. The key point is that forecasting error increases insofar as IR variability is not smoothed out. Ramping. IR ramping capabilities certainly vary by season, but practically speaking, they vary hourly and in particular intra-hourly. As the mix of IR in the resource stack increases, the less reliable ramping characteristics of IR drives ramping support requirements to other resources, including, possibly, certain types of DR for Ancillary Services. Because the implications of increased IR in the resource stack are still evolving, the consequences in terms of everything from market pricing of DR to generator reliability (adjusted because increased ramping pressure may require changes in predictive, preventative, and ongoing operational maintenance) are open questions. Filling in for missing IR that instantly drops or fades quickly as wind drops also can change the volumes and values of CAISO s sourcing through its various markets, from Day- Ahead through Ancillary Services. IR Variability and Reserve Risk Management. The IR supply characteristics discussed above logically can drive operators to seek instant access to more reserves, and more types of reserves, because as the pattern variability increases, while locational impacts become more targeted as a function of MRTU locational marginal pricing and congestion management, more traditional ramp-rate types of resources may be insufficient to ensure sustained grid reliability. 1 Graphs sourced from CALIFORNIA INDEPENDENT SYSTEM OPERATOR RENEWABLE INTEGRATION STUDY, a presentation by Clyde Loutan. Document is undated, but was presented to a stakeholder working group in September or October 00.

59 Thermal Storage as an Offset for IR Variability. The CAISO has suggested that certain types of thermal storage may be attractive candidates for IR smoothing, as well. For instance, ice storage for cooling purposes may be an attractive source of spinning and non-spinning reserves that could be matched to specific IR sources and associated variability. Additionally, the CAISO has suggested that traditional ice storage solutions may be complimented, or substantially enhanced, by targeting new, smaller storage solutions, which might be distributed more widely and thereby aggregated to provide greater volume and call option flexibility for CAISO IR variability problem solving. Need for Rigorous Testing of Thermal Storage Matched to IR. This hypothesis requires a more thorough articulation of the ice storage model, followed by thorough research design, which includes a pilot project of sufficient magnitude to truly test the hypothesis. The model needs engineering detail that would enable a reasonable costing of the equipment, systems, etc., necessary for end-users, DR Providers, LSEs, and the CAISO to assess whether, even at the design level before money is spent on a pilot, there is a compelling business case that can be made for doing such a pilot. Broaden the Definition of the IR Challenge. SCE suggests that part of the very real challenge of smoothing IR as it grows as a supply source is in how the challenge is defined. DR, using the CAISO s and Commission s definition of DR, may be able to contribute to IR smoothing to some degree. But the scope, scale, and effectiveness of DR matched to IR is, at this early stage of consideration, highly uncertain, because information appears to be limited from other markets and regions regarding how DR has been successfully matched to smooth out IR variability. A reasonable question to explore is whether IR smoothing should be examined in a broader context that includes EE, PLS, and various forms of DG. The volume potential for smoothing, and the ability to minutely shape counter-resources to offset for IR variability, may be greater than what might be achievable with DR alone. Further, a more broadly configured solution matrix that may include specific EE, PLS, and DG solutions as well as combinations thereof may enable the CAISO to increase the volume contributions to its NPL, PDR, and PL products as well; and an enhanced

60 understanding of such a solution matrix may nominate other types of market products that better fit the specific objectives related to mitigating IR variability. c) Solutions: Principles and Objectives Match the Scope of the Problems with the Severity of Impacts. Rather than proceeding with the assumption that DR, as defined by the CAISO and the Commission, is the best source of counter-resources that can smooth IR variability, SCE advocates that the scope of the potential solutions match the severity of the problem, which should be defined in terms of risks to grid reliability for California and for specific LSEs. Needed Value at Risk Analysis. SCE s view is that the quantification of grid reliability risks from growing IR contributions may merit further examination. The focus of further analysis should consider not only operational adjustments that may be required, but also the valuation of those adjustments and the effects of IR variability on wholesale market pricing. In other words, problem solving would be informed by a value at risk analysis of the CAISO s concerns. After completing such analysis, if such analysis has not been done already, the starting point for IR variability problem solving will be enhanced so that consideration of whether a broader scoping that most likely includes EE, PLS, and DG, as well as DR, better serves the CAISO s focus on solving the IR variability problem set. Need for Working Group Process. To ensure that the risks to grid reliability are well documented and understood by all concerned parties, SCE advocates for a working group process focused on impacts of IR on grid reliability, wherein specific products can be shaped to ensure maximum flexibility for the CAISO s efforts at smoothing IR variability. The subject matter could be folded into WG, as discussed previously in this volume; or it could be assigned to another working group in the mix of the five working groups originally specified by the CAISO for developing DR products for its wholesale markets; or in the alternative, it may merit a separate track that is not constrained by the specific focus on creating wholesale market DR programs. Guiding Principles. The guiding principles for this approach should include the following:

61 Make sure that the risks and severity of the IR variability problem are fully articulated, understood by all impacted parties, and the array of impacts, or consequences, are mapped as comprehensively as reasonable given the early stage of this problem area. Make sure that the solution matrix includes all significant possible ways and means of mitigating problems that IR variability creates. Be realistic about what types of market solutions are possible in terms of products and alignments with the CAISO s market structure, and where gaps may exist after market solutions are configured; develop necessary solutions to ensure a complete treatment of the overall problem area. Act now to maximize the time available to thoroughly consider the problem area and develop appropriate solutions. If initial research and analysis determines that the problem area may solve itself through existing market activities, or that even at a twenty percent mix in the resource stack does not impact system reliability as hypothesized, terminate efforts and consideration of IR smoothing as an option. Objectives re Smoothing IR Variability. The objectives to be achieved in considering the important question of how to smooth IR variability as it grows, as a percentage of the resource stack, should include: Identifying the appropriate counter-resources for achieving optimal mitigation, or smoothing, of IR variability. Specifying the best products and fits with the CAISO s market structure for maximizing control of IR variability within a market context. Detailing the design, development, and deployment map for positioning the CAISO to maximize its grid control over IR variability (where that control is fully valued using market mechanisms to the extent feasible).

62 d) Issues and Barriers to Solving IR Challenges Refer to Issues Previously Discussed. Most of the issues and barriers discussed concerning PDR and PL for DR contributions to the CAISO s wholesale markets apply to using DR for smoothing IR as well. In other words, the infrastructure, systems, and process requirements necessary for enabling DR to serve the CAISO s wholesale markets as PDR and PL will be essential to any efforts to fit DR into serving as counter-resources that offset IR variability. Readiness to Address IR Variability viz-a-viz DR for Wholesale Markets. In addition to all of these PDR and PL related considerations, trying to smooth IR may be, as discussed above, somewhat behind where the CAISO and the Commission are concerning DR shaped as NPL, PDR, and PL for the CAISO s wholesale markets. Need for Specific Operational Definitions. More specific operational definitions of the IR problem set and its associated risks and impacts may be needed before the solution space can be confidently framed and scoped. The details on this point were covered above in Section 1... IR Variability and the Energy Supply Chain. The discussion of DR readiness and requirements that have been presented in this volume focus primarily on structural, systems, process, and work flow changes required to align IOUs with the CAISO around the CAISO s specified DR products (NPL, PDR, and PL). Also, issues associated with realigning the roles of LSEs and DR Aggregators has been specified; and the triangulation of the CAISO s market processes and requirements with both IOUs (LSEs) and DR Aggregators has been emphasized herein. In other words, if looked at from a supply chain perspective, the focus for DR readiness has been at the customer end of the supply chain, as it should be for DR. How Comprehensive are the Change Requirements? By contrast, it is less clear that the same space is appropriate for considering how to smooth IR variability. More of the energy supply chain may be impacted in terms of structural, systems, process and work flow changes. For instance, if DG technologies are included in the solution matrix for smoothing IR, then changes in generation structures, systems, processes and work flows may be necessary. If a generator that also has a trading arm seeks to incorporate DG into its bids into the CAISO s markets, targeted for spin and non-

63 spin, which could be called for mitigating IR variability, it may require an array of adjustments to how the enterprise is managed, submits bids, processes settlements, etc. Structure, Systems, Processes and Work Flows. The graph below visually depicts, at a high level, some of the main systems and processes associated with generators and DR sources being able to bid into the CAISO s wholesale markets. The graph is not comprehensive, it is indicative. It does not include adjustments that can be expected as MRTU Release 1 and MAP deploy. Figure I- 1 1 The point of the graph is to establish the current state so as to display below in a related graph some of the critical structural, systems, process and work flow adjustments that are necessary for aligning with the CAISO s MRTU Release 1 and subsequent MAP deployment. The additional requirements are indicative, not comprehensive; and while they focus on adjustments necessary for attempting to maximize DR contributions to wholesale markets as expeditiously as possible, this high level look at the required changes applies equally to adapting DR and other offset resources to the objective of smoothing IR.

64 Figure I- 1 Scope and Significance of IR Challenges. Offsetting IR variability may be as significant a challenge in itself as the current comprehensive effort to maximize, as fast as possible, the migration of DR programs into price-responsive, bid-based wholesale market resources. Scope of Impacts. There are similarities to these two challenges in terms of the scope of potential structural, systems, process, and work flow changes, but the impacts on the overall energy supply chain of IR smoothing may differ from the more focused effects of maximizing the capture of DR from the customer end of the energy supply chain. IR as a Broader Challenge than DR Per Se. These issues of scope, the early stage of IR problem definition, and the need for greater clarity regarding the risks and associated impacts of IR argue for carving out IR concerns and considering them as a separate matter. This may assist the

65 CAISO in more specifically articulating its vision of how DR can help offset for variability, and hopefully it opens the field for consideration of other related sources of IR smoothing. SCE Commitment. SCE supports the CAISO s concerns about IR impacts on grid reliability and is committed to assisting the CAISO in fully articulating the problem, and developing appropriate solutions. SCE advocates a working group process for achieving thorough problem articulation and the development of workable solutions. SCE believes the following steps will enable forward progress in a productive, efficient, and timely manner appropriate to the current stage in the evolution of the CAISO s concerns about IR variability. e) Proceeding Forward to Address IR Challenges First, SCE recommends, as stated previously in this volume, that a working group focused on IR problems be formed, or that WG s scope of work include consideration of the CAISO s concerns about IR. Second, SCE s proposed pilot project for DR as participating load (PLwill assist the CAISO in understanding the potential for DR to provide offsetting, or counter, resources that help smooth IR variability. SCE will include in its pilot project charter and implementation attention to IR variability issues, and will make sure that the results from its PL pilot are made available for analysis by the appropriate WG subgroups working on IR. This will give the CAISO a broader look on a market category basis at DR as a potential IR variability offset. Third, insofar as the CAISO has interests and the Commission so directs SCE, other pilot projects could be configured and tested during the 00-0 period. The test space is potentially broad and deep. As a way of illustrating how to develop possibilities, the following matrix is presented. The matrix is suggestive. Actual projects should be derived from a working group process that involves, at minimum, the IOUs, the CAISO, DR Aggregators, and affected customers.

66 Figure I-1 1 The above matrix also is useful for conveying a previous point, which is that achieving the CAISO s objective of maximizing the smoothing of IR variability may have a higher probability of success by broadening the scope of potential sources of offsetting resources f) Summary of What Can and Cannot Be Done Now and Why What Can be Done Immediately. Because SCE considers the case for IR smoothing to be more early stage than adapting existing retail DR programs to the CAISO s wholesale markets and DR products for those markets, what can be done immediately to advance the CAISO s interests in smoothing IR variability are primarily research, design, planning and pilot program developments. SCE, and presumably the other IOUs, is open to working with the CAISO to identify and detail specific pilot projects that can be submitted to the Commission for review and approval that will 0