1 Long-term Renewables Planning Methodology, Inputs and Assumptions for the 2010 Long-Term Procurement Plan Proceeding California Public Utilities Commission June 18, 2010

2 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

3 Scope and Purpose Workshop scope: RPS planning in the LTPP system proceeding –Not the broader LTPP system planning activities Workshop purpose: –33% RPS Implementation Analysis – Review study, comments –RPS Planning in the 2010 LTPP – Begin refining assumptions and methodologies for an updated study

4 Scope and Purpose Proceeding scope: LTPP system proceeding is not the forum for deciding RPS procurement obligations. Rather, results from RPS analyses could serve three purposes: –Identify type and quantity of new system resources needed reliably integrate renewables – Decided in LTPP system proceeding –Identify high-level (conceptual) transmission needs to meet renewable targets – Coordinated with ISO’s transmission planning process –Generate alternative RPS procurement strategies – Inform the RPS proceeding

5 Background Previous Commission Decisions –Since D , the Commission has stated its intent to integrate long- term RPS planning into the LTPP proceeding. –D (2006 LTPP decision) directed parties to work with Energy Division to refine a methodology for RPS resource planning and analysis February 14, 2008 Order Instituting Rulemaking (OIR) for the 2008 LTPP proceeding (R ) –OIR launched to integrate and refine procurement policies, including further analysis regarding the feasibility and cost of 33% renewables August 28, 2008 Assigned Commissioner’s Ruling (ACR) and Scoping Memo –Established separate track for Energy Division’s 33% RPS Implementation Analysis project June % RPS Implementation Analysis: Preliminary Report

6 Background (cont’d) July 1, 2009 Amended ACR and Scoping Memo released the Energy Division Straw Proposal on LTPP Planning Standards (Staff Proposal) Working Principle – Resource plans should consider the scale of investment in transmission and flexible fossil resources to integrate and deliver new renewables –Commission “expect[s] the data produced out of RETI…to be utilized in the [LTPP] proceeding.” (OIR at p. A-9) –Transmission permitting based on TEAM decision relies on use of assumptions that are consistent with resource plans and system assumptions used in the procurement proceedings (D at OP 1 and p. A-2) A single, statewide “Renewables and Transmission Study” is needed as a foundational element

7 R OIR (2010 LTPP) Three Tracks –Track 1: System –Track 2: Bundled –Track 3: Rule and Policy issues System track is “to indentify CPUC-jurisdictional needs for new resources to meet system or local resource adequacy need over the planning horizon, including issues related to long-term renewables planning”

8 R (cont’d) Outlines several staff proposals which would be issued by ruling: –Planning Standards (Part 1) –Planning Standards (Part 2): RPS –Planning Standards (Part 3): EE

9 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

10 Proposed Definitions Scenario – A possible future state of the world encompassing assumptions about policy requirements, market realities and resource development choices. Sensitivity – A test to measure the change in output variable (e.g., cost, resource need) due to a change in input assumptions and parameters. Case – A set of input assumptions and parameters (e.g., gas price, or electricity demand) under a given scenario that drives the selection of a given portfolio of resources. Portfolio – A set of electric resources, both supply-side and demand-side, that provides electric service to all system ratepayers, under a given scenario.

11 Proposed Guiding Principles for RPS Planning Inputs, Assumptions and Methodology 1.Assumptions should reflect the behavior of market participants, to the extent possible 2.Methodology should be consistent with previous regulatory decisions, to the extent applicable 3.Any proposal should explain the policy basis for the proposal 4.Any proposal must include supporting documentation

12 Proposed Guiding Principles for RPS Planning RPS Scenarios 5.RPS scenarios should be reasonably feasible and reflect plausible procurement strategies with associated (conceptual) transmission. 6.RPS scenarios should represent substantially unique procurement strategies resulting in material changes to corresponding (fossil) procurement needs and/or required (conceptual) transmission. 7.RPS scenarios should be limited to 3-5

13 Proposed Approach to Scenario Development RPS generation under dev’t & contracted to CA by Q Discounted Core Cost- constrained Scenario Time- constrained Scenario Environmentally- constrained Scenario Trajectory Scenario Balanced Scenario – determined 2011? RPS generation delivering to CA by Q % of 2020 Retail Sales

14 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

RPS Model Methodology Arne Olson, Partner Doug Allen, Consultant

16 Contents Updates to Previous 33% Implementation Analysis Resource Potential, Cost, and Performance Portfolio Selection Methodology

Updates to Previous 33% Implementation Analysis

18 Current Analysis Compared to 2009 Implementation Analysis 33% RPS Implementation Analysis was focused on informing state RPS policy Key tasks: 1. Develop “plausible scenarios” for achieving 33% by Estimate the net ratepayer costs associated with each plausible scenario 3. Highlight key obstacles to reaching the 33% goal, including the need for new transmission and the integration costs of new resources The current analysis is focused on informing state RPS planning Key tasks: 1. Develop “plausible scenarios” for achieving 33% by Estimate year-by-year resource build-outs under each alternative scenario 3. Feed information into LTPP proceeding to inform the Commission’s decisions regarding fossil procurement

19 Weaknesses of Previous 33% Implementation Study New transmission is assumed for most projects  No way to determine which projects could get built without new lines Renewable projects are selected in aggregated “bundles” and could not be selected individually Crude methodology for addressing potential out-of-state REC resources (“Out-of-State Early”, “Out-of-State Late”) Did not look at operational impacts of renewables at such high levels of penetration  Integration costs for intermittent resources based on a rule of thumb Lack of transparency in handling of short-listed CPUC projects Project viability ratings not very scientific

20 Key Improvements in Portfolio Development Methodology Improved handling of transmission requirements for new renewable resources  Allows some resource to be delivered over existing transmission or with minor upgrades  Allows individual selection of Non-CREZ resources Incorporation of unbundled REC resources  Can now model out-of-state REC-only resources as well as out-of- state CREZs Incorporation of “Discounted Core” of commercial projects  Projects in advanced stage of permitting and development

21 Other Updates to Analysis Improved detail on Commercial projects in CPUC ED Database  CREZ assignments  2009 solicitations  Confidential projects either aggregated or excluded to improve transparency New Aspen environmental scores Updated resource list based on RETI Phase 2B analysis Updated RPS Net Short based on:  2009 IEPR Load Forecast  New load decrements for EE and CHP calculated by CPUC  Existing renewable resources from 2008 Net System Power Report

Resource Potential, Cost, and Performance

23 General Approach Determine renewable resource gap (GWh) in 2020 Compile database of resources available to meet RPS target Rank available resources based on cost, commercial interest, environmental sensitivity and timeline Select resources to fill renewable resource gap

24 Determining Demand for Renewables Demand for renewables in California is based on 2020 RPS target, equal to 33% of eligible retail sales  2020 retail sales based on CEC 2009 IEPR load forecast  Excludes retail sales by small LSEs (<200 GWh/yr) Estimate quantity of renewable resources online in base year  Used renewable resource “claims” from CEC 2008 Net System Power Report  Added new resources online in 2009 based on ED Database RPS resource gap is the difference between the 2020 target and the 2009 renewables claims

25 Renewable Net Short

26 Sources of New Resources to Fill Resource Gap 1. Commercial Projects  ED Database of IOU projects  POU procurement plan data obtained from CARB 2. Additional “Theoretical” Projects  RETI pre-identified and proxy projects for California  WREZ projects for the rest of the WECC 3. Original Renewable DG resource potential estimates  Developed as part of 2010 LTPP

27 Updated ED Database Analysis incorporates the latest ED Database, including 2009 IOU solicitations  Contracted projects are included individually  Shortlisted projects are aggregated by resource type and zone in cases where there are at least three such projects to preserve confidentiality (otherwise they are left out) Distribution of projects among zones has changed since previous analysis

28 Changes in the ED Database

29 Breakdown of the ED Database

30 Treatment of Commercial Projects Commercial projects are divided into two categories: 1. Discounted Core: Project has obtained or has made significant progress towards obtaining a permit 2. Non-Discounted Core: Project has made limited progress towards obtaining a permit Discounted Core projects are given first priority in the resource selection sorts, reflecting their high probability of development Non-Discounted Core projects are given priority over generic resources but are not guaranteed development POU planned resources treated as Non-Discounted core

31 POU Resources E3 has included POU resource procurement plans in the development of the resource portfolios based on CEC data POU resources are included as Non-Discounted Core commercial resources

32 RETI Phase 2B Database The updated RETI Phase 2B Database contains site-specific information for renewable resource potential, cost, and performance in California Out-of-state resources from the WREZ Transmission Model have been incorporated with the RETI data  WREZ estimates of potential represent high-quality remote renewable resources that would require significant transmission upgrades to reach load centers

33 Out of State Renewable Resource Data from E3 Models E3 maintains a database of renewable resource cost and performance data in the West  Wind and solar data based on NREL GIS modeling  Geothermal and hydro data from EIA  Biomass aggregated from various sources  Additional resource data for BC and Alberta Used to supplement WREZ data for out-of-state resources (Montana, Colorado, BC, Alberta)

34 Estimates of Statewide DG Potential As part of the 2010 LTPP, E3 and Black & Veatch collaborated to develop original estimates of the statewide potential of solar PV The RPS model integrates these estimates, allowing it to evaluate the viability of the development of these resources

35 Resource Cost and Performance E3 used site-specific data on resource cost and performance where available (RETI and WREZ projects) Generic assumptions were developed for resources without specific information based on averages of the RETI data (shown below)

36 Generic Resource Cost Assumptions

37 Transmission and Geographic Classification of Resources Each resource is assigned one of three classifications 1. CREZ: resources located within one of the 48 Competitive Renewable Energy Zones (either in California or in other states) 2. Non-CREZ: resources in California or directly across the border that are not located within a CREZ and can be delivered with minor transmission upgrades 3. Out-of-State REC: out-of-state resources that would deliver energy to the local market

38 Transmission Bundles Resources in CREZs are aggregated into transmission bundles in the following order: 1. Existing transmission bundle 2. Minor upgrade bundle 3. New transmission bundle Discounted core projects given first priority to fill each transmission bundle  Non-core Commercial projects given next priority to fill the New Transmission bundle  Any remaining transmission capacity in the bundle is allocated to the lowest-scoring generic projects  Up to 3000 MW of new transmission allowed for each CREZ

39 Examples of Transmission Capacity Allocation

40 Examples of Transmission Capacity Allocation

41 Out-of-State REC Resources Out-of-State RECs previously included in two “zones” (“Out-of- State Early” and “Out-of-State Late”) New model can select REC resources individually Assume physical limitations on wind integration for each region REC resources priced at long-run “Green Premium” or Cost minus Value  REC resources optimized for access to transmission, not for resource quality – RECs are allowed for average quality resources, not best sites  Pricing based on long-run cost, not REC market price forecast (analogous to in-state resources)  No pricing distinction between different types of RECs (bundled vs. unbundled, with or without delivery requirement)

42 Unbundled Out-of-State REC-Only Transaction Leg 1: Developer to Mid-C Mid-C Market Leg 2: CAISO to load Pure REC transaction with no energy purchase requirement and no delivery requirement Developer sells energy at Mid-C California LSE purchases REC from developer at LCOE minus Mid-C price Separately, California LSE arranges for energy transaction from CAISO market to load California LSE never owns energy No incremental imports to California

43 Out-of-State REC with Delivery Requirement Leg 1: CA utility to Mid-C Mid-C Market Leg 2: CAISO to load REC transaction with energy purchase requirement and delivery requirement California LSE purchases energy and REC from developer at LCOE of wind facility and sells energy at Mid-C Separately, California LSE arranges for energy transaction from CAISO market to load California LSE rebundles REC with transaction from Mid-C to CAISO that would have occurred anyway! No incremental imports to California Leg 3: Mid-C to CAISO

44 Pricing of Out-of-State REC vs. In-State Resource In-State Resource priced at LCOE, ratepayer impact is cost relative to market value or “Green Premium” Out-of-State REC priced directly at “Green Premium” Energy and capacity values vary by market (higher in California) Pricing is the same for all flavors of RECs Example of In-State vs. REC-only Pricing

45 Physical Limits on Out-of-State REC Supply Market Value of Wind Energy $/MWh MW of Intermittent Renewables Initially, wind displaces gas resources More wind reduces market prices and raises integration costs Value decreases significantly as wind displaces baseload There is a practical limit to how much intermittent energy each zone can easily accept

46 Physical Limits on Out-of-State REC Supply There is a practical limit to how much intermittent energy each zone can easily accept Market Value of Wind Energy $/MWh MW of Intermittent Renewables Initially, wind displaces gas resources More wind reduces market prices and raises integration costs Value decreases significantly as wind displaces baseload E3 limited the REC supply based on a simplified representation

47 Limits on Wind Penetration Ability to easily absorb wind is limited to load served with flexible generation E3 estimated hourly flexible generation in each zone:  Load – Nuclear – Coal – Base Hydro + Export Transmission Capacity  Wind limit is min value Other regions also have RPS requirements – assume CA can soak up 50% of each region’s limit 6,461 MW 2,257 MW 738 MW 404 MW 1,231 MW 461 MW 229 MW 47 MW 0 MW 3,665 MW 1,371 MW 3,968 MW 1,939 MW 2,135 MW 947 MW 1,700 MW 850 MW 2,211 MW 808 MW 13,745 MW Total limit on wind in region Wind available to California

Portfolio Selection Methodology

49 Resource Selection Methodology 1. Calculate project score for each resource 2. Allocate lowest-scoring out-of-state theoretical projects to other states until all non-CA WECC RPS targets for 2020 are satisfied 3. Rank remaining CREZ projects and select to fill transmission bundles 4. Calculate aggregate score for each transmission bundle 5. Rank transmission bundles against individual non-CREZ and REC resources 6. Select resources and bundles to meet 33% RPS target in 2020

50 Detailed Portfolio Development Resource Sort for Local Use Potential CREZ Resources Potential Non-CREZ and REC Resources Resource Sort for CA Use as RECs Resource Sort for CA Use Towards RPS Resources Remaining After Local Sort Resource Sort for Existing Tx Resources on Existing Transmission Resources Remaining After Existing Tx Sort Resource Sort for New Tx New Transmission Bundles Resources Remaining After Local Sort Non-CREZ and REC Resource Rankings Resources Selected for Local Use Resources Selected for CA RPS Portfolio

51 Project Scoring Methodology Each project is scored on a scale based on four metrics (0 is better):  Net Cost  Environmental Score  Commercial Interest Score  Timing Score Final score for each project is a weighted average of the four individual metrics  Weights are user-defined and vary by scenario

52 Net Cost Score Cost score is based on a modified version of the RETI Ranking Cost Includes integration and T&D avoided costs Scores are converted to 0 – 100 scale, bounded by the model’s lowest and highest net cost resources Modified RETI Ranking Cost +Levelized cost of energy +Interconnection (gen-tie) costs +Deemed integration costs +Levelized, per-MWh incremental transmission costs –T&D avoided costs –Energy value –Capacity value =Final project ranking cost

53 Environmental Score Handicaps resources in areas where environmental issues might hinder development Considers a variety of factors:  Disturbed lands  Right-of-Way  Significant species  Air quality  Others Scores for each resource in each CREZ on scale

54 Commercial and Timing Scores Commercial Score: Scale of reflecting contracting activity of California utilities  Commercial projects receive a score of 0, while generic projects receive a score of 100  POU-planned projects considered “Commercial” and receive score of 0 Timing Score: Gives better score to resources that can be developed on a relatively short time scale  Online date 2021 gets 100  For ED database projects, online dates filed with the applications  For other resources, dates based on size and type of project

55 Selection of RPS Portfolio Each transmission bundle is assigned an aggregate score based on an average of the constituent resources and compared against individual non- CREZ and RECs resources Discounted Core Projects are selected first unless in New Transmission bundle After Discounted Core, resources & bundles with the lowest score are selected to fill the 2020 RPS gap

56 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

57 Proposed Approach to Scenario Development RPS generation under dev’t & contracted to CA by Q Discounted Core Cost- constrained Scenario Time- constrained Scenario Environmentally- constrained Scenario Trajectory Scenario Balanced Scenario – determined 2011? RPS generation delivering to CA by Q % of 2020 Retail Sales

58 Approach to Discounted Core Goals: –Capture the projects most likely to be online by 2020 –Transparency – rely on public, objective data Propose to include projects with both: –PPA signed and approved or under review by CPUC –Major permit filed and deemed data adequate at relevant permitting agency (Energy Commission, County, etc.) Tested other cores: –All projects with PPAs signed and approved/under review –All projects with certain objective level of developer experience –Model allows users to to test these other cores

59 Makeup of Possible Discounted Cores CoreDefinition Total GWh Total MW In State Solar PV GWh Solar Thermal GWh Wind GWh Geo- thermal GWh Biomass GWh ARRA Projects 1 New/repowered with PPA signed 34,20611,83075%18%35%33%5%4%22 2 New/repowered with PPA signed and permit application deemed data adequate 24,4068,90672%19%27%46%2%5%18 3 New/repowered with PPA signed and permit application approved 11,4423,79859%7%10%78%3% 5 4 New/repowered with PPA signed and a minimum levels of project development and ownership/O&M experience, and technical feasibility 22,0887,48779%13%32%44%6%4%15

60 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

61 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

62 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

63 Transmission Bundles (review) Resources in CREZs are aggregated into transmission bundles in the following order: 1. Existing transmission bundle 2. Minor upgrade bundle 3. New transmission bundle Discounted core projects given first priority to fill each transmission bundle  Non-core Commercial projects given next priority to fill the New Transmission bundle  Any remaining transmission capacity in the bundle is allocated to the lowest-scoring generic projects  Up to 3000 MW of new transmission allowed for each CREZ

64 Examples of Transmission Capacity Allocation (review)

65 Transmission sizing and cost Weakness of previous study: assumed no capacity on existing transmission system; upgrades essentially all single or double circuit 500kV lines “trunklines” Updates: –ISO high-level assessment of capacity from various CREZs 1.) over existing system; 2.) over relatively minor new upgrades –Additional lines on case-by-case basis: 500kV for large and out- of-state zones, smaller lines for smaller zones close to loads. Cost = function of length and capacity –Incorporation of RETI work has been difficult – many CREZs tied to single line segments; many line segments tied to single CREZs; but additional detail would help in cost and timing review

66 Transmission sizing and cost See: –California ISO assessment of capacity over existing system and minor upgrades –Zaininger Engineering assessment of capacity over RETI lines Available here:

67 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda

68 33% RPS Implementation Analysis Timelines created for the 33% Reference Case – not for any of other 3 cases –Timeline 1: Historical experience, reforms/risk –Timeline 2: Reforms, no risks –Timeline 3: Reforms + external risks Goal: Better understand the tradeoffs between cost, risk, and time associated with particular procurement strategies Identify market and regulatory barriers to renewable development Identify solutions and their impacts on achievement of a 33% RPS

69 Example – Aspen Environmental Group research into generation timelines

70 Example – Research into transmission timelines

71 33% RPS Reference Case – Timeline 3

72 Timeline Tool Black & Veatch tool under development to automate timing considerations, create scenario timelines and supply curves Generic generation scheduling factors according to: –Technology –Project size –Land type –Location Generic transmission schedules for several types of lines User has ability to change/define inputs and timing assumptions; change default schedules and MW counts based on technology and other factors, etc.

73 Timeline Tool Data already incorporated –Contracts approved and pending approval – basic project information –Proxy projects – RETI, others to be added ED staff updating tool for consistency with new scenarios still being added See mock-up of tool here:

74 Generic generation timing assumptions Little change from Implementation Analysis: Project Type Development Length (months) excluding transmission Estimated Commercial Online Date Biogas/Biomass < 50 MW > 50 MW Geothermal < 50 MW > 50 MW Small Hydro Solar Thermal < 50 MW > 50 MW Solar PV < 50 MW > 50 MW Wind < 50 MW > 50 MW ED Database projects –Filed/approved by CPUC (public) –Under negotiation (confidential) - Per public contract information - Per generic estimates above

75 Transmission timing assumptions Transmission Schedule Type Transmission Planning by CAISO/ POU/ WECC (months) Project Description Prep by Utility CEQA/ NEPA Review By CPUC/POU/ Feds Final Review and Approval by CPUC/ POU/Feds Final Design And Construction by Utilities Total Existing / Distributed Typical Typical - Short Typical - Long Long-Distance Tehachapi Tehachapi Sunrise Devers - CO River000030

76 Building scenario timelines Each zone and transmission bundle/increment is assigned to a transmission schedule → no generation in that bundle is available before transmission Non-CREZ, distributed resources, and resources accessing existing capacity not reliant on transmission → available per contract or generic generation schedule Each scenario is assigned an “online date” according to the availability of the last generation resource chosen for that scenario No “bottlenecks” considered now; what assumptions are appropriate?

77 Timeline Tool – Features to Come Aggregated portfolios for alternative cases presented with summary timelines and yearly generation charts Timelines for individual projects available for review Project development phases are broken out Example Only

78 Timeline Tool – Features to come (cont’d) Yearly generation charts can be broken out by viability class – high, medium, and low viability Example Only

79 Introduction, scope, context Terminology, high-level approach Methodology, Inputs, Assumptions –Portfolio development –Discounted core –Environmental scoring –PV assessment –Transmission sizing –Timing assessment Results Next steps for analysis Schedule, Summary Agenda