Loading...
HomeMy WebLinkAboutCouncil Reading File - Full Tri-County Feasibility StudyFINAL REPORT AUGUST 2017 TECHNICALFEASIBILITYSTUDY ON COMMUNITY CHOICE AGGREGATION FOR THE CENTRAL COAST REGION             This page intentionally left blank.  EXECUTIVE SUMMARY             This page intentionally left blank.  Executive Summary A. Community Choice Aggregation Overview Community Choice Aggregation (CCA) is a program for local jurisdictions in California to procure electricity supply for, and develop energy resources to serve, jurisdictional customers. According to the Local Government Commission,1 the most common reasons for forming a CCA program are to: ▪ Increase use of renewable generation, ▪ Exert control over rate setting, ▪ Stimulate economic growth, and ▪ Lower rates. When a CCA is formed, the local incumbent electric investor-owned utility (IOU) continues to deliver power through its transmission and distribution facilities to customers within its service territory. The IOU also provides monthly customer metering and billing services. The local CCA program procures the electric commodity and sells it to its customers, with the intent that the electricity is less expensive, more local, and/or uses more renewable generation than the current utility alternative. The two components, delivery and generation, already appear separately on customer bills. The incumbent utility continues to provide billing services, but the CCA’s generation rate replaces the IOU’s generation rate on customer bills. Jurisdictions in California have formed CCA programs in efforts to provide constituents the option to be served with a greater mix of renewable and carbon-free energy generation than is provided by the incumbent utility. Eight CCA programs are currently operational in California, with ten more launching in 2018. At least 17 additional jurisdictions are exploring and/or are in the planning stages for CCA. B. Study Scope and Purpose This technical feasibility Study for CCA for the Central Coast Region (Study) was directed by the Advisory Working Group (AWG), which was formed by eleven governments in the Santa Barbara, San Luis Obispo, and Ventura County (Tri-County) Region. The Advisory Working Group collectively has named the potential CCA “Central Coast Power.” The Study’s purpose is to advise and guide the Tri-County Region in understanding the feasibility of forming a new CCA program. This Study considers required startup and operational processes and evaluates multiple Ten local governments joined with the County of Santa Barbara to fund this Study, and the following jurisdictions formed an Advisory Working Group in December 2015: • Unincorporated San Luis Obispo County • Unincorporated Santa Barbara County, plus: o City of Carpinteria o City of Santa Barbara • Unincorporated Ventura County, plus: o City of Camarillo o City of Moorpark o City of Ojai o City of Simi Valley o City of Thousand Oaks o City of Ventura Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-2 procurement scenarios to determine whether a CCA program in the Tri-County Region is: a) financially feasible; and b) will meet its stated policy objectives. The Study results do not necessarily apply to one or more of the Tri-County local governments joining an existing CCA program. This Study evaluates the financial and economic viability of a CCA by: • Forecasting the CCA electricity demand requirements (load) and potential customers by class; • Estimating the costs of procuring the necessary electricity supply; and • Projecting the costs of starting up and administering a CCA program. The Study also enumerates the potential benefits and associated risks of a CCA program and discusses implementation requirements. C. Energy Procurement and Study Scenarios Energy procurement is complex and the total cost of procurement is subject to changes in both market conditions (price) and consumption (volume). Load Serving Entities (LSEs)—IOUs, CCAs, and Electricity Service Providers (ESPs)—must manage both load forecasting and energy procurement with a robust risk management approach to account for the dynamic and volatile nature of power markets and load. Given the uniqueness of multiple municipalities partnering to commission this feasibility Study, the Advisory Working Group established eight geographic participation scenarios. These eight scenarios were selected to explore the feasibility of different sizes and configurations for the CCA program and the potential effect of customer demographics. Although the entire Tri-County Region may not ultimately pursue CCA, certain jurisdictions may decide to move forward with CCA. The eight participation scenarios defined for this Study are: 1. All Tri-County Region 2. AWG Jurisdictions 3. All San Luis Obispo County 4. Unincorporated San Luis Obispo County 5. All Santa Barbara County 6. Unincorporated Santa Barbara County 7. All Ventura County 8. City of Santa Barbara In addition to the eight participation scenarios, three renewable energy content scenarios were considered. All scenarios include a customer option to opt-up to a 100% renewable energy product. For the purposes of this Study, 2% of customers were assumed to opt-up to the 100% renewable option. The three renewable energy content scenarios are as follows: Throughout the report, the term LSE is used to provide illustrative trends that are affecting the Tri-County Region as a whole, regardless of whether the electricity is provided by an IOU, ESP or CCA program. For our purposes, a CCA program is a subset of the more broad LSE term. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-3 • RPS Equivalent: This scenario assumes that Central Coast Power would offer its base electricity product to all customers starting at 33% renewable content in 2020 and ramping up to 50% renewable content by 2030 in alignment with the California minimum Renewable Portfolio Standard (RPS).2 • Middle of the Road: This scenario assumes that Central Coast Power would offer its base electricity product to all customers using 50% renewable content for the entire Study period. • Aggressive: This scenario assumes that Central Coast Power would offer its base electricity product to all customers using 75% renewable content for the entire Study period. This Study evaluates an eleven-year period from 2020 to 2030, although a potential CCA program could begin earlier than 2020. Figure ES-1 illustrates how the renewable energy content in the RPS Equivalent scenario grows over time, and in the other two scenarios remains constant across the Study period. These three scenarios were chosen to illustrate the relative differences in cost given different levels of renewable supply content. Actual CCA implementation may choose to follow a progression of increasing renewable generation over that period based on cost competitiveness. For example, Central Coast Power CCA may launch in 2020 with 50% renewable content and progress to 75% renewable content by 2030, assuming it can do so at a cost advantage to the IOUs. To enhance report readability, the main body of this report presents results for the AWG Jurisdictions participation scenario, for the RPS Equivalent, Middle of the Road, and Aggressive renewable energy content scenarios. Detailed results for the other seven participation scenarios are provided in Appendices C, and E through J. Figure ES-1 Renewable Energy Content Modeled in this Study Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-4 The fundamental operational role of a CCA is to forecast customer electricity needs and procure energy and associated energy related services. Power procurement consists of forecasting and risk management tasks. Power procurement planning and day-to-day decision making rely heavily on short-term and long-term forecasts of consumer demand for power. The procurement function must also evaluate and assess the inherent risks associated with demand forecasting and develop appropriate risk mitigation strategies. Though no one can predict future energy demand with 100% certainty, logical, data- driven, industry-standard methodologies to forecasting are available to provide a realistic outlook of energy demand under a variety of future scenarios. Brief discussions covering the forecasts for customer power demand and power procurement costs are provided in the following segments. D. Customer Demand As shown in Figure ES-2, Ventura County is the largest electricity consumer of the three counties considered in this Study, followed by Santa Barbara and San Luis Obispo Counties. Collectively, customers in the incorporated cities in San Luis Obispo and Ventura Counties consume more electricity than customers in the unincorporated county. The reverse is true in Santa Barbara County. The fundamental operational role of a CCA is to forecast customer electricity needs and procure energy and associated energy related services. Energy is measured in several units throughout this study: kilowatt-hours (kWh), which is the unit used on customer bills; megawatt-hours (MWh), where 1 MWh equals 1,000 kWh; and gigawatt- hours (GWh), where 1 GWh equals 1,000 MWh or 1,000,000 kWh. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-5 Figure ES-2 Annual Demand in Gigawatt-hours (GWh) by County Figure ES-3 shows the annual electricity consumption for each of the Study’s eight geographic participation scenarios. The consumption and number of accounts generally mirror each other, with the exception of unincorporated San Luis Obispo and Santa Barbara Counties. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-6 Figure ES-3 Annual Demand in GWh for Each Geographic Participation Scenario Electricity consumption is forecasted to grow moderately over the Study period, however continued customer adoption of distributed generation (DG) solar photovoltaic (PV) is expected to offset this growth. DG PV reduces the amount of energy that needs to be provided by the potential CCA. Figure ES 4 illustrates the growth of customer-owned DG PV since the year 2000 and illustrates a forecast for additional DG PV capacity if this trend continues. Table ES 1 lists the forecasted annual energy consumption, annual DG PV generation, and the annual net load (consumption-generation) served by the potential CCA for the AWG Jurisdictions participation scenario. In summary, a Central Coast Power CCA would likely sell less electricity each year given customer DG PV adoption. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-7 Figure ES-4 California Solar Initiative Incentivized Customer-Owned Solar Photovoltaic in the Region with 2030 Forecast Table ES-1 Load, Distributed Generation, and Net Load Forecast, AWG Jurisdictions Participation Scenarios Year Annual Energy Consumption (MWh) Annual DG Generation (MWh) Annual Net Load Served by LSE (MWh) 2020 6,698,164 164,987 6,533,177 2021 6,735,965 202,979 6,532,985 2022 6,777,276 244,414 6,532,862 2023 6,811,982 287,988 6,523,995 2024 6,868,761 335,074 6,533,686 2025 6,888,329 381,954 6,506,375 2026 6,930,669 431,948 6,498,721 2027 6,971,608 483,660 6,487,948 2028 7,026,296 538,288 6,488,008 2029 7,047,280 592,489 6,454,791 2030 7,085,173 650,280 6,434,893 Forecast Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-8 As explained in Section II Technical and Financial Analysis, the increasing amount of DG PV also creates more volatile customer load due to the variable nature of its energy output. Solar generation depends on solar irradiance, which can fluctuate significantly over very short periods of time (within seconds) due to weather patterns and resulting cloud cover. E. Power Procurement Cost Forecasts CCAs, like all LSEs, satisfy customer demand for electricity by managing a power supply portfolio, a collection of supply-side resources. For the purposes of this Study, a power supply portfolio is designed to acquire two distinct commodities: energy, typically measured in MWh, and resource adequacy capacity, typically measured in megawatts (MW). Energy resources include natural gas generation, RPS compliant renewable energy generation, energy storage, and California Independent System Operator (CAISO) day- ahead and real-time market purchases. Resource adequacy is used to make sure there is sufficient capacity to produce electricity during peak demand periods. This Study projects decreasing costs for all energy resources considered, except for energy procured in the CAISO markets, where average pricing remains constant and large fluctuations are due to variability in renewable generation for both utility scale resources and customer-owned DG PV. Actual CAISO real- time market prices from January 2014 through October 2016 for the Tri-County Region average around $36 per megawatt-hour (MWh). However, the range of prices around that mean varied greatly, reaching a high of $4,377 per MWh during shortages of supply relative to demand, and a low of -$1,277 per MWh— meaning that CAISO will pay participants to take power—when supply exceeds demand. The high level of DG PV penetration in California, combined with solar and wind energy’s variable nature, accounts for much of this market volatility. This Study has modeled renewable resource variability and the CCA’s associated exposure to CAISO market prices. Table ES-2 presents the Study forecast for the average annual power procurement cost for the AWG Jurisdictions participation scenario for the three renewable supply scenarios. As can be seen in these data, the average cost of power procurement for the CCA rises as more renewable energy content is added because renewable generation is forecast to be more expensive than alternative non-renewable resources, despite a slight downward trend in renewable energy prices. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-9 Table ES-2 Average Annual Power Procurement Costs ($ per MWh), AWG Jurisdictions Scenarios Year RPS Equivalent Middle of the Road (50% Renewable) Aggressive (75% Renewable) 2020 $67 $74 $87 2021 $66 $74 $85 2022 $66 $74 $85 2023 $66 $72 $85 2024 $66 $72 $84 2025 $66 $71 $84 2026 $67 $70 $84 2027 $68 $70 $84 2028 $68 $69 $83 2029 $68 $69 $82 2030 $68 $69 $81 The total energy requirements served by various power supply options, including PPAs, the CAISO day- ahead and real-time markets, among others, change depending on scenario, however, the price of each option does not. This is what would be expected in actuality, as the amount of energy procured by the CCA would have little to no bearing on the prevailing PPA and market prices on a long-term basis. In support of the power procurement cost forecast, data from the U.S. Department of Energy’s Energy Information Administration’s Annual Energy Outlook 2017,3 which provides estimates of renewable generation costs on a regional basis, were examined. This data is used by utilities, energy consultancies, and others to help understand current and future energy-related pricing trends and is based on real-world project construction, financing, ownership, and ongoing operations and maintenance costs. Table ES-3 shows the various costing components for a new solar photovoltaic project and a new wind project, assuming they are installed on sites where there is no need to work within the constraints imposed by existing buildings or infrastructure (greenfield projects). This cost data supports all-in pricing at around $67 per MWh for wind resources and $101 per MWh for solar PV resources. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-10 Table ES-3 Energy Information Administration Cost Estimates for New Wind and Solar Energy Resources in California Description Wind Farm – Onshore Utility-Scale Photovoltaic Configuration 100 MW; 56 turbines at 1.79 MW each 20 MW, Alternating Current, Fixed Tilt Installation Type Greenfield Installation Greenfield Installation Total Capacity (MW) 100 20 Capacity Factor (National Average, Jan. 2016-Apr. 2017) 36.59% 26.76% Total Project Cost, California-Mexico Region ($ per kW-installed) $2,010 $2,578 Total Project Cost, California-Mexico Region ($) $201,000,000 $51,560,000 Variable O&M ($ per MWh) $ - $ - Fixed O&M ($ per kW-year) $46.71 $21.66 Weighted Average Cost of Capital (%) 5.50% 5.50% Debt Finance Term (years) 20 20 Financing Costs per Year ($) $16,819,545 $4,314,506 Fixed O&M Costs per Year ($) $4,671,000 $433,200 Total Project Costs per Year ($) $21,490,545 $4,747,706 Energy Production per Year (MWh) 320,528 46,884 Per Unit Cost ($ per MWh) $67.05 $101.27 Like all energy price forecasts, the one used within the Study—just as those used within other CCA feasibility studies—may or may not accurately reflect actual future conditions, which are unknown and not predictable. Various market drivers may change resulting in different outcomes from those assumed here. The forecast used herein is a reasonable estimate for the purposes of analyzing the feasibility of CCA within the Tri-County Region, but no warranties as to the accuracy of forecast prices for power purchase agreements or CAISO market commodities are implied or should be inferred. For example, large hydroelectric generation resources owned and managed by the IOUs were not significantly utilized during the recent drought years through 2016. Rainfall in the winter of 2016-2017 filled the hydroelectric reservoirs, enabling a low cost, carbon-neutral generation resource for the IOUs. Generally speaking, all other things being equal, increased hydro production will lower IOU generation revenue requirements and could have a dampening effect on IOU rates, potentially lowering the rates required for the CCA to be competitive. F. Greenhouse Gas Emissions Impact This Study also evaluated the greenhouse gas (GHG) emissions impact of the renewable energy content of the CCA’s portfolio—including the 100% renewable energy product assumed to be chosen by 2% of customers—relative to that of the incumbent IOUs, Southern California Edison (SCE) and Pacific Gas and Electric (PG&E). For the purposes of this comparison, the IOU Base Case assumes the IOUs will progress from currently published 2020 RPS levels of renewable generation linearly to the 50% RPS goal in 2030. Although each IOU may elect to exceed RPS requirements as they have in recent history and relative to 2020 requirements, for example PG&E submitted a joint proposal to decommission the El Diablo nuclear power station and voluntarily reach 55% RPS by 2031,4 neither IOU has publicly stated firm plans to exceed RPS targets. California is currently considering Senate Bill 100, which would increase the renewable energy mandate to: 50% by December 31, 2026 and 60% by December 31, 2030.5 Figure ES-5 summarizes Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-11 the GHG impact analysis results for the IOU renewable scenario and three CCA renewable scenarios. Figure ES-5 GHG Emissions Impact Analysis, AWG Jurisdictions Participation Scenarios Large hydroelectric generation resources owned and managed by the IOUs do not count towards RPS goals and were also not significantly utilized during the recent drought years through 2016. Rainfall in the winter of 2016-2017 filled the hydroelectric reservoirs enabling a low cost, carbon-neutral generation component for the IOUs. In addition, the pumped hydro energy storage that can balance the variability of other sources of renewable generation also relies on rain to fill reservoirs. Future rainfall and drought conditions are unknown, and therefore the future utilization of large hydroelectric generation by the IOUs cannot be predicted. Additional use of hydro resources or increases to the IOU RPS content would result in lower GHG emissions for the IOUs, potentially decreasing the additional GHG reduction benefit of the CCA program. G. Cost of Service and Financial Pro Forma Analysis The cost of service analysis relied on traditional utility ratemaking principles and followed an industry standard methodology for creation of a financial pro forma to forecast the future economic and financial performance of the CCA program. The Study assessed financial feasibility in terms of the ability of the CCA program to realistically deliver competitive costs for customers while paying its substantial start-up Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-12 and agency formation costs and ongoing operating expenses. The first step in the cost of service analysis was developing the projected CCA program revenue requirement: the amount of revenue required to cover the costs of the CCA program, including all operating and non-operating expenses, debt-service payments, a contingency allotment, a working capital reserve, and a rate stabilization fund. The revenue requirement was based on a comprehensive accounting of all pertinent costs and projections of customer participation; assumptions and input development are described later in this report. Cost assumptions relied on historical publicly-available information, power cost forecasts conducted for this Study, data provided by PG&E and SCE, and subject matter expertise gained working with a host of public utilities and similar organizations. Table ES-4 summarizes the CCA program Test Year revenue requirements for the AWG Jurisdictions participation scenarios Table ES-4 Test Year CCA Revenue Requirements, AWG Jurisdictions Participation Scenarios CCA program customer participation was assumed to be constant for each participation scenario across the three renewable energy content scenarios examined. For all scenarios, an opt-out rate of 15% was used for all rate classes for all years, meaning that 15% of bundled customers by load in each rate class were assumed to opt out of the CCA program.6 This 15% opt-out rate is in addition to an estimated 23.5% of AWG Jurisdictions scenario load that represents typically large commercial customers who are RPS Equivalent Middle of the Road Aggressive REVENUE REQUIREMENT Baseload Total Operating Expenses Excluding Power Costs 10,146,683$ 10,256,373$ 10,482,215$ Total Non-Operating Expenses 16,959,517 18,158,147 20,239,969 Power Costs 461,419,035 489,933,855 549,930,521 Contingency/Rate Stabilization Fund 54,171,111$ 57,535,423$ 64,613,615$ BASELOAD REVENUE REQUIREMENT 542,696,345$ 575,883,798$ 645,266,320$ Opt-up to 100% RPS Total Operating Expenses Excluding Power Costs 207,075$ 209,314$ 213,923$ Total Non-Operating Expenses 346,113 370,574 413,061 Power Costs 12,617,576 12,617,576 12,617,576 Contingency/Rate Stabilization Fund 1,105,533$ 1,174,192$ 1,318,645$ OPT-UP TO 100% RPS REVENUE REQUIREMENT 14,276,297$ 14,371,657$ 14,563,205$ TOTAL REVENUE REQUIREMENT 556,972,642$ 590,255,454$ 659,829,525$ AWG Jurisdictions Participation Scenarios Description The Test Year is the future annualized period for which operating costs are analyzed and rate proxies established. The Study Test Year is based on forecasts of CCA operating conditions for years 2022, 2023, and 2024 and represents a twelve-month period of normalized operations selected to evaluate the cost of service for each customer class and the adequacy of rate proxies to provide sufficient revenue. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-13 likely to remain with their existing Direct Access (DA) ESP. Other CCA feasibility studies have supported the assertion that opt-out rates, within a reasonable range, have little bearing on CCA feasibility. Figure ES-6 and Figure ES-7 summarize Test Year customer accounts by rate class and Test Year customer usage by rate class for the AWG Jurisdictions participation scenarios, respectively. Average CCA Test Year customer profiles for the three AWG Jurisdictions participation scenarios are provided in Table ES-5. Figure ES-6 Test Year CCA Customer Accounts, AWG Jurisdictions Participation Scenarios 0 50 100 150 200 250 300 350 400 PG&E Customers, AWG Jurisdictions Scenarios SCE Customers, AWG Jurisdictions Scenarios Total CCA Customers, AWG Jurisdictions Scenarios Thousands Test Year Customer Accounts by Rate Class Traffic Control Residential CARE Residential Lighting Small Comm <200kW Med Comm 200<500kW Large Comm 500<1,000kW Very Large Comm >1,000kW Agriculture Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-14 Figure ES-7 Test Year CCA Customer Usage, AWG Jurisdictions Participation Scenarios 0 1,000 2,000 3,000 4,000 5,000 6,000 PG&E Customers, AWG Jurisdictions Scenarios SCE Customers, AWG Jurisdictions Scenarios Total CCA Customers, AWG Jurisdictions Scenarios GWH Test Year Customer Energy Usage by Rate Class Traffic Control Residential CARE Residential Lighting Small Comm <200kW Med Comm 200<500kW Large Comm 500<1,000kW Very Large Comm >1,000kW Agriculture Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-15 Table ES-5 Test Year CCA Customer Accounts and Usage, AWG Jurisdictions Participation Scenarios While rate design was not part of the Study scope, based on the detailed pro forma analysis, CCA rate proxies by customer class by IOU jurisdiction were developed. Rate proxies represent the amount of revenue by customer class required to make the CCA financially solvent, based on the Test Year. Based on this analysis, CCA baseline customers would have all-in rate proxies that are higher than both PG&E and SCE for most rate classes for all participation and renewable energy content scenarios examined. Table ES-6 through Table ES-8 present the generation rate differences between the CCA and PG&E and SCE for the AWG Jurisdictions participation scenarios for the RPS Equivalent, Middle of the Road, and Aggressive renewable energy content scenarios. The generation portion of customers’ bills is the only cost component for which the CCA competes with the incumbent utilities. Customer billing and delivery charges (transmission and distribution) are the same for both CCA and IOU bundled customers. Generation rate comparisons are provided for the first five years of the Study period by rate class.7 The Accounts Annual Load Average Monthly Load Line Description (MWh)(kWh/Account) 1 BASELOAD 2 Agriculture 6,454 490,772 6,337 3 Very Large Comm >1,000kW 13 718,495 4,673,350 4 Large Comm 500<1,000kW 405 441,022 90,742 5 Med Comm 200<500kW 576 297,829 43,094 6 Small Comm <200kW 40,034 1,124,051 2,340 7 Lighting 1,757 26,357 1,250 8 Residential 256,812 1,709,325 555 9 Residential CARE 22,929 124,036 451 10 Traffic Control 841 2,811 278 11 TOTAL BASELOAD 329,821 4,934,699 1,247 12 OPT-UP TO 100% RPS (MWH) 13 Agriculture - - - 14 Very Large Comm >1,000kW - - - 15 Large Comm 500<1,000kW 9 10,071 90,742 16 Med Comm 200<500kW 29 15,106 43,094 17 Small Comm <200kW 538 15,106 2,340 18 Lighting - - - 19 Residential 9,078 60,425 555 20 Residential CARE - - - 21 Traffic Control - - - 22 TOTAL OPT-UP TO 100% RPS 9,655 100,708 869 23 TOTAL CCA 339,476 5,035,407 1,236 CUSTOMERS OPTING UP TO 100% RENEWABLES Portion of Opt Up Portion of Total CCA 24 Agriculture 0%0.00% 25 Very Large Comm >1,000kW 0%0.00% 26 Large Comm 500<1,000kW 10%0.20% 27 Med Comm 200<500kW 15%0.30% 28 Small Comm <200kW 15%0.30% 29 Lighting 0%0.00% 30 Residential 60%1.20% 31 Residential CARE 0%0.00% 32 Traffic Control 0%0.00% 33 TOTAL 100%2.00% Test Year Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-16 total anticipated bill impact to residential customers in 2020 is included in Table ES 9. Table ES-6 Generation Rate Comparisons for PG&E, SCE, and CCA, AWG Jurisdictions RPS Equivalent Renewable Energy Content Scenario CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates Agriculture 0.1175 0.0742 0.1175 0.0753 0.1175 0.0749 0.1175 0.0747 0.1175 0.0754 Commercial/Industrial Small <200kW 0.1183 0.1049 0.1183 0.1065 0.1183 0.1059 0.1183 0.1055 0.1183 0.1065 Commercial/Industrial Medium 200<500 kW 0.1190 0.1097 0.1190 0.1113 0.1190 0.1107 0.1190 0.1103 0.1190 0.1114 Commercial/Industrial Large 500<1000 kW 0.1145 0.1107 0.1145 0.1124 0.1145 0.1118 0.1145 0.1114 0.1145 0.1124 Residential 0.1220 0.1003 0.1220 0.1018 0.1220 0.1013 0.1220 0.1009 0.1220 0.1018 Residential CARE 0.1152 0.0936 0.1152 0.0950 0.1152 0.0945 0.1152 0.0941 0.1152 0.0950 Residential Solar Choice 0.1920 0.1265 0.1920 0.1284 0.1920 0.1277 0.1920 0.1272 0.1920 0.1284 Weighted Average 0.1193 0.0961 0.1193 0.0975 0.1193 0.0970 0.1193 0.0967 0.1193 0.0976 CCA Rate Premium/ (CCA Savings)24.10%22.27%22.92%23.37%22.22% Rate Class CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates Agriculture 0.1050 0.0543 0.1050 0.0551 0.1050 0.0548 0.1050 0.0547 0.1050 0.0552 Commercial/Industrial Small <200kW 0.1072 0.0922 0.1072 0.0936 0.1072 0.0931 0.1072 0.0927 0.1072 0.0936 Commercial/Industrial Medium 200<500 kW 0.1064 0.0837 0.1064 0.0850 0.1064 0.0845 0.1064 0.0842 0.1064 0.0850 Commercial/Industrial Large 500<1000 kW 0.1057 0.0777 0.1057 0.0789 0.1057 0.0785 0.1057 0.0782 0.1057 0.0789 Residential 0.0999 0.0712 0.0999 0.0723 0.0999 0.0719 0.0999 0.0716 0.0999 0.0723 Residential CARE 0.0924 0.0635 0.0924 0.0645 0.0924 0.0641 0.0924 0.0639 0.0924 0.0645 Residential Green Tariff 0.1199 0.1127 0.1199 0.1144 0.1199 0.1138 0.1199 0.1134 0.1199 0.1144 Weighted Average 0.1034 0.0776 0.1034 0.0788 0.1034 0.0784 0.1034 0.0781 0.1034 0.0788 CCA Rate Premium/ (CCA Savings)33.23%31.26%31.97%32.44%31.21% 2026 Rate Class 2022 2023 2024 2025 Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-17 Table ES-7 Generation Rate Comparisons for PG&E, SCE, and CCA, AWG Jurisdictions Middle of the Road Renewable Energy Content Scenario Table ES-8 Generation Rate Comparisons for PG&E, SCE, and CCA, AWG Jurisdictions Aggressive Renewable Energy Content Scenario CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates Agriculture 0.1242 0.0742 0.1242 0.0753 0.1242 0.0749 0.1242 0.0747 0.1242 0.0754 Commercial/Industrial Small <200kW 0.1250 0.1049 0.1250 0.1065 0.1250 0.1059 0.1250 0.1055 0.1250 0.1065 Commercial/Industrial Medium 200<500 kW 0.1257 0.1097 0.1257 0.1113 0.1257 0.1107 0.1257 0.1103 0.1257 0.1114 Commercial/Industrial Large 500<1000 kW 0.1212 0.1107 0.1212 0.1124 0.1212 0.1118 0.1212 0.1114 0.1212 0.1124 Residential 0.1287 0.1003 0.1287 0.1018 0.1287 0.1013 0.1287 0.1009 0.1287 0.1018 Residential CARE 0.1219 0.0936 0.1219 0.0950 0.1219 0.0945 0.1219 0.0941 0.1219 0.0950 Residential Solar Choice 0.1987 0.1265 0.1987 0.1284 0.1987 0.1277 0.1987 0.1272 0.1987 0.1284 Weighted Average 0.1260 0.0961 0.1260 0.0975 0.1260 0.0970 0.1260 0.0967 0.1260 0.0976 CCA Rate Premium/ (CCA Savings)31.06%29.13%29.82%30.29%29.08% Rate Class CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates Agriculture 0.1117 0.0543 0.1117 0.0551 0.1117 0.0548 0.1117 0.0547 0.1117 0.0552 Commercial/Industrial Small <200kW 0.1139 0.0922 0.1139 0.0936 0.1139 0.0931 0.1139 0.0927 0.1139 0.0936 Commercial/Industrial Medium 200<500 kW 0.1132 0.0837 0.1132 0.0850 0.1132 0.0845 0.1132 0.0842 0.1132 0.0850 Commercial/Industrial Large 500<1000 kW 0.1124 0.0777 0.1124 0.0789 0.1124 0.0785 0.1124 0.0782 0.1124 0.0789 Residential 0.1066 0.0712 0.1066 0.0723 0.1066 0.0719 0.1066 0.0716 0.1066 0.0723 Residential CARE 0.0991 0.0635 0.0991 0.0645 0.0991 0.0641 0.0991 0.0639 0.0991 0.0645 Residential Green Tariff 0.1266 0.1127 0.1266 0.1144 0.1266 0.1138 0.1266 0.1134 0.1266 0.1144 Weighted Average 0.1102 0.0776 0.1102 0.0788 0.1102 0.0784 0.1102 0.0781 0.1102 0.0788 CCA Rate Premium/ (CCA Savings)41.87%39.78%40.53%41.04%39.72% 2026 Rate Class 2022 2023 2024 2025 CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates Agriculture 0.1382 0.0742 0.1382 0.0753 0.1382 0.0749 0.1382 0.0747 0.1382 0.0754 Commercial/Industrial Small <200kW 0.1390 0.1049 0.1390 0.1065 0.1390 0.1059 0.1390 0.1055 0.1390 0.1065 Commercial/Industrial Medium 200<500 kW 0.1397 0.1097 0.1397 0.1113 0.1397 0.1107 0.1397 0.1103 0.1397 0.1114 Commercial/Industrial Large 500<1000 kW 0.1352 0.1107 0.1352 0.1124 0.1352 0.1118 0.1352 0.1114 0.1352 0.1124 Residential 0.1426 0.1003 0.1426 0.1018 0.1426 0.1013 0.1426 0.1009 0.1426 0.1018 Residential CARE 0.1359 0.0936 0.1359 0.0950 0.1359 0.0945 0.1359 0.0941 0.1359 0.0950 Residential Solar Choice 0.2026 0.1265 0.2026 0.1284 0.2026 0.1277 0.2026 0.1272 0.2026 0.1284 Weighted Average 0.1399 0.0961 0.1399 0.0975 0.1399 0.0970 0.1399 0.0967 0.1399 0.0976 CCA Rate Premium/ (CCA Savings)45.56%43.41%44.18%44.70%43.35% Rate Class CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates Agriculture 0.1258 0.0543 0.1258 0.0551 0.1258 0.0548 0.1258 0.0547 0.1258 0.0552 Commercial/Industrial Small <200kW 0.1280 0.0922 0.1280 0.0936 0.1280 0.0931 0.1280 0.0927 0.1280 0.0936 Commercial/Industrial Medium 200<500 kW 0.1272 0.0837 0.1272 0.0850 0.1272 0.0845 0.1272 0.0842 0.1272 0.0850 Commercial/Industrial Large 500<1000 kW 0.1265 0.0777 0.1265 0.0789 0.1265 0.0785 0.1265 0.0782 0.1265 0.0789 Residential 0.1208 0.0712 0.1208 0.0723 0.1208 0.0719 0.1208 0.0716 0.1208 0.0723 Residential CARE 0.1132 0.0635 0.1132 0.0645 0.1132 0.0641 0.1132 0.0639 0.1132 0.0645 Residential Green Tariff 0.1308 0.1127 0.1308 0.1144 0.1308 0.1138 0.1308 0.1134 0.1308 0.1144 Weighted Average 0.1242 0.0776 0.1242 0.0788 0.1242 0.0784 0.1242 0.0781 0.1242 0.0788 CCA Rate Premium/ (CCA Savings)59.94%57.58%58.43%59.00%57.52% 2026 Rate Class 2022 2023 2024 2025 Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-18 Figure ES-8 and Figure ES-9 graphically depict the difference in generation rates between the CCA and PG&E and the CCA and SCE, respectively, for the AWG Jurisdictions scenario for the three renewable content scenarios. Figure ES-8 CCA and PG&E Generation Rate Comparison Summary for AWG Jurisdictions Participation Scenarios 0.0000 0.0500 0.1000 0.1500 0.2000 CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates AWG RPS EquivalentAWG Middle of The RoadAWG Aggressive2022-2026 Average Base Generation Rates ($/kWh) Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-19 Figure ES-9 CCA and SCE Generation Rate Comparison Summary for AWG Jurisdictions Participation Scenarios Table ES-9 shows the percentage change in average generation rates and the monetary change in monthly Residential bills for CCA customers versus PG&E and SCE, and the percent change in GHG emissions for all rate classes. This data is presented for year 2020. The previous Tables ES-6 through ES-8 present weighted average rate impacts across all seven customer classes examined for years 2022-2026. 0.0000 0.0500 0.1000 0.1500 0.2000 CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates AWG RPS EquivalentAWG Middle of The RoadAWG Aggressive2022-2026 Average Base Generation Rates ($/kWh) Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-20 Table ES-9 Summary of Forecasted Residential Class Outcomes by Renewable Energy Content Scenario, AWG Jurisdictions Participation Scenarios, Year 2020 Participation Scenario Included Jurisdictions Renewable Energy Content Pacific Gas & Electric Southern California Edison Proportional GHG Comparison Generation Rate Comparison (% Increase/ Decrease for CCA Customers) Monthly Bill Comparison ($ Increase/ Decrease for CCA Customers) Generation Rate Comparison (% Increase/ Decrease for CCA Customers) Monthly Bill Comparison ($ Increase/ Decrease for CCA Customers) All Tri-County Region All San Luis Obispo County All Santa Barbara County All Ventura County RPS Equivalent 22% $11.25 41% $14.55 6% 50% 29% $14.62 51% $17.93 -9% 75% 43% $21.72 71% $25.05 -55% Advisory Working Group Jurisdictions San Luis Obispo County Santa Barbara County Carpinteria Santa Barbara Ventura County Camarillo Moorpark Ojai Simi Valley Thousand Oaks Ventura RPS Equivalent 22% $12.21 41% $16.08 6% 50% 29% $15.92 50% $19.79 -9% 75% 43% $23.68 70% $27.64 -55% All San Luis Obispo County Arroyo Grande Atascadero Grover Beach Morro Bay Paso Robles Pismo Beach San Luis Obispo Unincorporated SLO County RPS Equivalent 29% $12.07 7% 50% 36% $14.89 -9% 75% 51% $20.77 -54% Unincorporated San Luis Obispo County Unincorporated SLO County RPS Equivalent 35% $15.70 7% 50% 42% $18.77 -9% 75% 56% $25.21 -54% All Santa Barbara County Buellton Carpinteria Goleta Guadalupe Santa Barbara Santa Maria Solvang Unincorporated Santa Barbara County RPS Equivalent 24% $11.15 45% $14.53 7% 50% 31% $14.27 55% $17.69 -9% 75% 45% $20.78 75% $24.22 -55% Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-21 Participation Scenario Included Jurisdictions Renewable Energy Content Pacific Gas & Electric Southern California Edison Proportional GHG Comparison Generation Rate Comparison (% Increase/ Decrease for CCA Customers) Monthly Bill Comparison ($ Increase/ Decrease for CCA Customers) Generation Rate Comparison (% Increase/ Decrease for CCA Customers) Monthly Bill Comparison ($ Increase/ Decrease for CCA Customers) Unincorporated Santa Barbara County Unincorporated Santa Barbara County RPS Equivalent 26% $15.08 47% $19.29 7% 50% 33% $18.97 56% $23.23 -9% 75% 47% $27.11 76% $31.44 -54% All Ventura County Camarillo Fillmore Moorpark Ojai Oxnard Port Hueneme Santa Paula Simi Valley Thousand Oaks Ventura Unincorporated Ventura County RPS Equivalent 41% $15.87 6% 50% 50% $19.54 -10% 75% 70% $27.35 -55% City of Santa Barbara Santa Barbara RPS Equivalent 69% $17.91 6% 50% 78% $20.42 -10% 75% 100% $25.98 -55% Table ES-10 shows annual operating results for the AWG Jurisdictions participation scenario for the RPS Equivalent renewable energy content scenario. Net operating margins are negative for all years of the Study period; meaning revenues are not sufficient to cover total operating and non-operating expenses plus the contingency and rate stabilization fund. In the initial years of the study period, this is due to the phasing in of customers and a lag in revenues versus expenditures. In later years, this revenue insufficiency is caused by rates remaining unchanged even though the CCA experiences an increase in operating costs. Rates were not increased because the CCA rate proxies were not competitive with IOU rates from the onset of the Study through 2026. Raising rates would make them less competitive. Although working capital initially is adequate, given the current debt assumptions that include a long-term bond financing in year 2020 of $288 million, starting in year 2024, working capital declines below targeted amounts and continues to decrease. The combination of increasingly negative net margins and a shortage of working capital would indicate the need for a rate increase around year 2026, again which would further harm the CCA program’s rate competitiveness relative to the IOUs. Table ES-11 presents this data for the AWG Jurisdictions Middle of the Road renewable energy content scenario and Table ES-12 presents this data for the AWG Jurisdictions Aggressive renewable energy content scenario. Generally speaking, results for these alternate renewable energy content scenarios are similar to the RPS Equivalent scenario, although Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-22 net margins and working capital deficiencies are better due to the higher rate proxies, which are set at the beginning and remain constant throughout the study period. Rate increases would still be required, but around the 2028 timeframe. Table ES-10 CCA Annual Operating Results, AWG Jurisdictions RPS Equivalent Scenario Table ES-11 CCA Annual Operating Results, AWG Jurisdictions Middle of the Road Scenario Year Operating Revenues ($000s) Total Operating Expenses Plus Contingency/ Rate Stabilization Fund ($000s) Non-Operating Revenues/ (Expenses) ($000s) Debt Service ($000s) Net Margin1 ($000s) Working Capital Fund ($000s) Working Capital Target ($000s) Working Capital Surplus/ (Deficiency) ($000s) Working Capital Surplus/ (Deficiency) (%) a b c d a - b + c - d e f e - f (e/f)-1 2020 110,694 139,109 1,145 11,515 (38,785) 211,653 47,077 164,575 350% 2021 445,293 469,267 2,227 11,515 (33,262) 189,905 159,570 30,335 19% 2022 545,838 533,627 2,046 17,276 (3,018) 186,887 181,993 4,894 3% 2023 556,361 541,735 2,028 17,276 (621) 186,266 184,808 1,458 1% 2024 556,922 543,639 1,925 17,276 (2,067) 184,199 185,916 (1,716) -1% 2025 555,121 543,720 1,985 17,276 (3,889) 180,310 186,453 (6,143) -3% 2026 554,190 551,493 1,903 17,276 (12,676) 167,634 189,470 (21,836) -12% 2027 553,316 556,757 1,721 17,276 (18,995) 148,639 191,885 (43,246) -23% 2028 553,165 566,687 1,396 17,276 (29,401) 119,238 195,934 (76,697) -39% 2029 550,808 569,985 1,183 17,276 (35,270) 83,967 198,148 (114,181) -58% 2030 548,923 581,521 386 17,276 (49,488) 34,479 203,224 (168,745) -83% NPV of Net Margin:(176,175) 1 Net Margin includes Net Operating Income less Debt Service. The net present value (NPV) of the Net Margin is determined using a 4% discount rate and is as of Year 2020. The discount rate is equal to the interest rate on the long-term debt. Year Operating Revenues ($000s) Total Operating Expenses Plus Contingency/ Rate Stabilization Fund ($000s) Non-Operating Revenues/ (Expenses) ($000s) Debt Service ($000s) Net Margin1 ($000s) Working Capital Fund ($000s) Working Capital Target ($000s) Working Capital Surplus/ (Deficiency) ($000s) Working Capital Surplus/ (Deficiency) (%) a b c d a - b + c - d e f e - f (e/f)-1 2020 117,525 150,875 1,235 12,330 (44,445) 223,724 50,583 173,141 342% 2021 472,491 504,655 2,323 12,330 (42,170) 193,883 170,117 23,766 14% 2022 579,072 568,848 2,082 18,499 (6,192) 187,691 192,494 (4,803) -2% 2023 590,222 575,366 2,044 18,499 (1,600) 186,092 194,836 (8,745) -4% 2024 590,817 570,966 1,962 18,499 3,314 189,406 194,067 (4,662) -2% 2025 588,906 566,609 2,098 18,499 5,896 195,302 193,284 2,019 1% 2026 587,918 570,586 2,132 18,499 966 196,268 195,171 1,096 1% 2027 586,991 571,282 2,109 18,499 (681) 195,587 196,227 (640) 0% 2028 586,831 576,506 1,991 18,499 (6,182) 189,405 198,875 (9,470) -5% 2029 584,330 574,978 2,033 18,499 (7,113) 182,292 199,652 (17,361) -9% 2030 582,330 581,643 1,541 18,499 (16,270) 166,022 203,279 (37,257) -18% NPV of Net Margin:(100,693) 1 Net Margin includes Net Operating Income less Debt Service. The net present value (NPV) of the Net Margin is determined using a 4% discount rate and is as of Year 2020. The discount rate is equal to the interest rate on the long-term debt. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-23 Table ES-12 CCA Annual Operating Results, AWG Jurisdictions Aggressive Scenario H. Feasibility Outcome Summary The two primary factors driving forecasted feasibility results for the CCA include: 1) the competitiveness of CCA rates against PG&E and SCE rates; and 2) the long-term financial viability of the enterprise. Under all participation scenarios, because the rate comparisons show most rate classes paying more for power supplied by the CCA than from the incumbent utilities and because the CCA does not maintain sufficient revenues and working capital throughout the Study period, the CCA is deemed infeasible Regarding rate competitiveness, forecasted CCA revenue requirements are primarily driven by power procurement costs and the Cost Responsibility Surcharge (CRS), which consists of the Competitive Transition Charge (CTC), the Department of Water Resources Bond Charge (DWR-BC), and the Power Cost Indifference Adjustment (PCIA). Together, these two components represent 78% of the total of the overall projected CCA revenue requirement and are thus primary drivers of rate competitiveness against the two incumbent utilities. Recent historical movements in the CRS and the allocation of incumbent utility revenue requirements between generation and delivery (i.e., transmission and distribution) appear to disadvantage the CCA program. The delivery portion of customers’ bills is paid equally by CCA and bundled IOU customers. Generally speaking, in recent years the incumbent utilities appear to have been shifting costs from generation to delivery, as discussed in more detail in Section II.E.1 Feasibility Drivers. The CCA only competes against the incumbent utilities on generation. Given the assumptions of this Study, SCE and PG&E forecasted generation rates are not high enough to support CCA feasibility at the forecasted level of CCA power procurement and operational costs. Regarding long-term financial viability, the CCA would Year Operating Revenues ($000s) Total Operating Expenses Plus Contingency/ Rate Stabilization Fund ($000s) Non-Operating Revenues/ (Expenses) ($000s) Debt Service ($000s) Net Margin1 ($000s) Working Capital Fund ($000s) Working Capital Target ($000s) Working Capital Surplus/ (Deficiency) ($000s) Working Capital Surplus/ (Deficiency) (%) a b c d a - b + c - d e f e - f (e/f)-1 2020 131,724 168,193 1,428 13,746 (48,788) 250,176 55,745 194,431 349% 2021 528,600 562,520 2,607 13,746 (45,059) 218,863 187,370 31,493 17% 2022 647,505 633,619 2,361 20,623 (4,375) 214,487 211,809 2,679 1% 2023 659,933 646,015 2,318 20,623 (4,388) 210,100 215,901 (5,801) -3% 2024 660,598 637,896 2,227 20,623 4,307 214,407 214,025 381 0% 2025 658,462 633,821 2,370 20,623 6,388 220,795 213,325 7,469 4% 2026 657,357 640,581 2,395 20,623 (1,452) 219,343 216,041 3,302 2% 2027 656,320 642,137 2,343 20,623 (4,096) 215,247 217,353 (2,106) -1% 2028 656,142 648,050 2,187 20,623 (10,344) 204,903 220,206 (15,303) -7% 2029 653,345 646,843 2,185 20,623 (11,936) 192,967 221,079 (28,111) -13% 2030 651,109 652,739 1,647 20,623 (20,605) 172,362 224,476 (52,114) -23% NPV of Net Margin:(120,434) 1 Net Margin includes Net Operating Income less Debt Service. The net present value (NPV) of the Net Margin is determined using a 4% discount rate and is as of Year 2020. The discount rate is equal to the interest rate on the long-term debt. In no participation or renewable energy content scenario were the CCA program’s rates competitive with PG&E or SCE. Given the underperformance of the CCA in terms of being rate competitive, consistently having negative net margins, and failing to meet the target for working capital, the CCA under the assumptions used in the Study is neither reliably solvent nor financially feasible. Executive Summary Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 ES-24 need additional rate increases around the year 2026 timeframe to maintain adequate working capital and increase net margins, further decreasing rate competitiveness. I. Sensitivity Analysis Results Upon completion of the Study outcomes for each participation and renewable energy content scenario, additional sensitivity cases were examined against the AWG Jurisdictions participation scenario to determine how changes in key inputs affect feasibility outcomes. These sensitivities included: (1) Decreases in power procurement costs; (2) Increases in IOU rate escalation; and (3) Decreases in staffing costs. Each sensitivity was examined individually to determine the point at which the CCA could be feasible. As discussed in more detail in Section II.E.2, Pro Forma Sensitivity Analysis, in order for the CCA to be feasible: • Power procurement costs would have to decrease 40% over the Study forecast, or • PG&E and SCE rates would have to escalate at an additional 4.0% per year above the Study forecast. A staffing cost reduction alone is not expected to affect program feasibility. Although not examined as part of this Study, some combination of changes to the Study assumptions could result in a more feasible outcome. Like all feasibility studies, assumptions used herein are based on a forecast of future conditions which may or may not occur. Various market and regulatory drivers may change resulting in different outcomes from those assumed herein. The assumptions used in the Study are reasonable for the purposes of analyzing the feasibility of CCA within the Tri-County Region, but no warranties as to the accuracy of outcomes are implied or should be inferred. FOR THE CENTRAL COAST REGION APPENDIX L: PEER REVIEW AND RESPONSE AUGUST 2017 TECHNICALFEASIBILITYSTUDY ON COMMUNITY CHOICE AGGREGATION             This page intentionally left blank.  APPENDIX L PEER REVIEW AND RESPONSE             This page intentionally left blank.  Table of Contents APPENDIX L: PEER REVIEW AND RESPONSE 1. MRW and Associates Peer Review .................................................................................................... L-1 2. MRW Extended Peer Review ............................................................................................................. L-23 3. Response to Peer Review ................................................................................................................. L-39 Exhibit A: Power Procurement Cost Comparison Results ........................................................................ L-64 Exhibit B: Decrease In Staffing Costs Comparison Results ........................................................................ L-69 Exhibit C: Annual Escalation of PG&E and SCE Rates Comparison Results .......................................... L-72 Exhibit D: Power Procurement Monte Carlo Simulation Model Questions ......................................... L-74 4. Response to Extended Peer Review ................................................................................................ L-83             This page intentionally left blank.  Appendix L: Peer Review and Response This Appendix provides the initial and extended peer reviews conducted by MRW and Associates, LLC of the Technical Feasibility Study on CCA for the Central Coast Region and the response of Willdan Financial Services and EnerNex to the initial peer review. 1. MRW and Associates Peer Review             This page intentionally left blank.  MEMORANDUM To: Jennifer Cregar, Project Supervisor, Energy and Sustainability Initiatives, County of Santa Barbara From: Mark Fulmer, David Howarth, Jeremy Waen, and Anna Casas Llopart Subject: Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report dated May, 2017 Date: May 31, 2017 In late 2015, the County of Santa Barbara Board of Supervisors authorized funds to perform a Draft Study and directed staff to explore regional interest in Community Choice Aggregation (CCA). Ten local governments joined with the County of Santa Barbara to fund the Draft Study, and the following jurisdictions formed an Advisory Working Group (AWG) in December 2015. The CCA Feasibility Study was requested to provide an in-depth technical, economic, and financial analyses of the potential costs, benefits, and risks of CCA for the Tri-county region (Santa Barbara, Ventura, and San Luis Obispo counties) under a variety of future outcomes, or scenarios. The Draft Study is intended to provide policy makers, stakeholders, and electricity consumers information for assessing the feasibility of a CCA program for the Tri-County region. On May 14, 2017, the County provided MRW & Associates, LLC (MRW) a draft report entitled “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report dated May 10, 2017 (the Draft Study), and requested MRW to provide a professional peer review of the Draft Study. This memorandum provides MRW’s review. Beyond the Summary of Conclusions, it is organized around the 10 questions concerning the Draft Study to which the County asked MRW to respond. Summary of Conclusions The Draft Study considered eight CCA composition scenarios, each with differing community memberships, ranging from the “All Tri-County Region” to the City of Santa Barbara alone (See Table ES- XIII). Like the Draft Study, MRW’s review effort concentrated on the AWG Jurisdictions scenario. Overall, the Draft Study is detailed and comprehensive. Its assessment of loads and load forecast are thorough and reasonable, and it provides an in-depth look into potential CCA operations. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-3 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 2 _____________________________________________________________________________________ MRW & Associates, LLC Unlike prior recent CCA technical studies, the Draft Study concluded that CCA was not economically feasible even when only the state-required minimum renewable energy content was assumed. MRW’s focused its review to identify areas where the Draft Study was potentially overly conservative or made questionable assumptions that might explain why its conclusion was negative while others have been affirmative. In this regard, MRW identified several areas where Willdan, the Draft Study’s author, should consider revising its assumptions: 1. CCA Renewable power contracts. The Draft Study’s use of utility-average renewable contract prices does not reflect the most recently-reported contract prices and does not reflect the general downward trend in renewable prices seen over the past few years. 2. “Uncollectible expenses.” The Study assumed from 5% to 8% of the revenues due to the CCA from its customers could not be collected. This is an order-of-magnitude higher than that experienced by either MCE Clean Energy (MCE),1 the longest-running CCA in the state, or Sonoma Clean Power (SCP), the second longest-running CCA in the state. CCAs do not observe the same level of uncollectible accounts as the IOUs due because CCAs are allowed to return non-paying accounts to the corresponding IOU’s bundled service. 3. Administrative labor costs. The number of employees assumed in the pro forma analyses, as well as their compensation, appear high relative to operating California CCAs. 4. CCA service fees. The incumbent utilities—Southern California Edison (SCE) and Pacific Gas and Electric (PG&E)—charge CCAs in their respective territories certain fees for billing conducted on behalf of the CCA as well as meter and data management. While the Draft Study reflects current tariffed rate for these services, it does not account for the proposed dramatic uncontested reductions being presented by both utilities. Similarly, it is unclear whether the ESP service fees section of the Draft Study properly accounts for critical operational services such as data management and scheduling coordination. 5. Assumed reserves funding. Beyond working capital, CCAs typically develop a “rate stabilization reserve fund” which can be drawn upon in years’ where the CCA might not otherwise be able to meet its rate targets. The Draft Study pro forma analysis appears to assume that approximately $78 million (14% of total expenses) is contributed each year, rather than setting a target (e.g., 15% of annual expenses), taking 3 to 5 years to achieve the fund, and then eliminate further contributions until replenishment is needed. 6. PG&E and SCE Rate Forecasts. A fundamental concern is that the forecast of SCE and PG&E rates is disconnected from the forecast of CCA rates. The utility rates against which the CCA rates are compared are simply the current rates escalated at 0-0.5%. It does not account for: (i) SCE’s or PG&E’s actual supply portfolio, (ii) the two utilities’ status with respect to State’s renewable power content mandates, (iii) fuel price trends, or (iv) any other underlying fundamentals. In particular, there is no explicit connection between the utilities’ generation 1 MCE began serving customers in May 2010 to select areas within Marin County. Presently serves approximately 255,000 accounts located within all of Marin and Napa Counties, as well as select cities within Contra Costa County (Richmond, San Pablo, El Cerrito, Lafayette, and Walnut Creek ) and the City of Benicia in Solano County. MCE serves a diverse customer base in terms of geographic, ethnic and socio-economic backgrounds. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-4 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 3 _____________________________________________________________________________________ MRW & Associates, LLC rates and the CCA generation cost, even though they would be purchasing from the same wholesale market and vying for the same incremental renewable generation sources. We are also concerned that the Draft Study assumes that the franchise fees (i.e., utility taxes) that would flow to the respective cities’ and counties’ general funds if SCE or PG&E were providing service is assumed to instead flow to the CCA. This treatment should be verified by the AWG or corrected. Lastly, we recommend that sensitivity cases used to explore the impact of lower SCE and PG&E rates and higher exit fees consider a wider range of potential values. Responses to Questions 1. Does the Study consider all pertinent factors to determine current and future electric energy requirements of the CCA? The Draft Study notes, “…historical utility level consumption data for 2001-2016 was pulled from EIA Form 861 for both PG&E and SCE. This data was analyzed and a logarithmic line of best fit was created and extended through 2030. This data was then compared with the California Energy Commission’s long-term procurement plan (LTPP)(sic) load forecasts, which are available through 2025 for the respective planning areas. Because the two sources showed very different results by 2030, the average between the LTPP sales projection and the EIA consumption data forecast was utilized for the load forecast for Central Coast Power.” The curve fit showed a much lower load growth rate than that from the CEC. Draft Study forecast shows modest load growth. That is, natural load growth from increased economic activity is generally offset by efficiency and behind-the-meter customer generation (e.g., rooftop solar). Particularly given the relatively short time frame in which it conducts the economic analysis, this load forecast is reasonable. Direct Access (DA): Since DA customers are not likely to join a CCA due to an existing contract with an Electric Service Provider (ESP), for purposes of this Draft Study DA customers have been excluded from the load forecast. Opt-out 15% base assumption. The Draft Study assumes that 15% of the eligible customers will opt-out of the CCA and remain on bundled utility service. This value is conservative relative to the actual opt- out rates experienced with the most recent CCAs. 2. Does the Study incorporate current power market conditions and reasonable projections of expected future conditions? The Draft Study provides a comprehensive review of current power market conditions, including a qualitative summary of power procurement considerations (e.g., renewable portfolio standard (RPS), resource adequacy and storage) as well as a quantitative analysis of recent historical pricing for Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-5 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 4 _____________________________________________________________________________________ MRW & Associates, LLC renewable energy, natural gas generation and California Independent System Operator (CAISO) day- ahead and real-time wholesale electricity markets. The Draft Study presents data on current expectations regarding the relative levelized cost of energy for different generation technologies and recent declines in solar photovoltaic (PV) costs. The Draft Study also presents data showing trends in utility RPS compliance costs, as reported annually to the California legislature (i.e., the Padilla report) and in the Biennial RPS reports. Renewable Energy Procurement. To forecast CCA renewable energy procurement costs, the Draft Study consultants developed a best-fit logarithmic curve using average utility RPS compliance costs depicted in Figure ES-40 of the Draft Study. The resulting RPS price forecast is likely a conservative estimate of CCA renewable energy procurement costs. This is because the data used to forecast RPS price trends do not necessarily reflect the market in which the CCA will operate since the data reflect utility procurement costs for energy delivered during a particular year. The renewable energy portfolios of utilities include contracts struck over a period of time during which technology costs have been rapidly decreasing. As a result, the decline in average costs incurred by the utilities for renewable energy deliveries has lagged behind the decline in costs for new (incremental) resources. This point is referred to in footnote 97 of the Draft Study, which quotes an explanation by California Public Utilities Commission (CPUC) staff. The 2016 Padilla report,2 issued May 1, 2017, presents time-of-delivery-adjusted renewable energy prices for bundled RPS contracts approved in 2016. The prices are aggregated to avoid revealing confidential data, and for SCE include wind, geothermal and biomass contracts in addition to solar. The weighted average prices for contracts approved in 2016 are $0.059/kWh for PG&E and $0.061/kWh for SCE, well below the average 2016 expenditures of $0.11/kWh and $0.094/kWh, respectively. The prices of contracts approved in 2016 are approximately 30% below the average RPS PPA cost of $88/MWh assumed in the Report for 2020. Since the CCA would be making RPS contract purchases at current and future market prices that are lower than the average utility RPS compliance cost as reflected in Figure ES-40, the Draft Study has likely overestimated RPS PPA costs in the pro forma analysis. The Monte Carlo model used for the Draft Study is useful for reflecting uncertainty in forecasts of procurement costs, by providing a statistically characterized range around this base forecast. The report does not provide information concerning the way in which RPS price uncertainty was characterized in the Monte Carlo model, so it is not possible to review the reasonableness of these assumptions. Natural Gas Generation. In the case of natural gas generation prices, the Draft Study fit a curve to 2002- 2016 CAISO market implied prices to forecast prices for the period through 2035. Based on this analysis, natural gas generation costs are forecast to decrease by 25% from $41/MWh in 2020 to $31/MWh in 2030. This trend analysis may be underestimating natural gas generation costs over the long term by not differentiating between trends in market heat rates (the implied rate of conversion of natural gas energy to electricity, in Btu/kWh) and natural gas prices, which may be driven by different market dynamics not captured by the trend analysis. Natural gas prices are relatively low at present. In its 2017 2 http://www.cpuc.ca.gov/uploadedFiles/CPUCWebsite/Content/About_Us/Organization/Divisions/Office_of_Gover nmental_Affairs/Legislation/2017/Final%20-%20Padilla%20Report%20-%20RPS%20Costs%202017.pdf Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-6 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 5 _____________________________________________________________________________________ MRW & Associates, LLC Annual Energy Outlook, the Energy Information Administration forecasts natural gas prices for electricity generation in the Pacific region to increase by an average of 3.5% per year between 2020 and 2030. Based on this forecast of natural gas prices, the forecast of natural gas generation costs used in the Draft Study suggests market heat rates will decrease by more than half between 2020 to 2030, or a compound average rate of -6.1%. While there may be downward pressure on market heat rates as additional renewable energy sources are brought on line, a 6% per year reduction in market heat rate is likely not sustainable since it would be difficult for natural gas generators to recover costs. The Draft Study would likely benefit from a review of this assumption and the associated discussion of the forecast. As with the RPS cost forecast, additional information on how natural gas price uncertainty was reflected in the Monte Carlo model would be needed to assess reasonableness. Other Cost Components. Following the cost of RPS procurement and natural gas generation, resource adequacy (RA) represents the remaining significant component of CCA procurement costs. The Draft Study provides a reasonable forecast of RA costs. The remaining components, including CAISO day- ahead and real-time markets and storage procurement represent a small fraction of total costs, just 2% in the 50% RPS case. The forecasts used in the Draft Study for these cost components appear reasonable. 3. Are the estimates of the GHG emissions intensity of the CCA scenarios relative to the incumbent investor-owned utilities (IOUs), namely Pacific Gas and Electric Company (PG&E) and Southern California Edison (SCE), reasonable and adequate? The Draft Study’s projections of CCA greenhouse gas emissions are generally reasonable. Figure 1 below replicates “Table ES-XL (sic) Jurisdictions scenario CO2 output comparison with IOU base case and trend” from the Draft Study. Note that the IOU Base Case line (orange) converges with the CCA 50% RPS line (green) by 2030. This reflects the fact that in 2030 the IOUs would be meeting the 50% RPS requirement in 2030, the same renewable content as the CCA. However, implicit in this figure is that the CCA also can procure non-RPS compliant carbon-free power (i.e., large hydroelectric) in an equal share to that which SCE and PG&E have. This is particularly important with respect to PG&E, which has significant nuclear and large hydroelectric resources3. Note also that this figure assumes that PG&E meets its goal of replacing the output of the retiring Diablo Canyon Nuclear Power Plant (2022-2023) with carbon-free resources. The “IOU Trend” line in the figure (yellow) is interesting and provides a conservative benchmark against which the CCA’s GHG emissions can be compared. However, it should not be used to provide the basis for a GHG analysis. 3 Note that the power content labels included in the Draft Study for the two IOUs are for 2015, which due to the drought conditions understates the typical hydroelectric output and thus overstates th e IOUs’ GHG emissions. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-7 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 6 _____________________________________________________________________________________ MRW & Associates, LLC Figure 1: Greenhouse Gas Emissions Consistent with other CCA analyses conducted or peer reviewed by MRW, the Draft Study illustrates that if a CCA wishes to reduce GHG emissions relative to remaining with the incumbent utility while maintaining competitive rates, it would need to explicitly contract for non-RPS complying, GHG-free power: that generated by large hydroelectric or nuclear facilities. 4. Does the Draft Study consider all pertinent factors in projecting future PG&E and SCE rates for comparison to CCA costs/payment/rate projections? MRW finds there are areas where the Draft Study can be improved and refined with respect to the forecast of PG&E and SCE rates. Error in Current IOU Rates. Table 1 compares current PG&E rates as presented in both the Draft Study and PG&E’s 5011-E-A advice letter. While some rates are reasonably similar, others, particularly the medium and large commercial and industrial rates, are not. The difference between these rates is attributable to the study’s use of differing “billing determinants.”4 It appears the Draft Study assumes a 4 “Billing Determinants” are the usage values one multiplies times the rates to arrive at the total bill. For residential customers, it is just the number of kilowatt -hours consumed. For large accounts, this include the seasonal on - peak and off peak use (in kilowatt-hours) as well as the maximum demand (kilowatts) that occur during various periods throughout the day and year. - 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 1,600,000 1,800,000 2,000,000 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030MT CO2CCA RPS Equivalent IOU Base Case CCA 50% Renewable IOU Trend CCA 75% Renewable Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-8 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 7 _____________________________________________________________________________________ MRW & Associates, LLC 17% load factor for Commercial/Industrial Large rate class; instead the average load factor for this rate class should be in the range of 45%-65%.5 This should be corrected. Table 1: Comparison of Draft Study’s estimated PG&E rates to PG&E’s actual rates PG&E (¢/kWh) Rate Class Schedule Draft Study Advice letter 5011-E-A Agriculture AG-5B 14.0 16.6 Very Large Commercial >1,000kW E-20-T 10.9 11.7 Commercial/Industrial Large 500<1000 kW E-19SV 33.5 17.8 Commercial/Industrial Medium 200<500 kW A-10S 24.3 20.4 Commercial/Industrial Small <200kW A-1 22.2 23.0 Residential E-1 23.1 23.1 Residential CARE EL-1 13.6 13.7 Table 2 below provides a similar comparison for SCE rates presented in the Draft Study relative to MRW’s estimated average rates. As was the case with Table 1, the rate differences occurring in Table 2 are due to differences in how the billing determinants are calculated. For example, for Commercial/Industrial Small, the Draft Study assumes a 11% load factor; instead the average load factor for this rate class should be in the range of 35-55%. Table 2: Comparison of Draft Study’s estimated SCE rates to MRW’s estimates of SCE rates SCE (¢/kWh) Rate Class Schedule Draft Study MRW estimates Agriculture TOU-PA-3 12.5 12.7* Very Large Commercial >1,000kW TOU-8 -T Option B 8.5 9.1 Commercial/Industrial Large 500<1000 kW TOU-8 -P Option B 28.2 12.8 Commercial/Industrial Medium 200<500 kW GS3-RTIME 17.5 14.5** Commercial/Industrial Small <200kW GS2-RTIME 31.3 16.9*** Residential D 19.6 19.4 Residential CARE D-CARE 12.1 12.1 * Average rate for agriculture rate class ** Rate for GS3-TOU-Option B *** Rate for GS2 –Option B 5 “Load Factor” reflects how much the customer uses relative to its peak demand. A customer who uses power at its peak demand level all time would have a “load factor” of 100%. Because large customer rates have per kW demand charges, the higher the load factor, the more kilowatt -hours the demand charges are averaged over and thus the lower the rate. Thus, there is a large difference between the average rate of a customer with a low load factor, like 17%, and a higher one, such as 65% or higher. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-9 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 8 _____________________________________________________________________________________ MRW & Associates, LLC a. IOU Rates Forecasts Figure 2 compares the PG&E generation rate forecast done by MRW for the CCA Technical Study for Contra Costa County6 (Contra Costa Study) and the Draft Study. In both cases the current generation rate, 2017, is based on the weighted average of Central Coast CCA Scenario 2 PG&E rate using the class averages generation rates from AL 5011-E-A. The Draft Study forecasted 0% annual increase between 2017 and 2020, and -0.25% between 2020 and 2030. This is based on the Draft Study’s annual increase of the power costs calculated using the Monte Carlo simulation. Instead, the Contra Costa Study forecast was developed on a fundamentals basis, considering PG&E’s generation portfolio, contracts, power markets, etc., and resulted in an annual average increase of 3% from 2017 to 2030. More precisely, the Contra Costa Study forecasts a 1.5% annual increase between 2017 and 2022, followed by a 1.5% annual decrease between 2023 and 2025 (due to the Diablo Canyon retirement), and finally a 5% annual increase between 2026 and 2030. Figure 2: Comparison of Draft Study’s and MRW’s forecasts of PG&E generation rates Furthermore, the Draft Study extends its calculated escalator for generation rates to non-generation rates. This is concerning because there is no direct relation between the cost drivers for generation and non-generation utility services. 6 http://www.cccounty.us/DocumentCenter/View/43588 0 2 4 6 8 10 12 14 16 2017 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030¢/kWhContra Costa Study Draft Study Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-10 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 9 _____________________________________________________________________________________ MRW & Associates, LLC 5. Does the Draft Study consider all pertinent factors in presenting a reasonably accurate investor-owned utility (IOU) vs. CCA cost/payment comparison? Our concerns regarding escalation of the PG&E and SCE delivery rate raised in response to Question 4 would not be material if the same delivery rate is used for both the utility and CCA rates. However, it is not clear from the Draft Study report that a common delivery rate was used in the comparison of SCE and PG&E rates and CCA costs. As noted above, the utility rate forecasts were based on the escalation of both the generation and delivery rates. What would be helpful would be a comparison table that showed, either on a class basis or on a system average basis the following (in $/kWh): YEAR PG&E/SCE CCA a b c = a+b d = a e f g h = d+e+f+g i=(h-a)/a Delivery Rate Genera- tion rate Total Rate Delivery Rate Ave. Power Cost Other Costs PCIA Total Rate Pct. difference 2022 2023 2024 2025 2026 … 6. Do the pro forma analyses consider all pertinent factors in projecting CCA’s operating results? Yes. However, the Draft Study may be treating the franchise fee revenues incorrectly. Franchise fees are a percentage of utility customers’ bills that are paid to cities or counties for the nonexclusive right to install and maintain equipment on streets and public rights of way (e.g., power poles, underground power or gas lines). The Draft Study assumes that the franchise fees collected by PG&E and SCE from CCA customers will be diverted from the general fund into the CCA. MRW is not aware of other CCAs diverting the franchise fee revenue stream from the participant’s general fund to the CCA. The AWG should verify that this is an acceptable treatment before it is included as a CCA revenue source. If it is not, or is at all questionable, franchise fee revenue should be removed from the pro forma analysis. Second, it is not clear that the franchise fees are correct. The rate modeling shows particularly high SCE franchise fees as part of the CCA rates: around 9% of CCA revenue. Later, and in the pro forma, the franchise fees are subtracted out. Power Costs: As discussed above, there is a great deal of uncertainty in forecasts of power costs. The base forecast of RPS procurement costs is likely conservative, while the forecasted costs of natural gas generation may be lower than expected over the forecast period. To the extent that the pro forma analyses include Monte Carlo simulation model results, the pro forma results may reasonably reflect the Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-11 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 10 _____________________________________________________________________________________ MRW & Associates, LLC expected range of power costs. It is difficult to assess the reasonableness of the Monte Carlo simulation model analyses with information presented in the Report. Other Operating Costs. Operating costs consist of all costs directly associated with provision of the business services and activities of the CCA—namely procuring and providing power to customers. The Draft Study thoroughly presented the operating costs of a hypothetical CCA. Salaries & Wages: Both the 45 FTE staff proposal and the average fully loaded salary costs seem excessive for this proposed CCA. MCE has the largest staff of any CCA present and this is largely due to two factors 1) they were the first CCA to form so resource sharing with other CCAs was not an option until very recently, and 2) they are engaged in administering Energy Efficiency programs utilizing ratepayer funds. The former is important because subsequent CCAs are finding they can operate with much leaner staffing than MCE. The latter is important to consider because the EE programs utilize a separate revenue stream from electricity sales. Additionally, EE (and customer facing programs in general) commands a higher staffing requirement than other core operations within a CCA. Additionally, based on this Draft Study the average loaded proposed salary for the Central Coast Power CCA would be $156,743. Whereas based on MCE’s projected FY 2016/17 financials their average fully loaded salary is $116,983. As a result, both factors cause the “Salaries and Wages” expense category to be significantly larger than would be prudent for a new CCA organization. As such, we suggest that Willdan consider the following revisions: 1) Adjust the anticipated FTE downward (perhaps 20-30 FTE), especially at the upper end of the staffing spectrum. 2) Adjusting the proposed salary costs downward. IOU Service Charges: Based on analysis it appears the Draft Study uses a $0.83/MWh/month multiplier to determine both PG&E’s and SCE’s service charges. Furthermore, this multiplier has a 2% annual escalator applied. These assumptions seem problematic. First, PG&E and SCE have notably different Meter Data Management Agent (MDMA) and Bill-Ready fees. (Note that MDMA charges are on a per meter per month basis. Bill-Ready charges are on a per customer per month basis.) PG&E’s present MDMA fee is dramatically higher than SCEs, though PG&E is proposing in its present General Rate Case (GRC) Phase 2 to dramatically reduce this fee from $7.67 to $0.14. PG&E has differing Bill-Ready fees based upon whether the CCA’s charges appear on a separate page of the bill or not. In contrast SCE has differing Bill-Ready fees depending upon whether the bill is delivered via printed or electronic means. Furthermore, both PG&E and SCE have proposals before the CPUC to reduce these charges because they observe increasing numbers of departing load customers over which these sorts of costs can be spread. There is no reason to believe this trend won’t continue as more CCAs form. As a result, IOU Service Charges seem a bit overestimated. The PG&E and SCE CCA Start Up and Opt-Out charges that also roll-up into this total IOU Service Charges category seem reasonable and do not require revising. As such, we recommend that Willdan consider the following revisions: Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-12 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 11 _____________________________________________________________________________________ MRW & Associates, LLC 1) Use PG&E’s and SCE’s proposed revised MDMA and Bill-Ready fees that are likely to be effective well before 2020 to more accurately approximate the resulting PG&E and SCE Service Fees. 2) Either keep these fees level or approximate some small de-escalation factor to account for the likelihood that these fees will further reduce as more load departs and as these metering and billing departments of the utilities adapt to automate these processes. ESP Charges: It was difficult to understand and extrapolate the various types of ESP services and related charges that could be used to justify the $1.50/account/month multiplier used to determine these overall charges. MCE presently has contracted a $1.15/account/month fee for Data Management services with Calpine. Scheduling Coordination is a separate service that also fits under this “ESP Charges” category and would add to the costs as well. It appears that this $1.50/account/moth factor is in the correct ballpark to approximate these types of costs; however, it is difficult to say if this figure is too high or too low. As such, we recommend that Willdan consider looking to existing CCAs’ public contract information to better approximate Data Management and Scheduling Coordination costs under this category. Jurisdictional Administration: It is atypical for a CCA to reimburse the local jurisdictions for staff- time spent interfacing with the CCA. The one area where this might be practiced is with Single Jurisdiction (rather than Joint Powers Authority) CCAs where staff is shared between local government and CCA operations. Even in those cases this “Jurisdictional Administration” category seems to overlap with the Salary & Wages category. As a result, these costs should not be considered part of the CCA’s operating expenses. We therefore recommend that Willdan consider excluding these costs from the Operation Expenses analysis. Uncollectable Accounts: Per the draft report it appears that a 5% uncollectable accounts rate is assumed for PG&E accounts and an 8% uncollectable accounts rate is assumed for SCE accounts. Neither rate seems reasonable. First and foremost, CCA uncollectable account rates are not directly comparable to IOU uncollectable account rates. If a CCA customer account is repeatedly uncollected or under-collected it permitted practice to return that customer’s account to bundled utility service.7 As such, CCAs observe a significantly lower uncollectable accounts rate than IOUs. For example, MCE presently observes a 0.5% uncollectable accounts rate for its 255,000 customer accounts across its four-county service area.8 SCP also observes and plans for 7 PG&E and SCE Electric Rule 23 section Q.2 both state: “[PG&E/SCE] shall not disconnect electric service to the customer for the non-payment of CCA charges. In the event of non-payment of CCA charges by the customer, the CCA may submit a CCASR requesting transfer of the service account to [PG&E/SCE] Bundled Service according to Section M. 8 See MCE fiscal year 2015/16 audited financial statements: https://www.mcecleanenergy.org/wp- content/uploads/2016/09/MCE-Audited-Financial-Statements-2015-2016.pdf Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-13 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 12 _____________________________________________________________________________________ MRW & Associates, LLC a 0.5% uncollectable accounts rate.9 As a result, the “Uncollectable Accounts” operating expense category is significantly overestimated. As such, we recommend that Willdan consider adjusting the uncollectable accounts rate downward from 5% for PG&E accounts and 8% for SCE accounts to 0.5% for both PG&E and SCE accounts. PCIA: Included in operating costs is the Power Cost Indifference Amount (PCIA). The PCIA is the state-mandated fee that SCE and PG&E imposes on all departed load (including CCA customers) to ensure that the rates of utility customers who do not—or cannot—choose CCA service do not increase because of CCA. The Draft Study relies upon a forecast of the PCIA rate from the utilities’ green tariff forecasts. Because the PCIA is difficult to accurately forecast, this assumption is not unreasonable, but as noted later, must be thoroughly explored in sensitivity analyses. Non-Operating Costs. Non-operating costs include initial capital outlays for longer-living assets required to get the CCA up and running as well as the associated debt issuance and annual debt service required to fund the CCA. Non-Operating Costs also include a contingency/rate stabilization fund. The Draft Study thoroughly presented the non-operating costs of a hypothetical CCA. The Study also assumes an initial long-term bond issuance for working capital equal to 5 months cash flow plus the rate stabilization fund. MRW is concerned that the debt amount appears to be unnecessarily high. Prior CCAs have started with an initial cash infusion of something closer to 3-4 months of cash flow only, and used rate revenue to build up the rate stabilization fund. Second, the Draft Study does not note who might issue the long-term bonds. The CCA, as a brand-new entity, would not have the financial history to issue long term bonds. Existing California CCAs have relied upon shorter-term loans (3-5 years) for the initial (smaller) working capital infusion and relied upon rate revenue to (slowly) fund the rate stabilization accout. Figure 3 depicts the contingency/rate stabilization fund proposed in the Draft Study for the Central Coast CCA. This fund is calculated every year as a sum of 10% of the total operating expenses (excluding power procurement costs) and 17% of the total power procurement costs. Based on this calculation, the contingency/rate stabilization fund increases every year and ultimately accumulates to $778 million dollars in 2030. The blue bars within Figure 3 illustrate this annual accumulation of the contingency/rate stabilization fund (even without the amount that seemed to be assumed in the initial bond). 9 See SCP fiscal year 2014/15 audited financial statements: https://sonomacleanpower.org/wp- content/uploads/2015/01/08b-2015-and-2014-Final-Audited-Financials.pdf Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-14 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 13 _____________________________________________________________________________________ MRW & Associates, LLC Figure 3: Draft Study’s proposed Central Coast contingency/rate stabilization fund The Contra Costa Study also accounted for a contingency/rate stabilization fund. A crucial difference is that the Contra Costa Study applied an accumulation cap of 15% of the annual operating cost to the contingency/rate stabilization fund. In this case once this cap is reached, no further revenues would be diverted to the contingency/rate stabilization fund unless the reserve funds were withdrawn. Creating a contingency/rate stabilization fund is critical for smooth CCA operations, but revenue allocations to this fund must be balanced against the ongoing need for the CCA’s rates to remain competitive with the local utility’s rates. In the case of the Contra Costa Study, MRW proposed using the contingency/rate stabilization fund to adjust the CCA’s generation rates so that it could remain competitive with PG&E rates. During periods when the total CCA customer rate (i.e. the CCA costs plus the PG&E exit fee) was below the projected PG&E generation rate, the Contra Costa Study proposed increasing the CCA rates upwards to layaway revenue into the contingency/rate stabilization fund up to the 15% cap, while still maintaining a discount. During periods when the total CCA customer rate would otherwise exceed the projected PG&E generation rate, the Contra Costa Study proposed drawing upon the revenue surplus within the contingency/rate stabilization fund to offset some of the costs that would otherwise have to be recovered from CCA customers through the CCA generation rate. Based on this methodology, the Contra Costa CCA would meet the 15% cap for its contingency/rate stabilization fund during the first three years of operation. After those first three years, there would be minimal additions to the fund due to load growth. Figure 4 illustrates MRW’s proposed accumulation of revenues for the Contra Costa CCA’s contingency/rate stabilization fund. $0 $100,000,000 $200,000,000 $300,000,000 $400,000,000 $500,000,000 $600,000,000 $700,000,000 $800,000,000 $900,000,000 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Accumulated Contingency/Rate Stabilization Fund Annual Costs Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-15 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 14 _____________________________________________________________________________________ MRW & Associates, LLC Figure 4: Contingency/rate stabilization fund accumulations for the Contra Costa CCA Study Pro-forma results and rate comparisons Figure 5 presents a graphical summary of Draft Study’s pro-forma results for its Central Coast Scenario 2.10 The vertical bars represent the CCA total cost per kilowatt-hour, the green line represents the fixed CCA rate (inclusive of the PCIA but not delivery charges or franchise fees), and the red line represents the IOU average generation rate for the total CCA load. Power costs (in orange) represent on average for 2020-2030 approximately 60% of the total costs. The PCIA (in yellow) represents 13% of the total costs during this same period, and other costs11 (in blue) represent 28%. Based on Figure 5, the formation of the Central Coast CCA seems infeasible for two reasons: 1) the IOU average rate is lower than the CCA average rate and 2) the negative difference between the CCA rate and the CCA total cost. Note, the IOU average rate is lower in the Draft Study than rates presented in other CCA feasibility studies based exclusively within PG&E’s service area, because 67% of the total potential load for the Central Coast CCA is within SCE’s service area. Presently, SCE generation rates are lower than PG&E’s generation rates (e.g. on average SCE generation rates are 6.8¢/kWh and PG&E’s are 9.2¢/kWh). 10 We have kept the franchise fee, CTC, DWB, and all the delivery services charges out of the analysis. 11 Other costs include: salaries and wages, IOU service charges, ESP charges, other start-up costs, professional services, jurisdictional administration, other operating expenses, uncollectable amounts, contingency/ rate stabilization fund, non-operating expenses, interest earnings, unrestricted funds, and debt service. $0 $100,000,000 $200,000,000 $300,000,000 $400,000,000 $500,000,000 $600,000,000 $700,000,000 $800,000,000 $900,000,000 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Accumulated Contingency/Rate Stabilization Fund Annual Costs Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-16 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 15 _____________________________________________________________________________________ MRW & Associates, LLC Figure 5: Central Coast Scenario 2 Pro-forma results In contrast with Figure 5, Figure 6 shows MRW’s pro-forma results from its Contra Costa Study, specifically the RPS equivalent scenario. In this case, the power costs represent 82% of the total costs, PCIA charges represent 13 % and other costs represent 6%. MRW’s Contra Costa Study concluded that the CCA program could be feasible because the CCA rates are lower than the IOU average generation rate. Note, the IOU average rate is higher in the Contra Costa Study than in the Draft Study because Contra Costa is located exclusively within PG&E’s service territory. Also note, another key difference between these analyses is that for the Contra Costa Study, the CCA rate was kept equal to the CCA total cost per kilowatt. CCA rate IOU rate 0 2 4 6 8 10 12 14 16 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030¢/kWhOther costs PCIA Power Costs Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-17 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 16 _____________________________________________________________________________________ MRW & Associates, LLC Figure 6: Contra Costa RPS equivalent Pro-forma results As one last point of comparison, MCE appears to have other costs equivalent to 9% of its total power procurement costs for 2016 (versus 7% forecasted for the Contra Costa CCA and 47% forecasted for the Central Coast CCA).12 7. Do you have any other suggestions for reducing CCA costs in light of the evolving California CCA market place? Please see MRW’s suggested revisions in response to questions 4, 5 and 6. 8. Does the Draft Study present an adequate analysis of potential economic benefits and challenges of various supply scenarios? And 9. Should any additional benefits or challenges be considered? The Draft Study considered the employment impacts of two separate mechanisms: those jobs created by the increased disposable income from lower electric bills and the jobs associated with local 12 Based on MCE’s FY2015/16 audited financials: https://www.mcecleanenergy.org/wp- content/uploads/2016/09/MCE-Audited-Financial-Statements-2015-2016.pdf Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-18 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 17 _____________________________________________________________________________________ MRW & Associates, LLC investment in renewable resources. Given that the Draft Study found no bill savings, it did not perform any analysis of employment impact associated with bill savings. If Willdan chooses to implement some of the suggestions made in this memo and finds the CCA to be able to offer lower rates than the incumbent utilities, then the bill savings-related jobs analysis should be conducted. The Draft Study assessed the potential economic development benefits associated with CCA building 1, 5 or 10 megawatts of solar projects or 100 MW of wind projects using the Jobs & Economic Impact Development (JEDI) model developed by the National Renewable Energy Laboratory. These projects are not explicitly included in the pro forma analyses, and must be seen as illustrative only. The JEDI model is the most commonly used tool to estimate these kinds of impacts of renewable power project development, and is appropriate. The Draft Study also acknowledged that the opportunity for larger-scale (i.e., not simple behind-the-meter rooftop) solar is limited within the study area. The estimated impacts depend on the number of jobs created and the salaries for each position. In addition, if the jobs are not sourced locally, but rely on workers from other areas of the country, state or region, the local direct impacts would diminish. The JEDI model uses “economic multipliers” to approximate impacts within the supply chain (e.g., manufacturing job creation). These multipliers are only estimates of potential effects and, perhaps more importantly, may not fully take into consideration that these effects may occur outside the local area. It is possible, for example, that the manufacturing jobs created because of power projects would be out of the local area or the U.S. entirely. The JEDI model estimates the direct, indirect and induced effects associated with new power projects, but does not take into consideration that there could be a negative “ripple” effect associated with higher rates necessary to pay for these projects over time. In other words, if residents and businesses pay higher rates for local projects, they could spend less money in the local economy, which could have negative indirect and induced multiplier effects. While we would not expect that these negative indirect and induced effects would cancel out benefits of local projects, they were not acknowledged or included in the analysis. 10. Does the Draft Study provide a thorough evaluation of the prospective CCA’s ability to achieve rate competitiveness with PG&E and SCE? What other factors, if any, should be considered? Because the Draft Study was not finding CCA to be cost-effective, it did not explore any explicit sensitivity cases. If Willdan chooses to implement some or all the recommendations and finds that the CCA rates can be competitive, sensitivity cases should be run to evaluate how robust the results are to reasonable variations in key inputs. These should include: • Lower SCE and PG&E rates • Higher PCIA • Higher Renewable costs • Higher gas prices The Monte Carlo simulation modeling approach used in the Draft Study also provides an opportunity to reflect uncertainty in CCA costs. It does not appear, however, that the rate comparisons in the Draft Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-19 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 18 _____________________________________________________________________________________ MRW & Associates, LLC Study report utilize the Monte Carlo simulation model results. It would be helpful to incorporate these results into the rate comparison. 11. Does the Draft Study consider all pertinent factors to assess the overall cost-benefit potential of CCA? Subject to the concerns and recommendations expressed in prior responses, all pertinent factors were included. 12. Does the Draft Study consider all pertinent risk factors involved with establishment and operation of the CCA program, and are such factors properly weighted and analyzed? Appendix B, sections 3 (technical risks) and 4 (external risks) of the Draft Study enumerate the major risks and presents reasonable mitigations to those risks. With respect to technical risks, the Power Procurement Risk: Power procurement risk includes wholesale power price spikes, uncertain load, intermittent renewable generation. The Draft Study suggests that the CCA can mitigate risk by “having a robust power supply plan, diversifying supply portfolios by production type, generation size and location, contract length, timing of contract purchases, and the use of hedging instruments ….” These are overall reasonable suggestions and should be refined and acted upon if the CCA moves forward. Regulatory Risk: The Draft Study accurately notes that the landscape for CCA is changing, and that these changes must be monitored. Exit Fee and Non-bypassable Charges: The Draft Study notes “The implication for the Central Coast Power CCA [of exit fees] is that even if the CCA’s primary power supply portfolio were cost-competitive with the existing supply costs, added PCIA and CRS charges may increase the overall costs such that the CCA’s offering would ultimately not be competitive with the IOU. This is especially true when considering the amount of load currently under consideration for CCA.” It further specifically identifies the ongoing application by SCE and PG&E (along with SDG&E) to revise the exit fee structure, which would likely increase further the IOU fees on CCA customers. The Draft Study further suggests, Given the relative size of the potential PCIA and CRS fees due to departing customers, Central Coast Power could attempt to procure excess IOU RPS contracts, which would both reduce the IOUs’ stranded costs and begin developing Central Coast Power’s renewable generation portfolio. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-20 Central Coast Region August 2017 Peer Review of CCA Feasibility Draft Study Page 19 _____________________________________________________________________________________ MRW & Associates, LLC While MRW finds the prospect of restructuring the IOU renewable contracts to be remote, we fully concur that it must be more fully evaluated if Central Coast Power moves forward towards CCA implementation. Opt-out risk: As shown in other CCA studies, the risk of higher- or lower-than expected initial opt-out is relatively modest. The Draft Study correctly states that opt-out risk once the CCA has begun service can be minimized by competitive rates (“economic advantage”), providing good customer services (“customer experience”), and offer products and services desired by the CCA customers (e.g., easy to implement solar rooftop agreements). Renewable Generation risk: The Draft Study extensively discusses solar “over-generation” (i.e., solar generating more power during some hours than is needed by the CCA) and what is needed to integrate the solar into its overall power procurement profile. The observations in this section are accurate, and should be addressed if the CCA pursues a portfolio with particularly high solar content. Appendix L: Peer Review and Response Technon Community Choice Aggregationical Feasibility Study L-21 Central Coast Region August 2017             This page intentionally left blank.  Appendix L Peer Review and Response Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 L-23 2. MRW Extended Peer Review             This page intentionally left blank.  MEMORANDUM To: Jennifer Cregar, Project Supervisor, Energy and Sustainability Initiatives, County of Santa Barbara From: Mark Fulmer, Anna Casas Llopart, and Jeremy Waen Subject: Willdan Pro-Forma with Alternative Assumptions Date: August 16, 2017 (Updated) The County of Santa Barbara (“County”) provided to MRW a community choice aggregation (CCA) pro- forma model that was originally created by Willdan Financial Services (“Willdan”) to inform Willdan’s preparation of a technical feasibility study (“Draft Study”) for the County and participating jurisdictions throughout San Luis Obispo, Santa Barbara, and Ventura Counties. At the request of the County, MRW edited the Draft Study pro-forma model according to MRW’s recommendations detailed in a peer review memorandum dated May 31, 2017. MRW made modifications to the pro-forma model for the following scenarios:  Advisory Working Group (AWG) Middle of the Road (50% renewable) Scenario, where the AWG includes 11 jurisdictions across San Luis Obispo, Santa Barbara, and Ventura Counties  Unincorporated Santa Barbara County Middle of the Road (50% renewable) Scenario  Unincorporated San Luis Obispo County Middle of the Road (50% renewable) Scenario MRW made changes to the underlying community choice aggregator (CCA) cost assumptions and updated Pacific Gas and Electric (PG&E) and Southern California Edison (SCE) rate forecasts based on its professional opinion. While the Willdan pro-forma model provides output comparisons for specific rate schedules, because of the fundamentally different approach that MRW takes with respect to the rate comparisons, the model’s specific rate output pages are not impacted by the changes MRW made to the CCA cost assumptions or PG&E/SCE rates. That is, some of the original model functionality is lost. Notably, changes made to the model do not allow an assessment of the annual net operating position. Instead, MRW established average rates to recover 100% of revenues. Each year, the CCA’s net operating position is, by definition, balanced by rate increases/decreases. To fully update the original pro forma Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-25 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 2 _____________________________________________________________________________________ MRW & Associates, LLC model according to MRW’s rate-setting approach would require significant modification to the spreadsheets, which was beyond the scope of our task.1 Summary of Conclusions Using the MRW alternative assumptions, the average CCA operational costs (i.e., the average rate the CCA could offer while covering all costs) for the AWG Middle of the Road Case is approximately 23% lower, on average, than that with the base assumptions (see Figure 1). Nearly half of the decrease is associated with the lower renewable power cost assumption; the bulk of the remaining cost reduction comes from reduced uncollectible expenses, elimination of the franchise fees as an expense (as well as a revenue) and revisions to the reserve fund. Some changes, including the cost of natural gas generation and updates to the power cost indifference adjustment (PCIA), modestly increased the CCA costs. See Table 1 for a summary of CCA cost impacts from the changes made by MRW. This decrease in operating costs (and therefore CCA rates), coupled with the alternative PG&E/SCE rate forecasts, shows, for the AWG Middle of the Road Case, the CCA initially would need to set its rates higher than the investor-owned utilities (IOUs) in order to cover its costs in 2020 to 2022. The CCA may be able to offer nominally similar rates as the IOUs for 2023 to 2027 and modestly lower rates thereafter. See Table 2 and Figures 2, 3 and 4 for rate comparisons for the AWG Middle of the Road Case. An important factor in the analysis is that PG&E’s generation rates are significantly higher than SCE’s generation rates. This has two implications for the analysis. First, it is more difficult for a CCA to offer competitive rates in communities located in SCE’s service territory than those in PG&E’s. Second, the CCA being considered here may choose to set different rates for customers located in PG&E’s service territory versus those in SCE’s service area. The net result of this differential between the two utilities’ generation rates is that a CCA is more likely to be rate-competitive—or even offer a rate savings—for customers located in PG&E territory (i.e., San Luis Obispo County and northern Santa Barbara County); whereas, the CCA is not likely to be able to offer rates that are competitive with SCE for customers located in SCE territory (i.e., southern Santa Barbara County and Ventura County). Because San Luis Obispo County and parts of Santa Barbara County are in PG&E territory, where a CCA may be more competitive, MRW also used the Willdan pro-forma model to compare the potential CCA’s rates for the Unincorporated San Luis Obispo County Middle of the Road and Unincorporated Santa Barbara County Middle of the Road Cases. In both cases, after the first year phase-in, the CCA’s rates are projected to be generally comparable to the weighted average of the SCE and PG&E rates (Santa Barbara County) or PG&E rates (San Luis Obispo County). Please note that MRW conducted this analysis using a tool which it did not design and an analytical approach which MRW does not typically take. While the results for the unincorporated counties may 1 Sheets in red became nonfunctional after MRW edits. Also, in “CCA Operating Results,” “PG&E Escalation,” and “SCE Escalation” sheets, cells inside a red square are nonfunctional. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-26 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 3 _____________________________________________________________________________________ MRW & Associates, LLC suggest that the CCA could offer competitive rates, MRW would need to perform additional, independent analyses before offering a conclusion. Model Changes All the adjustments are highlighted using orange color2 in the MRW edited version of the pro-forma model. Table 1 summarizes the quantitative impacts of the adjustments. The adjustments applied are the following3: 1. CCA renewable contracts. The Draft Study’s use of utility-average renewable contract prices does not reflect the most recently-reported contract prices and does not reflect the general downward trend in renewable prices seen over the past few years. According to the 2016 Padilla report4, the weighted average prices for renewable contracts approved in 2016 are $59/megawatt-hour (MWh) for PG&E and $61/MWh for SCE. Based on this and the flat tendency showed in Table ES - I from the Draft Study, MRW considered $60/MWh as a price for the renewable contracts for 2016-2030 (30% lower than Draft Study price estimates). MRW edited column N from “Tri County RPS Equiv” sheet. 2. CCA natural gas generation. Based on the Draft Study’s analysis, natural gas generation costs are forecast to decrease by 25% from $41/MWh in 2020 to $31/MWh in 2030. This trend analysis may be underestimating natural gas generation costs over the long term. Natural gas prices are relatively low at present, but according to the U.S. Energy Information Administration’s (EIA’s) 2017 Annual Energy Outlook, natural gas prices for electricity generation in the Pacific region are expected to increase by an average of 3.5% per year between 2020 and 2030. Since natural gas generation is typically on the margin in the California wholesale power market, power production costs for market power are driven by the price for natural gas. MRW forecasted natural gas prices based on current New York Mercantile Exchange (NYMEX) market futures prices for natural gas and PG&E’s tariffed natural gas transportation rates. MRW used a standard methodology of multiplying the natural gas price by projected heat rate for a gas-fired generator in the EIA’s 2017 Annual Energy Outlook5 and adding in variable operations and maintenance costs to calculate total power production costs. In addition, MRW added the cost of the greenhouse gas allowances calculated based on the auction floor price stipulated by the California Air Resources Board’s cap-and-trade regulation. Following this methodology, MRW estimated natural gas generation costs equal to $33/MWh for 2020, increasing on average 3% annually. MRW edited cells T19:V29 and column N from “Tri County RPS Equiv” sheet. 2 Cells with edited formulas are highlighted in light orange. 3 MRW edited row 24 from “CCA Expenses” expenses. 4http://www.cpuc.ca.gov/uploadedFiles/CPUCWebsite/Content/About_Us/Organization/Divisions/Office_of_ Governmental_Affairs/Legislation/2017/Final%20-%20Padilla%20Report%20-%20RPS%20Costs%202017.pdf 5 EIA 2017 AEO, Supplemental Table 55.20 (California) Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-27 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 4 _____________________________________________________________________________________ MRW & Associates, LLC 3. Jurisdictional administration. It is atypical for a CCA to reimburse the local jurisdictions for staff-time spent interfacing with the CCA. The one area where this might be practiced is with Single Jurisdiction (rather than Joint Powers Authority) CCAs where staff is shared between local government and CCA operations. Even in those cases, this “Jurisdictional Administration” category seems to overlap with the Salary & Wages category. As a result, these costs should not be considered part of the CCA’s operating expenses. MRW excluded these costs from the Operation Expenses analysis, editing cell D7 from “General Assumptions” sheet. 4. Administrative labor costs. The number of employees (45 full-time equivalents [FTEs]) assumed in the Draft Study pro-forma analysis, as well as their compensation, appear high relative to operating California CCAs. MRW lowered the staff to 35 FTE, editing column E from “Labor input worksheet” sheet. MRW did not adjust the compensation. 5. CCA service fees. MRW updated the service fees based on more recent fee data from the Meter Data Management Agent (MDMA), PG&E’s testimony6 and SCE’s settlement agreement.7 MRW edited cells K15, K18, and K19 from “PG&E Annual Service Costs” sheet and K14, K18, and K20 from “SCE Annual Service Costs” sheet. 6. Franchise Charges. The Draft Study pro-forma analysis appears to assume the franchise fees as an operating expense but not as a revenue for the CCA. Franchise fees are collected from CCA customers by IOUs, not the CCA, using the Franchise Fee Surcharge. This means that the same franchise fees are collected from CCA customers that would be collected from them had they been bundled customers. As such, it has no impact on the bundled versus CCA rate comparison. Therefore, MRW excluded from the analysis the franchise fees expense, editing row 30 from “CCA Operating Results” sheet. 7. PG&E and SCE PCIA escalation. The Draft Study relies upon a forecast of the PCIA rate from the utilities’ green tariff forecasts. This is not an unreasonable assumption, but doesn’t account for CCA departure in 2020-2022. In general, in the 2020’s, MRW sees the PCIA rates tending to decrease year to year. For conservatism, MRW kept PG&E and SCE’s PCIA constant starting in 2021. In addition, MRW updated the 2018 PCIAs according to the IOUs’ 2018 Energy Resource Recovery Account (ERRA) applications. While these rates are not adopted, the ERRA applications provide a good estimate as to what the upcoming year’s rates will be. MRW edited I6:R14 and F17:F25 from “PG&E Escalation” and “SCE Escalation.” 6 PG&E 2017 General Rate Case, Phase 2 (CPUC Application 16-06-013), Testimony Exhibit PG&E-2, Appendix C. June 30, 2016. https://pgera.azurewebsites.net/Regulation/ValidateDocAccess?docID=378139 7 SCE 2017 General Rate Case, Phase 1 (CPUC Application 16-09-001), Joint Motion of Southern California Edison Company (U 388-E) and the City of Lancaster for Adoption of Settlement Agreement. January 19, 2017. http://www3.sce.com/sscc/law/dis/dbattach5e.nsf/0/26C44E0FA545EC37882580AD0081F6BD/$FILE/A1609001- Joint%20Motion%20for%20Adoption%20of%20Settlement%20Agreement%20City%20of%20Lancaster%20and%20 COS.pdf Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-28 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 5 _____________________________________________________________________________________ MRW & Associates, LLC 8. Reserve Fund. The Draft Study pro-forma analysis appears to assume that approximately $54 million (11% of total annual expenses) is contributed each year to the reserve fund, resulting in a total accumulation of more than $597 million in 2030 (113% of total 2030 expenses). This approach is incorrect. MRW rather set a target amount (e.g., a percent of annual expenses), assumed 3 to 5 years to achieve the fund, and then eliminated further contributions until replenishment is needed. MRW estimated the reserve fund to be set at 10% of the non-power procurement expenses, plus 12% of the power procurement costs. Once this amount is achieved, it is adjusted nominally to account for CCA cost escalation. MRW edited row 34 from “CCA Expenses” sheet. 9. Interest earnings. The Draft Study pro-forma analysis accounts for the interest resulting from the net annual balance. According to MRW’s methodology to evaluate the feasibility of the CCA (explained under “Feasibility” on page 7), MRW simplified and didn’t account for any interest. MRW edited row 45 from “CCA Operating Results” sheet. Startup and Initial Financing Costs MRW’s initial review of the Draft Study called out that the assumed 30-year bond financing was unusual and the amount financed was relatively high. Because we did not offer specific alternatives, we did not include any in our analysis. Nonetheless, as proposed, the start-up cost and financing is particularly high. In general, CCAs begin operations—finding executive staff, office space, etc.—using County funds. Once they have a solid plan in place to deliver power (e.g., an implementation plan, power contractor in place, indicative bids for power), the CCA would arrange for a short-term (5-year) loan to cover the costs already paid for by the County, plus an amount for working capital to cover operating expenses until the first electricity bill revenues are received. A fully-funded rate stabilization fund would not typically be included in an initial financing; instead, the fund would be built with revenues over time. The initial start-up costs would fall in the order of a few million dollars, with the working capital equal to about 90 days of cash flow, or $107 million for the AWG Middle of the Road Case.8 This need for cash flow contributes to CCAs’ desire to phase in implementation. Results of Changes Table 1 and Figure 1 show the impacts on CCA total costs for each one of the MRW adjustments detailed above. As Table 1 shows, using the MRW alternative assumptions, the average CCA operational costs (i.e., the average rate it could offer while covering all costs) is approximately 24% lower, on average, than that with the base assumptions. Nearly half of the decrease is associated with the lower renewable power cost assumption; the bulk of the remaining reduction comes from reduced elimination of the franchise fees as an expense (as well as a revenue) and revisions to the reserve fund. Some changes, including the cost of natural gas generation and updates to the PCIA, modestly increased the CCA costs. 8 This figure is 90 days working capital for the fully-implemented AWG case (i.e., after all customers had been phased in). Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-29 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 6 _____________________________________________________________________________________ MRW & Associates, LLC Table 1 Impact of MRW adjustments on CCA costs, AWG Middle of the Road Case Adjustments Average 2020- 2030 CCA costs [$/MWh] Change [%] Willdan CCA costs - starting point 118.1 1. CCA renewable contracts 102.0 -13.6% 2. CCA natural gas generation 103.9 1.6% 3. Jurisdictional administration 103.8 0.0% 4. Administrative labor costs 103.5 -0.3% 5. CCA service fees 102.9 -0.5% 6. Franchise fees 95.7 -6.1% 7. PCIA escalation and 2018 update 99.1 3.0% 8. Reserve fund 90.0 -7.7% 9. Interest earnings 90.6 0.5% MRW CCA costs (=CCA rate) 90.6 -23.2% Based on the changes described above, the average CCA per-MWh cost obtained from the Draft Study pro-forma has been reduced by 23% on average. Figure 1 shows the differences between both results. The upper green line shows the CCA cost9 from the Draft Study pro-forma; the lower blue line shows the average CCA cost with MRW modifications to the pro-forma. The average per-MWh CCA cost is higher in 2020 because the debt service is relatively constant year to year; whereas, only 30% of the CCA’s load (MWh) is in place in 2020 due to Willdan’s assumptions about phasing in larger commercial and industrial customers first. With fixed costs ($) spread over lower sales (MWh), the average per-MWh cost is higher than later years when the full customer base is phased in. 9 The figures use “average CCA cost” interchangeably with “average CCA rate,” as we assume that rates will cover costs, no matter their relation to SCE and PG&E rates. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-30 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 7 _____________________________________________________________________________________ MRW & Associates, LLC Figure 1 Comparison of CCA Average Cost (Rate) from Draft Study Pro-forma and MRW Edited Pro- forma, AWG Middle of the Road Case AWG Middle of the Road Case Rate Comparison Results MRW used a different methodology than Willdan to assess the CCA feasibility. MRW considers a CCA “feasible” if the CCA average per-MWh cost (i.e., average CCA rate) is lower, on average, than the weighted average IOU generation rate.10 The MRW changes to evaluate the rate-competitiveness of the CCA are detailed below: 10. Comparative IOU generation rates and CCA expenses. The Draft Study sets the CCA rates based on the CCA expenses for 2022-2024 period. MRW assumes that CCA rates will be set to cover the CCA expenses in each year. To account for our different rate-setting approach, MRW created six new sheets “CCA IOU rates”, “PG&E RATES”, “SCE RATES”, “CCA IOU CTC+DWR”, “PG&E CTC+DWR”, “SCE CTC+DWR” and added rows 13-17 to “CCA Operating Results.” 11. PG&E and SCE rate escalation. The Draft Study uses for the rate comparison the total IOU rates (generation plus delivery). To forecast the generation plus delivery IOU rates, the Draft Study uses the annual change in CCA power procurement costs. Instead, MRW only analyzes the generation portion of the IOU rates.11 The MRW IOU generation rate forecast starts with 2018 rates from the IOUs’ 2018 ERRA applications and extends them using internally calculated escalators.12 MRW entered the IOUs’ 2018 ERRA generation rates in cells P12:P20 from “PG&E RATES” and “SCE RATES” sheets and the rate escalators in cells H65:S67 from “CCA IOU rates”. 10 To be consistent with the Willdan analysis, the comparison includes CTC and DWR in the IOU rate and in the CCA expenses. Excluding both is equally valid. 11 See footnote 4. 12 The internal escalators are aligned with the CCA natural gas generation and the CCA renewable contract prices assumed in this report. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-31 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 8 _____________________________________________________________________________________ MRW & Associates, LLC Table 2 compares the CCA’s average cost (i.e., generation rate) with each IOU’s generation rate separately and as a combined weighted average for the AWG Middle of the Road Case.13 For jurisdictions that are located in PG&E’s service area, the Average CCA Cost column can be compared to the “Average PG&E Rate” column. Alternatively, for AWG regions located in SCE’s service area, the Average CCA Cost column should be compared to the Average SCE Rate Column. The IOUs’ rates are lower in 2020 because of the Draft Study assumption that larger commercial and industrial accounts are transferred first to CCA service. Because these customers tend to have the lowest generation rates, the CCA is having to compete with the IOUs’ lowest rate classes while facing high start-up costs. This makes it particularly hard to compete in the first year of operations. Table 2. Rate Comparisons ($/MWh), AWG Middle of the Road Case Average SCE Rate ($/MWh) Average PG&E Rate ($/MWh) Weighted Average Utility Rate ($/MWh) Average CCA Cost ($/MWh) 2020 63.1 90.9 73.3 90.3 2021 71.8 103.2 79.7 86.0 2022 71.2 106.8 79.6 82.2 2023 73.9 105.3 81.5 81.9 2024 74.9 104.5 82.2 82.8 2025 75.9 97.6 81.6 83.3 2026 78.8 98.3 84.1 84.0 2027 79.6 103.8 85.9 84.8 2028 80.7 110.2 88.1 85.7 2029 81.9 117.6 90.5 86.4 2030 84.6 127.0 94.6 87.3 13 For Table 2, 3, 4, Figure 3, 4, 5, and 6, MRW didn’t include the CTC and DWR in the IOU generation rates or in the CCA rates. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-32 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 9 _____________________________________________________________________________________ MRW & Associates, LLC MRW’s comparison between the IOU weighted average generation rate and the average CCA total costs (rate) is shown in Figure 2. Through 2026, the expected IOU weighted generation rate14 (red line) is below average CCA costs (blue line). After 2027, the expected IOU weighted generation rate is higher than the average CCA costs, meaning the CCA may be able to offer competitive, or lower, rates after this 2027 transition point. Figure 2 Comparison of Average CCA Cost (Rate) and Weighted Average IOU Rate, AWG Middle of the Road Case Figures 3 and 4 show the expected PG&E and SCE average generation rates compared to the CCA average costs (generation rate), respectively. Because PG&E generation rates are higher than SCE generation rates, the CCA may choose to set different rates for customers located in PG&E versus SCE service area. The CCA is more likely to be rate-competitive—or even offer a rate savings—for CCA customers located in PG&E territory (i.e., San Luis Obispo County and northern Santa Barbara County); whereas, the CCA is not likely to be able to offer rates that are competitive with SCE for CCA customers located in SCE territory (i.e., southern Santa Barbara County and Ventura County). 14 The IOU rate depicted corresponds to generation rate plus CTC plus DWR. MRW included CTC and DWR because both charges are included as CCA expenses. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-33 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 10 _____________________________________________________________________________________ MRW & Associates, LLC Figure 3. Comparison of Average CCA Cost (Rate) and PG&E Average Rate, AWG Middle of the Road Scenario Figure 4. Comparison of Average CCA Cost (Rate) and SCE Average Rate, AWG Middle of the Road Scenario As discussed on page 6, the particularly low SCE and PG&E average rates in 2020 are attributable to the way that the original Willdan Study phased in the CCA’s customers starting with the largest commercial customers, who also have the lowest IOU generation rates. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-34 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 11 _____________________________________________________________________________________ MRW & Associates, LLC Unincorporated Santa Barbara and San Luis Obispo Counties Middle of the Road Rate Comparison Results MRW was also asked to use the modified Willdan pro forma model to derive CCA-utility rate comparisons assuming stand-alone CCAs covering either unincorporated Santa Barbara County or unincorporated San Luis Obispo County. These analyses used the model changes noted above, plus reflected the load and customer profiles of the unincorporated parts of the respective counties. The analyses did not change any of the underlying CCA costs, which while predominantly fixed, could potentially scale downward with the smaller CCAs. Table 3 and Figure 5 show the results of the analysis for unincorporated Santa Barbara County. After the first year phase-in, the Unincorporated Santa Barbara County CCA’s rates are projected to be generally comparable to the weighted average of the SCE and PG&E rates. This is because of the large number of PG&E accounts in the unincorporated area, where PG&E has higher generation rates relative to SCE. Table 4 and Figure 6 show the results of the analysis for unincorporated San Luis Obispo County. After the first-year phase-in, the Unincorporated San Luis Obispo County CCA’s rates are projected to be generally comparable to the PG&E rates, although with a three-year period from 2025 through 2027 where the CCA rates are projected to be slightly higher than PG&E rates. This anomaly is due to the retirement of the two Diablo Canyon Nuclear Power Plant generators, the output of which is expected to be replaced with power that has a lower average cost than the power currently being generated by Diablo Canyon. Figures 5 and 6 also break down the CCA costs into the major components. This highlights the impact of both the fixed costs and the PCIA. Because unincorporated San Luis Obispo County has smaller loads than the AWG or unincorporated Santa Barbara County, the average fixed costs (upper teal segments of the bar charts) are larger. Because SCE’s PCIA is lower than PG&E’s, Figure 5 shows that the green PCIA segment of the bar charts are slightly smaller for unincorporated Santa Barbara County (which is partially in SCE territory) than unincorporated San Luis Obispo County. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-35 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 12 _____________________________________________________________________________________ MRW & Associates, LLC Table 3. Rate Comparisons ($/MWh), Unincorporated Santa Barbara County Middle of the Road Case Average SCE Rate ($/MWh) Average PG&E Rate ($/MWh) Weighted Average Utility Rate ($/MWh) Average CCA Cost ($/MWh) 2020 61.3 90.3 80.3 95.1 2021 67.8 101.3 88.1 89.7 2022 67.6 104.6 89.4 87.8 2023 70.2 103.1 89.8 87.9 2024 71.2 102.3 89.8 88.9 2025 72.1 95.6 86.5 89.5 2026 74.9 96.3 88.1 90.2 2027 75.7 101.6 91.5 91.1 2028 76.7 108.0 95.5 92.1 2029 77.8 115.2 100.0 92.8 2030 80.4 124.4 106.3 94.0 Figure 5. Rate Comparisons ($/MWh), Unincorporated Santa Barbara County Middle of the Road Case Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-36 Central Coast Region August 2017 Pro-forma results with alternative assumptions Page 13 _____________________________________________________________________________________ MRW & Associates, LLC Table 4. Rate Comparisons ($/MWh), Unincorporated San Luis Obispo County Middle of the Road Case Average SCE Rate ($/MWh) Average PG&E Rate ($/MWh) Weighted Average Utility Rate ($/MWh) Average CCA Cost ($/MWh) 2020 N/A 92.9 92.9 114.7 2021 N/A 106.1 106.1 105.0 2022 N/A 109.7 109.7 102.8 2023 N/A 108.2 108.2 102.5 2024 N/A 107.3 107.3 103.8 2025 N/A 100.3 100.3 104.5 2026 N/A 101.0 101.0 105.5 2027 N/A 106.6 106.6 106.7 2028 N/A 113.2 113.2 108.0 2029 N/A 120.8 120.8 109.0 2030 N/A 130.5 130.5 110.9 Figure 6. Rate Comparisons ($/MWh), Unincorporated San Luis Obispo County Middle of the Road Case Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-37 Central Coast Region August 2017             This page intentionally left blank.  Appendix L Peer Review and Response Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 L-39 3. Response to Peer Review             This page intentionally left blank.  MEMORANDUM 407.872.2467 | 200 South Orange Avenue, Suite 1550, Orlando, FL 32801 | www.willdan.com OVERVIEW The County of Santa Barbara (The County) forwarded to Willdan and EnerNex the above referenced peer review prepared by MRW & Associates (MRW) dated May 31, 2017 (MRW Report). The MRW Report identifies six recommended changes to Willdan’s pro forma analysis. Additionally, the MRW Report cites a concern over the treatment of franchise fees and offers a recommendation concerning the need for additional sensitivity analyses. This memorandum responds to these six suggested revisions and two additional comments. The MRW Report also answers twelve questions posed by the AWG; this memorandum responds to MRW’s responses to these AWG questions in the final section. BACKGROUND The peer-reviewed draft Study was prepared by Willdan Financial Services (Willdan), who conducted the pro forma analysis, and EnerNex, who forecasted load and power procurement pricing. Initial Study results found that the Central Coast Power (CCP) Community Choice Aggregation (CCA) program was not feasible as it resulted in forecasted rate proxies1 that in most cases were higher than those of the incumbent investor owned utilities (IOUs)—Pacific Gas and Electric (PG&E) and Southern California Edison (SCE)—by rate class. As noted on page 2 of the MRW review: Unlike prior recent CCA technical studies, the Draft Study concluded that CCA was not economically feasible even when only the state-required minimum renewable energy content was assumed. MRW’s [sic] focused its review to identify areas where the Draft Study was potentially overly conservative or made questionable assumptions that might explain why its conclusion was negative while others have been affirmative. Each of MRW’s six proposed changes, as discussed below, results in outcomes that favor CCP CCA feasibility. Not one of MRW’s six recommended pro forma analysis changes negatively impacts CCP CCA feasibility. Importantly, the two largest drivers of feasibility results are power pricing and IOU rate forecasts. The former because power prices comprise nearly 70% of CCA annual operating costs; the latter because IOU rate forecasts create the yardstick against which CCA rate proxies are measured. With respect to the former, a large portion of Study effort was devoted to in depth load analysis using actual data obtained from each IOU and power price forecasting as described more fully in the report and 1 The technical Study did not include rate design, rather rate proxies, the unitized revenue requirement by rate class needed to meet the CCA programs financial obligations, were calculated based on cost of service principles. TO: Jen Cregar FROM: Willdan and EnerNex DATE: August 1, 2017 RE: Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-41 Central Coast Region August 2017 Page 2 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 appendices thereto. MRW has conducted no similar analysis. With respect to the latter, the primary scope of the Study was modeling CCA operating costs. Although providing reference rate comparisons was part of the scope, forecasting IOU rates was not part of the scope of work and would require significant additional resources and cost. Even with a considerable budget devoted specifically to forecasting IOU rates, results would at best be tenuous. IOU rates are driven by internal decision making, investor concerns, the Public Utilities Commission, and a host of other factors in addition to wholesale power market prices, all of which can fluctuate considerably. Lack of IOU rate forecasts is a challenge lacking resolution that impacts all CCA feasibility studies. Willdan, therefore used publicly available information and applied reasonable assumptions. Willdan and EnerNex conducted an unbiased, third party review of CCP CCA feasibility. Given, as stated on page 2 of MRW’s peer review—and included on page 1 of this memo—MRW specifically “focused its review to identify where the draft Study was potentially overly conservative or made questionable assumptions that might explain why its conclusion was negative,” we are concerned that the peer review appears biased in favor of CCP CCA feasibility and caution that results based on these recommendations may also be biased accordingly. RESPONSE TO PEER REV IEW 1. CCA RENEWABLE POWER CONTRACTS MRW SUGGESTION The Draft Study’s use of utility-average renewable contract prices does not reflect the most recently- reported contract prices and does not reflect the general downward trend in renewable prices seen over the past few years. WILLDAN RESPONSE Power markets are volatile and dynamic, in particular for the regions addressed in this Study. For example, the recent rain in California has filled the large hydroelectric reservoirs owned and managed by both PG&E and SCE. In 2015, only 2% of SCE’s power content and 6% of PG&E’s power content was produced by large hydroelectric resources.2 In contrast, these resources provided 18% of electricity for PG&E and 7% of electricity for SCE in 2011.3 As a result, recent rainfall is likely to decrease the overall portfolio cost for IOU generation. This weather-dependent cost variable for hydroelectric generation is just one example of IOU power portfolio and retail 2 Power Content Label required by AB 162 (Statute of 2009) and Senate Bill 1305 (Statutes of 1997): http://www.energy.ca.gov/pcl/labels/ 3 Utility Annual Power Content Labels 2011: http://www.energy.ca.gov/pcl/labels/2011_index.html Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-42 Central Coast Region August 2017 Page 3 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 cost volatility. Similar weather dependence applies to both sunshine and wind for renewable generation portfolios. Renewable Generation The Study was initiated in the summer of 2016 using the 2016 Padilla Report,4 among other resources; the preliminary results were released in May of 2017. The 2017 Padilla Report5 was released in May 2017, more than four months after the Study forecast was finalized. As in any Study of this nature, data must be analyzed as of a point in time. The forecast used in the Study does capture the downward trend as of the forecast date and the team stands by the forecasts presented as of the time of the Study. As discussed below, the forecast is not inconsistent with the updated findings of the 2017 Padilla Report. MRW cites the 2017 Padilla Report versus the Study as follows: The weighted average prices for contracts approved in 2016 are $0.059/kWh for PG&E and $0.061/kWh for SCE, well below the average 2016 expenditures of $0.11/kWh and $0.094/kWh, respectively. The prices of contracts approved in 2016 are approximately 30% below the average RPS [Renewable Portfolio Standard] PPA [Purchase Power Agreement] cost of $88/MWh [$0.088/kWh] assumed in the Report for 2020. However, this information must be considered in light of the full set of data presented in the report and against all trends reported. The 2017 Padilla Report notes that certain actual 2016 procurement costs increased over 2015: bundled renewable supply to $0.104/kWh from $0.101/kWh in 2015. PG&E paid a premium for bundled RPS in 2016, an average of $0.1119/kWh. SCE paid $0.0942/kWh that same year. SCE’s actual average cost for 2015 was revised upward to $0.0905 from the $0.087 originally reported in the 2016 Padilla Report. The corresponding chart in the CCP CCA study has been updated accordingly, is included below as Figure 1, and illustrates that the RPS costs for all three IOUs are actually higher than the CCA forecast price for 2016. 4 http://www.cpuc.ca.gov/uploadedFiles/CPUC_Website/Content/Utilities_and_Industries/Energy/ Reports_and_White_Papers/Padilla%20Report%202016%20-Final%20-%20Print.pdf 5 http://www.cpuc.ca.gov/uploadedFiles/CPUCWebsite/Content/About_Us/Organization/Divisions/ Office_of_Governmental_Affairs/Legislation/2017/Final%20-%20Padilla%20Report%20-%20RPS%20Costs%202 017.pdf Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-43 Central Coast Region August 2017 Page 4 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Figure 1 IOU RPS compliance cost.6 With significant solar generation growth in California, from both utility scale and distributed customer owned photovoltaic resources, solar generation output sometimes exceeds electricity demand during periods of peak solar output. California is entering an over-capacity condition for solar generation during certain daylight periods which means that additional solar generation capacity is not needed and that solar is no longer displacing fossil fuel generation. This over- capacity condition results in negative pricing in the CAISO day-ahead and real-time markets during periods when excess solar production exceeds demand. Battery energy storage is one 6 The basis of the renewable RPS cost analysis included data from the May 2016: Report on 2015 Renewable Procurement Costs in Compliance with Senate Bill 836 (Padilla, 2011) Table A -2 Weighted Average TOD-Adjusted RPS Procurement Expenditures (Bundled Energy Only) for 2015 http://www.cpuc.ca.gov/uploadedFiles/CPUC_Website/Content/Utilities_and_Industries/Energy/Reports_and_W hite_Papers/Padilla%20Report%202016%20-Final%20-%20Print.pdf; Subsequent to the analysis an updated report was produced and the data was consistent with the forecast analysis previously performed: May 2017: Report on 2015 Renewable Procurement Costs in Compliance with Senate Bill 836 (Padilla, 2011) Table B -2 Weighted Average RPS Procurement Expenditures (Bundled Energy Only) for 2016 http://www.cpuc.ca.gov/uploadedFiles/CPUCWebsite/Content/About_Us/Organization/Divisions/Office_of_Gove rnmental_Affairs/Legislation/2017/Final%20-%20Padilla%20Report%20-%20RPS%20Costs%202017.pdf. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-44 Central Coast Region August 2017 Page 5 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 technology being pursued to help mitigate this overcapacity challenge. For reference, the LA Times article: “California invested heavily in solar power. Now there's so much that other states are sometimes paid to take it,”7 provides a clear discussion of this situation. Natural Gas Generation The supply cost for natural gas generation used in the Study incorporated two factors: 1) a decreasing cost for the natural gas commodity as a result of increasing supplies from shale gas and fracking; and 2) an improved heat rate efficiency for natural gas electric generation. However, the cost of natural gas is also volatile as illustrated in the corresponding figures “California natural gas generation cost based on natural gas price and heat rate conversion” and “Natural gas generation supply cost” in the Study. The curve fitting regression analysis in the “Natural gas generation supply cost” is an averaging and flattening of the recent natural gas generation cost trend with actual historical prices being both above and below the cost forecast. The Monte Carlo simulation model estimates the corresponding volatility of natural gas prices ($/MWh) based on the 2002-2016 data source. CCA Renewable Power Contracts The 2016 approved contracts referenced in the 2017 Padilla Report are primarily for supplies that will be provided in the future, and likely after 2020, for deals entered today. Given the dynamic nature of this market, prices may move in either direction. The forecast used in the Study stands as reasonable. Summary Comments Finally, MRW indicates that the Study is over-estimating the cost of future renewables and under-estimating the cost of natural gas generation. Although MRW suggests that we revise downward the renewables forecast, it does not similarly suggest that we also revise upward the natural gas generation price forecast. This one-sided recommendation further evidences a bias towards a feasible outcome, which must be rejected. Exhibit A hereto presents the results of sensitivity analyses conducted against Participation Scenario 2: Advisory Working Group (AWG) Jurisdictions – Middle of the Road scenario that illustrate the impact of changes in power costs to feasibility results. Demonstrating that, all other assumptions held constant, a 40% reduction in power costs is required to achieve rate proxies lower than both IOUs. 7 L.A. Times “California invested heavily in solar power. Now there's so much that other states are sometimes paid to take it” by Ivan Penn, June 22, 2017: http://www.latimes.com/projects/la-fi-electricity-solar/ Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-45 Central Coast Region August 2017 Page 6 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 2. UNCOLLECTABLE EXPENS ES MRW SUGGESTION a) The Study assumed from 5% to 8% of the revenues due to the CCA from its customers could not be collected. This is an order-of-magnitude higher than that experienced by either MCE Clean Energy (MCE), the longest-running CCA in the state, or Sonoma Clean Power (SCP), the second longest-running CCA in the state. b) CCAs do not observe the same level of uncollectible accounts as the IOUs due because CCAs are allowed to return non-paying accounts to the corresponding IOU’s bundled service. WILLDAN RESPONSE a) The Study assumption was based on the actual filings by PG&E and SCE using the ratio of Uncollectable Account allowance to total Receivables. In response to MRW’s suggestion, additional research was conducted that revises this assumption. In the 2014 General Rate Case Decision 14-08-0321, the California Public Utilities Commission (Commission or CPUC) adopted a revised methodology to determine PG&E’s uncollectibles factor, which is based on a 10-year rolling average using recorded uncollectible data. The 2015 uncollectibles factor using historical data from 2004 through 2013 is 0.003325.SCE’s authorized uncollectibles factor for 2010 and 2011 was 0.00240 and for 2012 to 2013 was 0.00204. However, SCE’s actual uncollectible expense exceeded the authorized amount in each of these years and exhibits an increasing trend. Based on these analyses, Willdan agrees that it makes sense to revise the pro forma assumption to reflect the actual expense set by the CPUC for PG&E of 0.3325%; this factor has been applied to both IOUs. Revision of this assumption in isolation does not materially impact forecasted feasibility outcomes. b) Willdan does not concur with MRW’s assertion in practice nor in principle. Although a CCA is technically allowed to return clients to the IOU for non-payment, such treatment appears to conflict with the CCA’s role in the public power paradigm. CPUC Code Section 366.2(c)(3) lists requirements for CCAs that indicate if a public agency seeks to serve as a CCA, it shall offer the opportunity to purchase electricity to all residential customers within its jurisdiction. Furthermore, for purposes of a feasibility study, such an assumption defies industry standards and practice and is, therefore, indefensible. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-46 Central Coast Region August 2017 Page 7 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 3. ADMINISTRATIVE LABOR COSTS MRW SUGGESTION The number of employees assumed in the pro forma analyses, as well as their compensation, appear high relative to operating California CCAs. WILLDAN RESPONSE Willdan based its labor analysis on the regional labor markets and a functional analysis of required positions. Figure 2 below demonstrates the level of staffing is reasonable when compared to other CCAs.8 Labor costs include benefits. Figure 2: CCA Staffing Comparison Figure 3 below illustrates the size of the CCP CCA relative to other currently operating CCAs by Participation Scenario, illustrating the extreme range between scenarios assessed. Staffing 8 Based on Participation Scenario 2: AWG Jurisdictions. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-47 Central Coast Region August 2017 Page 8 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 assumptions are adjusted by scenario and range from a low of 24 for Participation Scenario 8: City of Santa Barbara and a high of 57 for Participation Scenario 1: All Tri-County Region. Figure 3: Summary of CCA Size (GWh and Customer Accounts) Willdan conducted sensitivity analyses concerning staffing levels. Exhibit B hereto presents the results of this sensitivity analysis. Decreasing staffing by over 70% in isolation did not materially alter feasibility outcomes. 4. CCA SERVICE FEES MRW SUGGESTION a) The incumbent utilities—Southern California Edison (SCE) and Pacific Gas and Electric (PG&E)— charge CCAs in their respective territories certain fees for billing conducted on behalf of the CCA as well as meter and data management. While the Draft Study reflects current tariffed rate for these services, it does not account for the proposed dramatic uncontested reductions being presented by both utilities. b) Similarly, it is unclear whether the ESP service fees section of the Draft Study properly accounts for critical operational services such as data management and scheduling coordination. WILLDAN RESPONSE a) As noted by MRW, the Study relies upon current tariffed rates for CCA Service Fee at the time of the Study. No other assumption concerning pending proposals would be defensible. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-48 Central Coast Region August 2017 Page 9 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 b) The Study adequately accounts for all required CCA functions as more fully described in the report. 5. ASSUMED RESERVE FUND ING MRW SUGGESTION Beyond working capital, CCAs typically develop a “rate stabilization reserve fund” which can be drawn upon in years’ where the CCA might not otherwise be able to meet its rate targets. The Draft Study pro forma analysis appears to assume that approximately $78 million (14% of total expenses) is contributed each year, rather than setting a target (e.g., 15% of annual expenses), taking 3 to 5 years to achieve the fund, and then eliminate further contributions until replenishment is needed. WILLDAN RESPONSE A contingency fund is budgeted for unanticipated occurrences over the course of a year. The pro forma assumes that each year a certain amount is set aside to cover unanticipated increases in operating costs. In the most recent version of the pro forma, the annual amount set aside for the rate stabilization fund was lowered to 12% of power costs (previously 17%). The contingency fund remains at 10% of non-power O&M. Usage of the contingency fund was not modeled—there are no withdrawals—so MRW’s assumption that the fund continues to grow is incorrect. The purpose of the contingency fund is to provide adequate funding given a reasonable increase in operating costs; given that the opt-out rate was set conservatively high and power procurement costs can fluctuate significantly, it should be assumed that the contingency fund will be used. Altering the level of contingency and reserve funding (while maintaining reasonable levels) in isolation would not materially alter feasibility outcomes. 6. PG&E AND SCE RATE FO RECASTS MRW SUGGESTION A fundamental concern is that the forecast of SCE and PG&E rates is disconnected from the forecast of CCA rates. The utility rates against which the CCA rates are compared are simply the current rates escalated at 0-0.5%. It does not account for: (i) SCE’s or PG&E’s actual supply portfolio, (ii) the two utilities’ status with respect to State’s renewable power content mandates, (iii) fuel price trends, or (iv) any other underlying fundamentals. In particular, there is no explicit connection between the utilities’ generation rates and the CCA generation cost, even though they would be purchasing from the same wholesale market and vying for the same incremental renewable generation sources. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-49 Central Coast Region August 2017 Page 10 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 WILLDAN RESPONSE In the prior section MRW contends that the renewable rates for the IOUs for 2016 are high and not representative of the market from which the CCA would be purchasing. However, here MRW contends that there should be an explicit connection between the IOU generation rates and the CCA generation cost. Renewables are currently the most expensive resource in the IOUs’ supply portfolio. On the one hand MRW contends that CCA prices for renewables should be much lower than the IOUs are currently paying, but at the same time that IOU rates and CCA rates should be connected. This appears to be contradictory, and depending on interpretation, could bias results in favor of feasibility. As noted in the Background section of this memorandum, forecasting IOU rates was not part of the scope of work of this Study. Additionally, lack of insight into IOU rate forecasts is a challenge faced by all CCAs. Furthermore, CCAs compete only on the energy-related component of rates. CCA and IOU bundled service customers alike pay the delivery portion of the IOU bill which covers transmission and distribution. Additionally, CCA customers pay an exit fee to reimburse the IOU for generation related costs “stranded” when the CCA load leaves the IOU—i.e., the Cost Recovery Surcharge (CRS), in particular the Power Charge Indifference Adjustment(PCIA). When discussing rate forecasts and escalations, the non-energy component of IOU rates could escalate by 15%, and not impact Study outcomes (independent of other potential adjustments to Study assumptions) because both CCA and non-CCA customers would pay that increase. As discussed in more detail with the following tables and figures, Willdan has demonstrated that both PG&E and SCE have, over the last few years, been moving more of the revenue requirement from generation to transmission and distribution costs—in other words shifting costs to the fixed delivery charge paid by both CCA and non-CCA customers. Table 1 shows historical energy and delivery charges for SCE for the Residential rate class since 2014, for the baseline consumption. Overall for this period, the delivery charge has increased 89% while the energy component has decreased 13%. Table 1: SCE Rate Changes Since 2014, Residential Baseline 2014 2015 2016 2017 % Change 2014-2017 RESIDENTIAL, Baseline Usage Basic Service Fee $/Meter/Month 0.94292 0.94292 0.94292 0.94292 Energy Summer $/kWh 0.08555 0.0899 0.06887 0.07477 Winter $/kWh 0.08555 0.0899 0.06887 0.07477 Increase/Decrease 5%-23%9%-13% Delivery Summer $/kWh 0.04678 0.0586 0.08221 0.0884 Winter $/kWh 0.04678 0.0586 0.08221 0.0884 Increase/Decrease 25% 40%8%89% California Climate Credit $0.00 ($4.83) ($6.33) ($5.17) Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-50 Central Coast Region August 2017 Page 11 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Table 2 and Table 3 on the following pages show the historical rate changes occurring for the Medium and Large Commercial classes, respectively. Overall for this period, the delivery charges increased and the generation charges decreased for both classes. Table 2: SCE Rate Changes Since 2014, Medium Commercial 2014 2015 2016 2017 % Change 2014-2017 GENERAL SERVICE, TOU-GS-3 Basic Service Fee $/Meter/Month 444.790 441.930 493.360 446.130 Increase/Decrease -1%12%-10%0% Energy Summer On-Peak $/kWh 0.30087 0.33132 0.23913 0.28916 Increase/Decrease 10%-28%21%-4% Mid-Peak $/kWh 0.10158 0.1119 0.08078 0.08281 Increase/Decrease 10%-28%3%-18% Off-Peak $/kWh 0.03227 0.03555 0.02568 0.03226 Increase/Decrease 10%-28%26%0% Winter Mid-Peak $/kWh 0.05581 0.06148 0.04537 0.04662 Increase/Decrease 10%-26%3%-16% Off-Peak $/kWh 0.03681 0.04055 0.02927 0.03712 Increase/Decrease 10%-28%27%1% Voltage Discount, Energy 50kV<220kV $/kW (0.00404) (0.00440) (0.00320) (0.00461) Increase/Decrease 9%-27%44%14% Delivery Summer On-Peak $/kWh 0.02332 0.02691 0.02557 0.02718 Increase/Decrease 15%-5%6%17% Mid-Peak $/kWh 0.02332 0.02691 0.02557 0.02718 Increase/Decrease 15%-5%6%17% Off-Peak $/kWh 0.02332 0.02691 0.02557 0.02718 Increase/Decrease 15%-5%6%17% Winter Mid-Peak $/kWh 0.02332 0.02691 0.02557 0.02718 Increase/Decrease 15%-5%6%17% Off-Peak $/kWh 0.02332 0.02691 0.02557 0.02718 Increase/Decrease 15%-5%6%17% Demand Charges Facilities Related $/kW $16.14 $16.07 $18.45 $17.81 Increase/Decrease 0% 15%-3%10% Voltage Discount, Demand Facilities Related 50kV<220kV $/kW (6.76000) (6.71000) (7.46000) (6.79000) Increase/Decrease -1%12%-21% -12% Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-51 Central Coast Region August 2017 Page 12 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Table 3: SCE Rate Changes Since 2014, Large Commercial Unfortunately, this type of historical delivery data was not available for PG&E; PG&E does not post historical tariffs on its website and provides only bundled data for previous years’ rates. However, the California Public Utilities Commission April 2016 report entitled “Electric and Gas Utility Cost Report” provides illustrative data comparisons between the rates and Revenue Requirements of the three state IOUs: PG&E, SCE, and San Diego Gas and Electric (SDG&E). Information from that report has been inserted into this memo for discussion purposes. 2014 2015 2016 2017 % Change 2014-2017 GENERAL SERVICE-LARGE, TOU-8-Option B Basic Service Fee $/Meter/Month 321.60 319.47 356.41 303.25 Increase/Decrease -1%12%-15%-6% Energy Summer On-Peak $/kWh 0.10485 0.11445 0.08309 0.07072 Increase/Decrease 9%-27% -15% -33% Mid-Peak $/kWh 0.05449 0.05948 0.04318 0.04730 Increase/Decrease 9%-27%10%-13% Off-Peak $/kWh 0.03241 0.03537 0.02568 0.03165 Increase/Decrease 9%-27%23%-2% Winter Mid-Peak $/kWh 0.05616 0.06130 0.04451 0.04579 Increase/Decrease 9%-27%3%-18% Off-Peak $/kWh 0.03738 0.04081 0.02963 0.03645 Increase/Decrease 9%-27%23%-2% Demand Charges Time Related Summer On-Peak $/kW 28.23 30.81 22.38 22.55 Increase/Decrease 9%-27%1%-20% Mid-Peak $/kW 0.00 0.00 0.00 3.63 Increase/Decrease 0%0%N/A Delivery Summer On-Peak $/kWh 0.02162 0.02463 0.02331 0.02426 Increase/Decrease 14%-5%4%12% Mid-Peak $/kWh 0.02162 0.02463 0.02331 0.02426 Increase/Decrease 14%-5%4%12% Off-Peak $/kWh 0.02162 0.02463 0.02331 0.02426 Increase/Decrease 14%-5%4%12% Winter Mid-Peak $/kWh 0.02162 0.02463 0.02331 0.02426 Increase/Decrease 14%-5%4%12% Off-Peak $/kWh 0.02162 0.02463 0.02331 0.02426 Increase/Decrease 14%-5%4%12% Demand Charges Facilities Related $/kW 11.64 14.88 16.89 18.34 Increase/Decrease 28% 14%9%58% Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-52 Central Coast Region August 2017 Page 13 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Figure 4 shows the overall rate levels for the three California IOUs for 2015 and the component parts. SCE and SDG&E appear to have about half of their rates attributable to the generation component, with PG&E having more than half, estimated around 60%. Figure 4: From CPUC, 2015 Rate Components for the Three California IOUs Table 4 shows that in 2015 for PG&E, Distribution and Transmission account for approximately 44% of its total Revenue Requirement, in line with SCE at 43% and SDG&E at 44%. Generation accounts for 48% of its Revenue Requirement, in line with SCE at 48% and higher than SDG&E at 40%. Table 4: From CPUC, 2015 Electric IOU Revenue Requirements ($000) Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-53 Central Coast Region August 2017 Page 14 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Figure 5 and Figure 6 show transmission and distribution Revenue Requirements over time, which have been more or less consistently growing for each of the three IOUs since 2005. Figure 5: From CPUC, Trends in Transmission Revenue Requirements for the Three California IOUs Figure 6: From CPUC, Trends in Distribution Revenue Requirements for the Three California IOUs Figure 7 shows the generation Revenue Requirements over time; year 2015 generation Revenue Requirements are lower than 2014 and currently near the 2011 levels. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-54 Central Coast Region August 2017 Page 15 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Figure 7: From CPUC, Trends in Generation Revenue Requirements for the Three California IOUs Assuming PG&E follows the combined trends for the three utilities, this data would indicate that transmission and distribution is making up a larger portion of the total Revenue Requirement for the utility. This would, theoretically, justify a higher fixed component of rates—shifting revenues from generation-related charges to delivery-related charges. On April 14, 2017 Lancaster Choice Energy (LCE) filed a protest against SCE claiming inappropriate shifting of generation related costs into the distribution component, and thus to CCA customers.9 LCE’s filing supports the analysis presented above and the trend of cost shifting to the distribution portion of the electric bill, reducing the margin against which the CCA competes. In addition, Exhibit C provides the results of sensitivity analyses of CCA results against rate escalation relative to the IOUs. 7. FRANCHISE FEE TREATM ENT MRW SUGGESTION We are also concerned that the Draft Study assumes that the franchise fees (i.e., utility taxes) that would flow to the respective cities’ and counties’ general funds if SCE or PG&E were providing service is assumed to instead flow to the CCA. This treatment should be verified by the AWG or corrected. 9 Protest of Lancaster Choice Energy in the Application of Southern California Edison Company (U 338 -E) for Approval of its Proposal to Implement Residential Default Time-Of-Use Rates, Application No. 17-04-015. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-55 Central Coast Region August 2017 Page 16 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 WILLDAN RESPONSE Franchise fees are generally not collected by public power entities. General fund transfers, payments in lieu of taxes, or payments in lieu of franchise fees are typically made by a public power entity. Ultimately, treatment of franchise fees would be a policy decision determined by the participating jurisdictions. Willdan has removed flowback of the franchise fees to the CCA. This change in isolation did not alter feasibility results materially. 8. ADDITIONAL ANALYSES MRW SUGGESTION Lastly, we recommend that sensitivity cases used to explore the impact of lower SCE and PG&E rates and higher exit fees consider a wider range of potential values. WILLDAN RESPONSE The sensitivities and supporting analyses conducted adequately bound the realm of outcomes and exceed the contracted scope of services. MRW RESPONSE TO QUESTIONS The MRW Report answered twelve questions posed by the AWG. Willdan’s responses to this material follow. QUESTION 1: DOES THE STUDY CONSIDER ALL PERTINENT FACTORS TO DETERMI NE CURRENT AND FUTUR E ELECTRIC ENERGY RE QUIREMENTS OF THE CCA? MRW RESPONSE TO QUES TION 1 MRW finds the analyses reasonable. WILLDAN RESPONSE No response required. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-56 Central Coast Region August 2017 Page 17 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 QUESTION 2: DOES THE STUDY INCORPORATE C URRENT POWER MARKET CONDITIONS AND REASO NABLE PROJECTIONS OF EXPEC TED FUTURE CONDITIONS? MRW RESPONSE TO QUES TION 2 Renewable Energy Procurement MRW finds the analysis overestimates the cost of renewable energy and is unable to determine the reasonableness of the Monte Carlo Simulation results. Natural Gas Generation MRW finds the analysis underestimates the cost of natural gas generation and is unable to determine the reasonableness of the Monte Carlo Simulation results. Other Cost Components MRW finds study results reasonable. WILLDAN RESPONSE These items are addressed in other sections of this memorandum. No additional response required. QUESTION 3: ARE THE ESTIMATES OF THE GHG EMISSIONS I NTENSITY OF THE CCA SCENARIOS RELATIVE T O THE INCUMBENT INVE STOR -OWNED UTILITIES (IOUS), NAMELY PACIFIC GAS A ND ELECTRIC COMPANY (PG &E) AND SOUTHERN CALIFORNIA EDISON (S CE), REASONABLE AND ADEQUATE? MRW RESPONSE TO QUES TION 3 MRW finds the analyses reasonable. WILLDAN RESPONSE No response required. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-57 Central Coast Region August 2017 Page 18 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 QUESTION 4: DOES THE DRAFT STUDY CONSIDE R ALL PERTINENT FACT ORS IN PROJECTING FUTURE PG &E AND SCE RATES FOR COMPARISON TO CCA COSTS/PAYMENT/RATE P ROJECTIONS ? MRW RESPONSE TO QUES TION 4 Error in Current IOU Rates MRW identifies an anomaly in load data, based on demand factors, for medium and large commercial and industrial customers for PG&E and SCE. a) IOU Rates Forecasts i. MRW finds that the IOU rate forecast used in the Study is not consistent with a forecast of PG&E rates prepared by MRW in March 2017 for the Contra Costa CCA Feasibility Study that predicts PG&E annual changes as follow: an increase of 1.5% per year for 2017 to 2022; a decrease of 1.5% per year from 2023 to 2025; and annual increases of 5% thereafter. ii. The Draft Study extends its calculated escalator for generation rates to non-generation rates. This is concerning because there is no direct relation between the cost drivers for generation and non-generation utility services. WILLDAN RESPONSE Error in Current IOU Rates The demand level data anomalies resulted from the raw data set used in the load analysis. These anomalies were being researched parallel to MRW’s review. The analysis presented in the final report uses demand proxies to rectify this issue. This issue does not impact load forecasts used in the Study, rather it results from attempting to retro-fit load forecasts into current IOU rate structures. a) IOU Rates Forecasts i. Willdan, lacking access to the underlying data and analysis, cannot verify MRW’s forecast. MRW claims the forecast is based on PG&E’s actual generation resources, however it is not clear what portion of the rate escalation is associated with generation assets that would ultimately be included in the PCIA charge and thus recovered from CCA customers. Some, or all, of the PG&E escalation could appear not in the energy portion of PG&E rates but instead be allocated to the PCIA component, that applies only to CCA customers. The forecast is not consistent with the rate of change in PG&E’s Green Tariff Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-58 Central Coast Region August 2017 Page 19 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Shared Renewables 20-year Rate Forecast—Feb. 2017, page 8 of pdf10–which is the only long-term forecast publicly available. The approach used in the Study is reasonable and consistent. ii. The rate escalation applied to the non-generation portion of rates applies equally to CCA and non-CCA customers and therefore the impact cancels out, having no impact on Study outcomes. QUESTION 5: DOES THE DRAFT STUDY CONSIDE R ALL PERTINENT FACT ORS IN PRESENTING A REASONA BLY ACCURATE INVESTO R -OWNED UTILITY (IOU ) VS. CCA COST/PAYMENT COMPARI SON? MRW RESPONSE TO QUES TION 5 MRW’s concern is that it is not clear that the same delivery rate (and escalation) was used for both IOU and CCA rates. WILLDAN RESPONSE The same delivery rate and escalation was used for both CCA and IOU customers, thus canceling out. QUESTION 6: DO THE P RO FORMA ANALYSES CO NSIDER ALL PERTINENT FACTOR S IN PROJECTING CCA’S OPERATING RESULTS? MRW RESPONSE TO QUES TION 6 Franchise Fees i. MRW believes the Study may be treating franchise fees incorrectly by flowing them back to the CCA. ii. MRW believes the level of SCE franchise fees is incorrect. Power Costs MRW finds it difficult to assess the reasonableness of the Monte Carlo simulation model analyses based on information presented in the report. 10 PG&E Green Tariff Shared Renewables 20 Year Rate Forecast: https://www.pge.com/pge_global/common/pdfs/solar-and-vehicles/options/solar/Forecast.pdf Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-59 Central Coast Region August 2017 Page 20 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Other Operating Costs Salaries and Wages MRW suggests the Study decrease both the number of FTEs and the salary costs. IOU Service Charges MRW suggests the Study decrease the charges below current IOU tariff rates based on the expectation that these charges will decrease or be reduced in the future. ESP Charges MRW concedes that the fee used in the Study is reasonable assuming it includes Scheduling Coordination. Jurisdictional Administration Charges MRW recommends that these costs be removed from CCA operating expenses. Uncollectable Account Charges MRW recommends that these costs be reduced to 0.5% based on rates experienced by operating CCAs. PCIA MRW recommends sensitivity analyses around the level of the PCIA be conducted. Non-Operating Costs MRW takes issue with the Study’s assumptions around contingency funding and financing assumptions. Pro Forma Results and Rate Comparisons MRW concurs that the CCP CCA is infeasible for two reasons: 1) the IOU average rate is lower that the CCA average rate; and 2) the CCA average rate does not cover costs starting in year 2026. MRW cites Contra Costa CCA study results that indicate power costs are 82% of total costs, PCIA charges are 13% and other costs are 6%. MRW claims that other non-power costs comprise 47% of Study costs. WILLDAN RESPONSE Franchise Fees i. The treatment of franchise fees has been revised as discussed in this memorandum under the response to Item No. 7. ii. Based on the tariff applicable to CCAs, SCE’s franchise fees are correct. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-60 Central Coast Region August 2017 Page 21 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 Power Costs Exhibit D provides a memorandum concerning the Monte Carlo simulation prepared for the AWG. Other Operating Costs Salaries and Wages Refer to the response to Item No. 3. IOU Service Charges Refer to the response to Item No. 4. ESP Charges Willdan confirms that the ESP charges include Scheduling Coordination. Jurisdictional Administration Charges These charges are for external CCA coordinators located at member sites or to reimburse members for use of FTEs performing coordination efforts needed to facilitate CCA operations. These charges represent an additional labor requirement for members resulting from creation of the CCA and are not captured elsewhere. Willdan does not concur with removing such costs from CCA operating expenses but also notes that such costs in isolation are immaterial to feasibility Study results. Uncollectable Account Charges Refer to the response to Item No. 2. PCIA Refer to the response to Item No. 8. Non-Operating Costs With respect to a contingency/rate stabilization fund, MRW incorrectly asserts that the Study would accumulate $778M in contingency funds by 2030 (refer to Figure 3). Contingency funds are intended to cover unanticipated events. Therefore, the Study prudently includes a contingency amount in yearly budgeted amounts and assumes such funding is used to routinely cover power cost fluctuations and other expenditures in excess of budgeted amounts. It is an erroneous belief that such amounts would accrue in an account over time. Pro Forma Results and Rate Comparisons Willdan finds it difficult to respond to MRW’s cited percentages absent understanding what items are included in cited amounts and the basis of comparison. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-61 Central Coast Region August 2017 Page 22 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 For this Study, power costs represent approximately 70% of operating expenses—97% when adding IOU service charges, the CRS component, and franchise fees—leaving other non-power and non-IOU costs totaling approximately 3% of operating expenses. Given that CCA rates were higher than the IOU rates in the first five years, no further adjustment was made to CCA rates in outer years as the enterprise was deemed infeasible. QUESTION 7: DO YOU HAVE ANY OTHE R SUGGESTIONS FOR RE DUCING CCA COS TS IN LIGHT OF THE EVOL VING CALIFORNIA CCA MARKET PLACE? MRW RESPONSE TO QUES TION 7 MRW’s suggestions appear in its responses to Questions 4, 5, and 6. WILLDAN RESPONSE Refer to Willdan’s responses to Questions 4, 5, and 6. QUESTION 8: DOES THE DRAFT STUDY PRESENT AN ADEQUATE ANALYSIS OF POTENTIAL ECONOMIC B ENEFITS AND CHALLENG ES OF VARIOUS SUPPLY SCENARIOS? AND QUESTION 9: SHOULD A NY ADDITIONAL BENEFI TS OR CHALLENGES BE CONSIDERED? MRW RESPONSE TO QUES TIONS 8 AND 9 MRW believes that the Study failed to model the negative indirect and induced effects canceling out the benefits of local projects. WILLDAN RESPONSE Willdan believes that the entities involved are rational economic actors that would not proceed with an infeasible enterprise and therefore no negative economic impacts would be realized. QUESTION 10: DOES TH E DRAFT STUDY PROVID E A THOROUGH EVALUAT ION OF THE PROSPECTIVE CCA’S ABILITY TO ACHIEVE RATE COMPETITIVENES S WITH PG&E AND SCE? WHAT OTHER FACTORS, IF ANY, SHO ULD BE CONSIDERED? MRW RESPONSE TO QU ESTION 10 MRW suggests additional sensitivities should have been run. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-62 Central Coast Region August 2017 Page 23 Response to MRW Peer Review of “Technical Feasibility Study on Community Choice Aggregation for Central Coast Region” Draft Report Dated May 10, 2017 August 1, 2017 WILLDAN RESPONSE Refer to Willdan’s response to Item No. 8. QUESTION 11: DOES TH E DRAFT STUDY CONSID ER ALL PERTINENT FAC TORS TO ASSESS THE OVERALL C OST -BENEFIT POTENTIA L OF CCA? MRW RESPONSE TO QUESTION 11 MRW has no additional factors to include. WILLDAN RESPONSE No additional response is needed. QUESTION 12: DOES TH E DRAFT STUDY CONSID ER ALL PERTINENT RIS K FACTORS INVOLVED WITH ESTABL ISHMENT AND OPERATIO N OF THE CCA PROGRAM , AND ARE SUCH FACTORS PROPER LY WEIGHTED AND ANAL YZED? MRW RESPONSE TO QUES TION 12 MRW finds the Study addressed all pertinent risk factors. WILLDAN RESPONSE No additional response is needed. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-63 Central Coast Region August 2017 EXHIBIT A Original analysis conducted in May 2017; revised in August 2017 to reflect changes incorporated into the final report. POWER PROCUREMENT CO ST COMPARISON RESULT S At the request of the AWG, all sensitivity analyses considered the AWG Jurisdictions Middle of the Road scenario against changes in key input assumptions, including power procurement costs, staffing costs, and IOU rate escalation. This Exhibit A presents the results of the power procurement cost sensitivity analyses. Table A-1 depicts the difference in average power procurement costs between the AWG Middle of the Road scenario and the 30% decrease in power procurement costs and 40% decrease in power procurement costs sensitivity cases. Table A-1: Average Power Procurement Costs, AWG Jurisdictions - Middle of the Road Scenario, with 30% Decrease in Power Procurement Costs, and with 40% Decrease in Power Procurement Costs AWG Jurisdictions Middle of the Road Scenario Year Original Power Procure ment Cost ($ per MWh) With Power Procure ment Cost 30% Lower ($ per MWh) With Power Procurem ent Cost 40% Lower ($ per MWh) 2020 74.54 52.18 44.72 2021 74.81 52.37 44.89 2022 73.55 51.48 44.13 2023 74.33 52.03 44.60 2024 72.80 50.96 43.68 2025 71.73 50.21 43.04 2026 71.69 50.18 43.01 2027 70.93 49.65 42.56 2028 70.56 49.39 42.34 2029 69.18 48.43 41.51 2030 68.64 48.05 41.18 Table A-2 presents the AWG Middle of the Road scenario average rate comparisons between the CCA and PG&E and SCE over the rate comparison period of 2022 through 2026. Tables A-3 and A-4 present this information for the 30% decrease in power procurement cost and 40% decrease in power procurement cost cases, respectively. As shown in Table A-3, the 30% decrease in power procurement costs results in CCA rate proxies that are still not below both PG&E and SCE. The average rates for the CCA are between 2.93% and 4.51% higher than PG&E and between 7.26% and 8.91% higher, depending on the year. While the premium across the Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-64 Central Coast Region August 2017 Page 25 EXHIBIT A POWER PROCUREMENT COST COMPARISON RESULTS classes between the CCA and the SCE has gone down over the AWG Middle of the Road scenario, shown in Table A-2, the CCA power procurement costs still need to be even lower to be competitive with either IOU. Table A-4 shows that CCA rate proxies become competitive against both PG&E and SCE once power procurement costs are decreased for the CCA by 40%. Compared to PG&E rates, a CCA rate proxy savings (CCA customer pay less) of between 4.34% and 5.79%, results depending on the year. Compared to SCE rates, a CCA rate proxy savings of between 2.01% and 3.50% results. Table A-2: Rate Comparisons, Participation Scenario 2: AWG Jurisdictions - Middle of the Road Scenario Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-65 Central Coast Region August 2017 Page 26 EXHIBIT A POWER PROCUREMENT COST COMPARISON RESULTS Table A-3: Rate Comparisons Participation Scenario 2: AWG Jurisdictions - Middle of the Road, with Power Price Forecast Sensitivity set at -30% Table A-4: Rate Comparisons, Participation Scenario 2: AWG Jurisdictions - Middle of the Road, with Power Price Forecast Sensitivity set at -40% Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-66 Central Coast Region August 2017 Page 27 EXHIBIT A POWER PROCUREMENT COST COMPARISON RESULTS Tables A-5 and A-6 show the operating results for the AWG Middle of the Road scenario and the 40% decrease in power procurement costs sensitivity, respectively. Table A-5: Operating Results, AWG Jurisdictions Middle of the Road Scenario Table A-6: Operating Results, Participation Scenario 2: AWG Jurisdictions - Middle of the Road, with Power Price Forecast Sensitivity set at -40% Overall, financial performance is similar between the cases, with a sustained period of negative net margins Iasting through 2023, followed by a few years of positive net margins (from 2024 to 2026 in the AWG Middle of the Road scenario and 2024 to 2025 in the sensitivity), and then negative net margins for all remaining years of the study period. The net present value of net margins is $108 million in the 40% Year Operating Revenues ($000s) Total Operating Expenses Plus Contingency/ Rate Stabilization Fund ($000s) Non-Operating Revenues/ (Expenses) ($000s) Debt Service ($000s) Net Margin1 ($000s) Working Capital Fund ($000s) Working Capital Target ($000s) Working Capital Surplus/ (Deficiency) ($000s) Working Capital Surplus/ (Deficiency) (%) a b c d a - b + c - d e f e - f (e/f)-1 2020 117,525 150,875 1,235 12,330 (44,445) 223,724 50,583 173,141 342% 2021 472,491 504,655 2,323 12,330 (42,170) 193,883 170,117 23,766 14% 2022 579,072 568,848 2,082 18,499 (6,192) 187,691 192,494 (4,803) -2% 2023 590,222 575,366 2,044 18,499 (1,600) 186,092 194,836 (8,745) -4% 2024 590,817 570,966 1,962 18,499 3,314 189,406 194,067 (4,662) -2% 2025 588,906 566,609 2,098 18,499 5,896 195,302 193,284 2,019 1% 2026 587,918 570,586 2,132 18,499 966 196,268 195,171 1,096 1% 2027 586,991 571,282 2,109 18,499 (681) 195,587 196,227 (640) 0% 2028 586,831 576,506 1,991 18,499 (6,182) 189,405 198,875 (9,470) -5% 2029 584,330 574,978 2,033 18,499 (7,113) 182,292 199,652 (17,361) -9% 2030 582,330 581,643 1,541 18,499 (16,270) 166,022 203,279 (37,257) -18% NPV of Net Margin:(100,693) 1 Net Margin includes Net Operating Income less Debt Service. The net present value (NPV) of the Net Margin is determined using a 4% discount rate and is as of Year 2020. The discount rate is equal to the interest rate on the long-term debt. Year Operating Revenues ($000s) Total Operating Expenses Plus Contingency/ Rate Stabilization Fund ($000s) Non-Operating Revenues/ (Expenses) ($000s) Debt Service ($000s) Net Margin1 ($000s) Working Capital Fund ($000s) Working Capital Target ($000s) Working Capital Surplus/ (Deficiency) ($000s) Working Capital Surplus/ (Deficiency) (%) a b c d a - b + c - d e f e - f (e/f)-1 2020 82,848 105,426 760 8,677 (30,495) 158,236 37,030 121,205 327% 2021 334,087 355,046 1,651 8,677 (27,985) 138,928 125,496 13,432 11% 2022 409,860 402,108 1,493 13,019 (3,774) 135,154 142,754 (7,600) -5% 2023 417,805 407,071 1,470 13,019 (814) 134,340 144,631 (10,291) -7% 2024 418,226 405,861 1,383 13,019 730 135,069 144,811 (9,741) -7% 2025 416,874 404,605 1,477 13,019 727 135,796 144,949 (9,153) -6% 2026 416,175 408,857 1,460 13,019 (4,242) 131,555 146,919 (15,364) -10% 2027 415,518 411,523 1,376 13,019 (7,647) 123,907 148,560 (24,653) -17% 2028 415,405 417,554 1,189 13,019 (13,980) 109,927 151,447 (41,520) -27% 2029 413,635 419,905 1,133 13,019 (18,156) 91,771 153,379 (61,607) -40% 2030 412,220 428,262 527 13,019 (28,534) 63,237 157,508 (94,271) -60% NPV of Net Margin:(107,507) 1 Net Margin includes Net Operating Income less Debt Service. The net present value (NPV) of the Net Margin is determined using a 4% discount rate and is as of Year 2020. The discount rate is equal to the interest rate on the long-term debt. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-67 Central Coast Region August 2017 Page 28 EXHIBIT A POWER PROCUREMENT COST COMPARISON RESULTS decrease in power procurement cost sensitivity versus negative $101 million in the AWG Middle of the Road scenario. In terms of surplus funds available for investment, both cases show the CCA has issues maintaining adequate working capital for all but a few years of the study period. This larger working capital shortage is attributable to several factors including a lowering of debt issuance amount and the decrease in average rate revenue resulting from lower rates which is sustained throughout the study period (debt issuance and rates are both driven lower due to the power procurement costs being lower). Thus, the lowering of available cash and rates at the onset result in negative financial impacts which worsen through time. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-68 Central Coast Region August 2017 EXHIBIT B Original analysis conducted in May 2017; revised in August 2017 to reflect changes incorporated into the final report. DECREASE IN STAFFING COSTS COMPARISON RE SULTS This Exhibit B presents the results of the staffing cost sensitivity analyses. Again, at the request of the AWG, this analysis and all sensitivity analyses considered the AWG Jurisdictions Middle of the Road scenario against changes in key input assumptions. Table B-1 shows the total staffing costs between the AWG Middle of the Road scenario and the 70% decrease in staffing costs case. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-69 Central Coast Region August 2017 Page 30 EXHIBIT B DECREASE IN STAFFING COSTS COMPARISON RESULTS Table B-1: Test Year Staffing Costs, AWG Jurisdictions - Middle of the Road Scenario and with a 70% Decrease in Salary and Benefits Costs Description Number of Positions Salary and Benefits Base Case ($) Salary and Benefits 70% Decrease in Staffing Costs Case ($) Executive Management Positions: General Manager 1 350,868 105,260 Assistant General Manager 1 241,563 72,469 Chief Financial Officer 1 301,680 90,504 Customer Service Manager 1 241,563 72,469 Human Resources Manager 1 241,563 72,469 Attorney 1 334,472 100,342 Total Executive Management Positions:6 1,711,709 513,513 Other/Departmental Management Positions Accounting and Budget Manager 1 163,957 49,187 Rates and Regulatory Affairs Manager 1 226,260 67,878 Customer Information and Billing Manager 1 226,260 67,878 Key Accounts Manager 1 226,260 67,878 DSM Program Manager 1 174,887 52,466 Communications and Public Relations Manager 1 174,887 52,466 Power Supply and Planning Manager 1 213,144 63,943 Information Technology Manager 1 226,260 67,878 Procurement and Contracts Manager 1 163,957 49,187 Total Other/Departmental Management Positions 9 1,795,873 538,762 Analyst, Technical, Engineering Positions Contracts Analyst 1 128,979 38,694 Accounting and Budget Analyst 3 386,938 116,081 Rates and Regulatory Affairs Analyst 0 - - Power Supply Analyst 2 277,633 83,290 DSM Analyst 2 277,633 83,290 Total Analyst, Technical, Engineering Positions 8 1,071,184 321,355 Administrative, Customer Service, and Other Positions Executive Administrative Assistant 3 341,030 102,309 Administrative Assistant 4 314,797 94,439 Customer Service Representative 4 314,797 94,439 Key Account Representative 7 994,671 298,401 Communications Specialist 1 122,421 36,726 IT Specialist 2 244,842 73,453 Human Resources Specialist 1 142,096 42,629 Total Administrative, Customer Service, and Other Positions 22 2,474,654 742,396 Total, All Positions 45 7,053,421 2,116,026 Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-70 Central Coast Region August 2017 Page 31 EXHIBIT B DECREASE IN STAFFING COSTS COMPARISON RESULTS Table B-2 depicts the rate comparisons under the 70% decrease in staffing costs case. Even with this large reduction in staffing costs, the CCA rate proxies under the AWG Middle of the Road scenario are not competitive with PG&E and SCE. Table B-2: Rate Comparisons Participation Scenario 2: AWG Jurisdictions - Middle of the Road, with Staffing Costs Sensitivity set at -70% CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates Agriculture 0.1230 0.0742 0.1230 0.0753 0.1230 0.0749 0.1230 0.0747 0.1230 0.0754 Commercial/Industrial Small <200kW 0.1238 0.1049 0.1238 0.1065 0.1238 0.1059 0.1238 0.1055 0.1238 0.1065 Commercial/Industrial Medium 200<500 kW 0.1245 0.1097 0.1245 0.1113 0.1245 0.1107 0.1245 0.1103 0.1245 0.1114 Commercial/Industrial Large 500<1000 kW 0.1200 0.1107 0.1200 0.1124 0.1200 0.1118 0.1200 0.1114 0.1200 0.1124 Residential 0.1275 0.1003 0.1275 0.1018 0.1275 0.1013 0.1275 0.1009 0.1275 0.1018 Residential CARE 0.1208 0.0936 0.1208 0.0950 0.1208 0.0945 0.1208 0.0941 0.1208 0.0950 Residential Solar Choice 0.1975 0.1265 0.1975 0.1284 0.1975 0.1277 0.1975 0.1272 0.1975 0.1284 Weighted Average 0.1248 0.0961 0.1248 0.0975 0.1248 0.0970 0.1248 0.0967 0.1248 0.0976 CCA Rate Premium/ (CCA Savings)29.84%27.93%28.62%29.08%27.88% Rate Class CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates Agriculture 0.1106 0.0543 0.1106 0.0551 0.1106 0.0548 0.1106 0.0547 0.1106 0.0552 Commercial/Industrial Small <200kW 0.1127 0.0922 0.1127 0.0936 0.1127 0.0931 0.1127 0.0927 0.1127 0.0936 Commercial/Industrial Medium 200<500 kW 0.1120 0.0837 0.1120 0.0850 0.1120 0.0845 0.1120 0.0842 0.1120 0.0850 Commercial/Industrial Large 500<1000 kW 0.1112 0.0777 0.1112 0.0789 0.1112 0.0785 0.1112 0.0782 0.1112 0.0789 Residential 0.1056 0.0712 0.1056 0.0723 0.1056 0.0719 0.1056 0.0716 0.1056 0.0723 Residential CARE 0.0979 0.0635 0.0979 0.0645 0.0979 0.0641 0.0979 0.0639 0.0979 0.0645 Residential Green Tariff 0.1256 0.1127 0.1256 0.1144 0.1256 0.1138 0.1256 0.1134 0.1256 0.1144 Weighted Average 0.1091 0.0776 0.1091 0.0788 0.1091 0.0784 0.1091 0.0781 0.1091 0.0788 CCA Rate Premium/ (CCA Savings)40.46%38.39%39.13%39.63%38.33% 2026 Rate Class 2022 2023 2024 2025 Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-71 Central Coast Region August 2017 EXHIBIT C Original analysis conducted in May 2017; revised in August 2017 to reflect changes incorporated into the final report. ANNUAL ESCALATION OF PG&E AND SCE RATES COMPARISON RESULTS This Exhibit C presents the results of the PG&E and SCE rates escalation sensitivity analyses. Again, at the request of the AWG, this analysis and all sensitivity analyses considered the AWG Jurisdictions Middle of the Road scenario against changes in key input assumptions. Table C-1 depicts the difference in PG&E and SCE generation rate escalation (the same escalation rates are applied to all classes for both IOUs) between the AWG Middle of the Road scenario and the 4.0% increase in annual escalation of PG&E and SCE rates case. Table C-1: IOU Rates Escalation, AWG Jurisdictions - Middle of the Road Scenario and with a 4.0% Increase Year Study’s Assumed Rate Escalation With IOU Rates Escalated at Additional 4.0% 2020 0.00% 4.00% 2021 0.85% 4.85% 2022 -0.49% 3.51% 2023 1.50% 5.50% 2024 -0.53% 3.47% 2025 -0.36% 3.64% 2026 0.94% 4.94% Table C-2 depicts the rate comparison results of the 4.0% increase in annual escalation of PG&E and SCE generation rates case. The increase of 4.0% in IOU generation rate escalation results in CCA rate proxies being more competitive compared to the AWG Middle of the Road scenario (shown in Table A-2). Compared to PG&E, CCA average generation rate proxies are less than PG&E beginning in year 2024; savings continue to increase in years 2025 and 2026. CCA average generation rate proxies still are higher than SCE rates through year 2025, and then become lower than SCE in 2026. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-72 Central Coast Region August 2017 Page 33 EXHIBIT C ANNUAL ESCALATION OF PG&E AND SCE RATES COMPARISON RESULTS Table C-2: Rate Comparisons Participation Scenario 2: AWG Jurisdictions - Middle of the Road, with IOU Rates Escalation Sensitivity set at +4.0% CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates CCA Rates PG&E Rates Agriculture 0.1242 0.0903 0.1242 0.0952 0.1242 0.0985 0.1242 0.1021 0.1242 0.1072 Commercial/Industrial Small <200kW 0.1250 0.1276 0.1250 0.1346 0.1250 0.1393 0.1250 0.1443 0.1250 0.1515 Commercial/Industrial Medium 200<500 kW 0.1257 0.1334 0.1257 0.1408 0.1257 0.1456 0.1257 0.1509 0.1257 0.1584 Commercial/Industrial Large 500<1000 kW 0.1212 0.1347 0.1212 0.1421 0.1212 0.1470 0.1212 0.1524 0.1212 0.1599 Residential 0.1287 0.1220 0.1287 0.1287 0.1287 0.1332 0.1287 0.1380 0.1287 0.1448 Residential CARE 0.1219 0.1138 0.1219 0.1201 0.1219 0.1243 0.1219 0.1288 0.1219 0.1351 Residential Solar Choice 0.1987 0.1539 0.1987 0.1623 0.1987 0.1680 0.1987 0.1741 0.1987 0.1827 Weighted Average 0.1260 0.1169 0.1260 0.1233 0.1260 0.1276 0.1260 0.1323 0.1260 0.1388 CCA Rate Premium/ (CCA Savings)7.74%2.13%-1.30%-4.76%-9.25% Rate Class CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates CCA Rates SCE Rates Agriculture 0.1117 0.0661 0.1117 0.0697 0.1117 0.0721 0.1117 0.0748 0.1117 0.0785 Commercial/Industrial Small <200kW 0.1139 0.1122 0.1139 0.1183 0.1139 0.1224 0.1139 0.1269 0.1139 0.1331 Commercial/Industrial Medium 200<500 kW 0.1132 0.1018 0.1132 0.1074 0.1132 0.1112 0.1132 0.1152 0.1132 0.1209 Commercial/Industrial Large 500<1000 kW 0.1124 0.0946 0.1124 0.0998 0.1124 0.1032 0.1124 0.1070 0.1124 0.1123 Residential 0.1066 0.0866 0.1066 0.0914 0.1066 0.0945 0.1066 0.0980 0.1066 0.1028 Residential CARE 0.0991 0.0773 0.0991 0.0815 0.0991 0.0844 0.0991 0.0874 0.0991 0.0918 Residential Green Tariff 0.1266 0.1371 0.1266 0.1446 0.1266 0.1496 0.1266 0.1551 0.1266 0.1627 Weighted Average 0.1102 0.0944 0.1102 0.0996 0.1102 0.1031 0.1102 0.1068 0.1102 0.1121 CCA Rate Premium/ (CCA Savings)16.63%10.55%6.84%3.09%-1.76% 2026 Rate Class 2022 2023 2024 2025 Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-73 Central Coast Region August 2017 EXHIBIT D POWER PROCUREMENT MO NTE CARLO SIMULATION MODEL QUESTIONS OVERVIEW On Friday, May 19, 2017, EnerNex was sent a detailed inquiry from the Central Coast Power (CCP) Advisory Working Group (AWG) related to the methodology utilized to establish the power procurement cost component of the CCA Feasibility Study (Study) for the Tri-County region of Santa Barbara County, San Luis Obispo County, and Ventura County. The full text of that inquiry is included below along with EnerNex responses and clarifications. EnerNex welcomes any additional questions that may be needed to further clarify the statistical analysis and Monte Carlo simulation model (MCSM) utilized to estimate electricity usage, demand, and power procurement cost for the CCP feasibility study. INQUIRY/RESPONSE AWG PREAMBLE This comment has to do with “Table XXXV. Weekday electricity usage Monte Carlo confidence interval” and the narrative around it (and it is relevant to several other sections). We do not understand how the Monte Carlo simulations are being applied here, and we are confused about the use of confidence interval vs confidence level. Figure ES - XXXV. Weekday electricity usage Monte Carlo confidence intervals. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-74 Central Coast Region August 2017 Page 35 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS Figure ES - XXXVI. Weekend/holiday electricity usage Monte Carlo confidence intervals. ENERNEX PREAMBLE The statistical analysis that provides the basis for the load analysis and cost of power analysis utilize both confidence intervals and statistical based simulations to estimate and forecast future electricity demand and usage as well as power supply cost. The figures/tables in question are an output rather than an input to the analysis. In this case, the figures are depicting a range of possible electricity demand (kW) for each weekday and weekend hour in the year 2020 based on the historical CCA-Info data provided by the CCAs, the load forecast, and a projection of customer owned distributed generation. The figures are intended to illustrate the range of historic variability, as forecasted to future years. The majority of load stays fairly close to the average, but outliers exist on the high and low ends. Managing the cost exposure when serving the high and low extremities of the demand range can be the difference between a successful and unsuccessful power procurement strategy. The depiction of the 95% simulation upper end illustrates that the majority of electricity demand is close to the average and ut ilizing that number for power procurement planning is a conservative approach. However, the power procurement approach embedded in the Monte Carlo model does not utilize power purchase agreements (PPAs) to procure power at this upper end 95% confidence level. Instead, the model procures energy through PPAs to a lower bound 90% confidence level. Confidence levels as low Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-75 Central Coast Region August 2017 Page 36 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS as the 80% lower bound were explored as a PPA procurement strategy with the intent to minimize exposure to California Independent System Operator (CAISO) market volatility – especially when there is a correlation between high renewable supply content and the extreme high and low prices in the CAISO markets. Specific responses to the inquiries regarding the confidence intervals, confidence levels and Monte Carlo simulation follow. AWG INQUIRY 1 • We are working off the following definitions: o Confidence interval: refers to the range of values around the sample value within which we expect the population value to lie. ENERNEX RESPONSE 1 • Correct. The CCA-Info data set was analyzed for every hour of every day of every month (with differentiation between weekdays and weekends) to calculate the average demand and standard deviation and confidence intervals. This specified range (low end to high end) i s the confidence interval which is expressed in percentages. Put another way, with a 95% confidence interval, there is a 95% statistical probability that the average price within a given hour is between the low end of the range and the high end of the range based on historical sample data. AWG INQUIRY 2 • Confidence level: refers to the degree of certainty (or probability, allows us to claim significance at certain levels). It indicates how confident we are about the projected value lying within our confidence interval. ENERNEX RESPONSE 2 • Correct. Confidence level refers to the capability of the analysis to produce accurate confidence intervals. Intervals and levels go hand in hand. As utilized in the figure, the confidence level provides an estimate with a 95% probability that the actual result will be at or below the 95% confidence level based on the historical data. • Again, the figures in question are intended to be illustrative with the majority of load staying fairly close to the average, while outliers exist on the high and low ends. AWG INQUIRY 3 • And the following general understanding of MC simulation: o 1. You want to model the output (in our case, cost) based on input (in our case, customer load profile) ENERNEX RESPONSE 3 • The load forecast and power purchase cost forecasts are developed independently. • The forecasts are used as inputs to the MCSM. The output of the MCSM is a total cost of power estimate for related products to serve the estimated future load. o Load Forecast Inputs: Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-76 Central Coast Region August 2017 Page 37 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS ▪ CCA-Info data for historical load • Load forecast includes a statistical Monte Carlo simulation for every hour in the Study to model load variability. ▪ Load forecast based on historical load, EIA data and IOU long term procurement plan ▪ Forecast for continuing consumer adoption of rooftop solar which reduces the load forecast • The intermittency of customer owned rooftop solar is estimated with a statistical Monte Carlo simulation for every hour in the Study to model distributed generation variability. o Power Purchase Cost Inputs ($/MW) ▪ Estimated cost forecast for natural gas generation ▪ Estimated cost forecast for RPS compliant renewable generation ▪ Estimated cost forecast for monthly resource adequacy ▪ Estimated cost forecast for meeting mandated energy storage procurement ▪ Estimated cost of day ahead and real-time CAISO market participation • Power procurement cost forecast for CAISO includes a statistical Monte Carlo simulation using a beta distribution for every hour in the Study to model market volatility. AWG INQUIRY 4 • 2. Given an input (average load) you can model the output (average cost to meet that load) ENERNEX RESPONSE 4 • The Power Purchase Cost is estimated by multiplying the projected load (MW) for each hour by the forecasted supply costs for that hour ($/MWh for energy and $/MW for capacity/resource adequacy). • The Monte Carlo model attempts to simulate the power purchase progression with increasing certainty over shorter timeframes and the intent to minimize exposure to the CAISO wholesale market. o PPAs are utilized to purchase energy to serve the load forecast at the 90% lower bound confidence level – with 90% certainty that at least that amount of energy will be needed. ▪ The 90% confidence level was utilized after a few iterations to minimize CAISO market exposure. o CAISO markets are then utilized to true-up the load forecast and energy supply to meet the day-ahead forecast and the real-time demand. AWG INQUIRY 5 • 3. But you don't know the impact that changes in input have on (model) output. (In our case: how changes in customer load affect cost) ENERNEX RESPONSE 5 • This is exactly the variable that the MCSM is intended to estimate – the electricity demand variability relative to the forecast. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-77 Central Coast Region August 2017 Page 38 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS o The output is the exposure to volatile CAISO wholesale electricity markets: ▪ Purchasing additional energy in CAISO markets when PPAs are short relative to actual (simulated) electricity demand. ▪ Selling excess energy in CAISO markets when PPAs are long relative to actual (simulated) electricity demand. ▪ The Monte Carlo calculates the energy transacted in CAISO as the differential between the amount of energy procured through PPAs and the simulated electricity usage for every hour during the 10 year study timeframe. ▪ The CAISO cost estimate utilizes a beta distribution aligning with the skewed distribution of historical CAISO prices for each hour of each month with differentiation between weekdays and weekends/holidays. AWG INQUIRY 6 • 4. So, what you do is to first compute average load and use this to compute average cost ENERNEX RESPONSE 6 • The procurement costs on a per unit basis is determined by the cost and forecast for: o Natural Gas Generation o RPS Compliant Generation (renewables) o Resource Adequacy o Storage • Rather than using the average load, the Study estimates electricity demand using statistical calculation of confidence intervals and application of confidence levels. o PPAs are then used to procure energy/capacity to meet the lower bound 90 % confidence level of the load forecast. o The “per unit” cost is translated to an overall cost. • The Monte Carlo then estimates the exposure to CAISO prices based on the differential between the load forecast and the simulated actual load for each hour of the 10 year study period. AWG INQUIRY 7 • 5. Then you try to model the variance in the load and use this to compute the variance in the cost - but since you don't have an analytical formula to do this, you use MC simulations ENERNEX RESPONSE 7 • Yes. This is the CAISO market exposure component (See the answer to item 3 above). AWG INQUIRY 8 • 6. You generate hundreds, if not thousands, of model load profiles by assuming a normal distribution around the load mean with the variance that's determined by statistical analysis of the customer load profiles provided. ENERNEX RESPONSE 8 • Yes. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-78 Central Coast Region August 2017 Page 39 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS o For SCE, standardized tariff specific SCE load profiles11 are applied to the CCA-Info electricity usage data within each customer tariff as provided in the CCA-Info data. o For PG&E, the CCA-Info data contained actual hourly or 15 minute interval meter data for each potential CCA customer. The interval data were combined into tariff totals. o The CCA-Info data was then analyzed for any given hour of any given month with differentiation between weekdays and weekends. AWG INQUIRY 9 • 7. For each of these randomly generated load profiles, you generate your model output (in this case, the cost of providing the energy) ENERNEX RESPONSE 9 • The PPA cost and forecast ($/MW) is determined independently for each resource type and applied to the load forecast for every hour of every day for the 10 year study period. • The estimated exposure to CAISO markets is based on the MCSM simulated differential between the energy procured through PPAs (with the lower bound 90% confidence interval hourly load forecast) and the simulated actual demand for a specific hour. AWG INQUIRY 10 • 8. After several thousand random input profiles you will have several thousand different model output costs, but the average of these should be the same as the average given in (2) above. However, what you have learned from all these simulations is the variation in the output (cost) around this average. ENERNEX RESPONSE 10 • A simple average can skew the estimate because of the variance in demand. o For example, in the “All 27 Jurisdiction tri-county region” scenario, the average weekday noontime electricity usage forecasted for August 2020 is 1,434,446 kWh. However, the range of usage includes a minimum of 1,236,638 kWh, a maximum of 1,616,966 kWh and a standard deviation of 67,053. • The confidence level approach was utilized to procure PPAs to serve the lower bound 90% confidence level load of 1,420,550 kWh with an embedded strategy to manage CAISO market exposure in the Monte Carlo simulation. o The average simulated usage for August 2020 on a weekday at noontime is 1,417,853 kWh – a little less than the forecast average as well as the 90% confidence interval lower bound. o The resulting CAISO market exposure for a simulated month of August 2020 at noontime on weekdays is: ▪ $15,364 day-ahead market income from CAISO in the day ahead market to sell excess energy while meeting the day-ahead forecast. 11 SCE Load Profiles http://bit.ly/LoadProfiles Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-79 Central Coast Region August 2017 Page 40 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS ▪ $18,162 real-time market expenditure to purchase energy to meet simulated estimates of actual customer electricity needs. AWG INQUIRY 11 • 9. So now you have a distribution of output values (costs). The 95th percentile of these tells us that we're 95% certain that our costs will lie at or below this value. ENERNEX RESPONSE 11 • Correct. Utilization of the upper bound 95% confidence interval output for both the load forecast and the power procurement forecast adds in some contingency margin for the many factors that are unknown and not modeled for a forward looking estimate. AWG INQUIRY 12 The way that the MC analysis is presented in this report doesn't make sense based on the above understanding. • Firstly, placing a 95% percentile on something where there are only 10 values makes no statistical sense. ENERNEX RESPONSE 12 • For each Monte Carlo run, each of the variables are calculated multiple times: o The weekday noon estimate for August 2020 was calculated 22 times (22 weekdays in that month) for each simulation or 220 times within the 10 run simulation. o The scenarios also provide additional simulation runs to compare the procurement cost estimates for each renewable scenario on a per unit basis: ▪ 8 participation scenarios x 22 occurrences of 5:00 in August x 10 simulation runs per scenario = 1,760 simulations of that hour of each year o Of course additional data would be nice, but the Study utilized the lowest resolution data available including the effort to aggregate all customers’ interval meter data provided by PG&E. AWG INQUIRY 13 • Second, finding the 95th percentile of load curves is not an MC simulation. You actually have the load data. The actual statistical curve can be calculated (including its 95th percentile). ENERNEX RESPONSE 13 • Correct. The confidence interval is calculated using average and standard deviation. o However, the MCSM is the model where the load forecast and cost forecasts are combined even if the specific data is not simulated. EnerNex will update the label of the figures accordingly. • The Monte Carlo simulation sensitivity analysis is performed when simulating the hourly demand and cost based on those same statistics (average and standard deviation): o Load variability; o Distributed generation intermittency; and Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-80 Central Coast Region August 2017 Page 41 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS o CAISO price volatility AWG INQUIRY 14 • Thirdly, the MC simulation should take us from load profile to procurement cost, not from load profile to model load profile. ENERNEX RESPONSE 14 • The MCSM combines data elements into the load forecast by taking the historical load profile, applying a load forecast and then deducting the customer adoption of distributed generation to derive the amount of demand and energy needed in future years. • The Monte Carlo analysis is designed specifically to address one of the fundamental challenges for a load-serving entity: to develop a load forecast and procure energy to serve that load. o The load forecast is straightforward using statistics as you suggest. o Determining the accuracy of that load forecast and related exposure to CAISO market prices requires a Monte Carlo simulation of both the customer demand forecast and the CAISO market prices. AWG INQUIRY 15 • When we look at the green line in the graph, there is only one, which makes it seem like we are talking about a value, not an interval. ENERNEX RESPONSE 15 • Correct. This should be relabeled “95% Confidence Level Upper End” as the graph is illustrating the upper bound of the confidence interval for a 95% probability that the demand will be less than or equal to the 95% confidence level based on the historical data. AWG INQUIRY 16 So, it seems like the graph and supporting narrative is saying that the green line is in the 95th percentile of model runs (again only based on 10 runs, so how is that statistically significant?). If that is true, why would we select such an outlier as our projection? Shouldn’t we use something like the average, or maybe something 1 standard deviation from average to be conservative? In short, I’m very confused by their methodology. I don’t understand how the 95th percentile value was used. Was it used to estimate procurement costs and if so, was the average profile used at all? And – we should be looking at the 95th percentile in procurement costs, not the 95th percentile in load profile, because these are 2 very different things. ENERNEX RESPONSE 16 • The figures are intended to illustrate the range of historic variability projected forward to future years. The majority of load stays fairly close to the average, and that outliers exist on the high and low ends. The figures are an output rather than an input for the procurement strategy within the MCSM. • The actual procurement strategy embedded within the MCSM is intended to model how power procurement progresses with long term, near term, day-ahead and real-time timeframes. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-81 Central Coast Region August 2017 Page 42 EXHIBIT D POWER PROCUREMENT MONTE CARLO SIMULATION MODEL QUESTIONS o As stated, PPAs are utilized to meet the lower bound 90% confidence level. The model is not over-procuring energy relative to the load forecast as implied in the question. AWG INQUIRY 17 And finally, referencing a “95% confidence interval” seems to communicate a false sense of certainty. It sounds like the study implies you are 95% sure that outcomes will fall within a certain range, which suggests the findings are statistically significant. However, aren't you really saying that 95% of model outcomes fall below that line, which is very different and may have important impacts on our outcomes? ENERNEX RESPONSE 17 • Utilization of statistical confidence intervals and confidence levels are intended to account for the variability within both load and cost. Proceeding with a simple average is likely to under estimate the actual cost given the variability in demand illustrated in the figures. For example, with the normal distribution analysis of the load data, there is an equal probability of the demand being higher or lower than the average. Utilizing the upper bound 95% confidence level incorporates some contingency to factor in unknown variables that can impact cost. However, we will provide the full range of MCSM results for the load forecast and power procurement cost including the maximum, minimum and average in the Study in order for the AWG and stakeholders to make an informed decision. • Just like a forecast and statistical analysis for stock prices, past behavior is not necessarily an indicator of future performance. As stated in the Study, risk management is the primary focus of developing a power procurement portfolio. The Study attempts to describe the wide variety of risks associated with power procurement. The only thing that is certain is that the load and cost forecast will not match reality. o However, the Study results are as statistically accurate as possible given supporting data and statistical/Monte Carlo model certainties to estimate a load forecast and simulate the cost of power by including CAISO market exposure. Appendix L: Peer Review and Response Technical Feasibility Study on Community Choice Aggregation L-82 Central Coast Region August 2017 Appendix L Peer Review and Response Technical Feasibility Study Central Coast Region on Community Choice Aggregation August 2017 L-83 4. Response to Extended Peer Review Willdan received the MRW Extended Peer Review appearing in Section 2 of this Appendix L on August 24, 2017. Concerning the Extended Peer Review: Willdan has neither reviewed nor vetted assumptions; was afforded no opportunity to review or question MRW’s methodology; and makes no representations concerning the validity of its results, as related to this Study and outcomes. Further, Willdan has not been provided the MRW-revised version of the Pro Forma Model. Willdan cannot therefore opine as to the reasonableness of MRW’s alternative assumptions nor can Willdan determine the extent to which changes to the Pro Forma Model implemented by MRW impaired functionality or the validity of outputs therefrom. Any reliance upon the results of MRW’s alternative pro forma analysis presented in Section 2 of this Appendix L is neither supported nor endorsed by Willdan.