Demand Charges to Deal With Net Energy Metering: Key Considerations

Demand Charges to Deal With Net Energy Metering: Key Considerations Amparo Nieto Vice President Presented at EUCI Residential Demand Charges Symposium Calgary, Canada December 1, 2015

Key Rate Design Principles Efficiency (in usage and investment decisions) Equity (inter and intra-class) Cost recovery Rate transparency Predictability 1

Cross-Subsidies in Traditional Standard Rates Residential and Small Commercial electricity rates are overly simplified No daily time-differentiation and two-part structures mean that high-load factor customers subsidize low-load factor customers in the same class Well-designed TOU rates can reduce cross-subsidization although the actual rate structure matters Other cross subsidies (location-based) may remain if rates are not geographically-differentiated 2

Net Metering Exacerbates the Problem Under net metering excess solar output (exports) are compensated at full retail kwh rate (which includes energy plus capacity value) Little or no presence of solar PV generation at the time of system peak means no incremental capacity-related cost savings to the system Intermittent generation may bring grid stability problems, and ramping capacity needs, lacking appropriate smart inverters The unrecovered cost is eventually shifted to nonsolar customers, including low income households 3

An Example of Typical Impact of Rooftop Solar on Load Factor SRP, Summer 2014 1,545 kwh 7.7 kw 7.3 kw 491 kwh Energy Before Solar After Solar Demand The solar customer s total monthly energy usage drops considerably but the maximum demand imposed on the grid barely changes 4

Solar Output Declines as the Utility System Peaks 4 3 Average Residential Load Before and After PV Solar, SMUD July 2014 Summer Peak Period 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24-1 -2 RSGH PV Output Net Metered Usage 5

Equity Considerations With Regard To Solar Customers Ensuring that solar customers continue to pay for their fair share of costs of service Compensation for solar should be based on the utility s opportunity costs (marginal costs) May include supply, delivery, and environmentalrelated opportunity costs May be easier to accomplish setting a separate rate for new net metering customers, with its own revenue target 6

The Method Used to Allocate Costs Matters Key Differences Embedded Costs Top-down approach Allocates test-year costs according to kwh (energy), various measures of CP (capacity and transmission) and NCP (distribution) Energy costs may not be time-differentiated Marginal Cost Basis Bottom-up approach Starts with marginal unit costs of service Final class revenue target relies on information on class price elasticity and relative contribution to overall marginal costs Hourly load profiles determine both energy and demand-related class cost responsibility 7

Traditional ECOS Do Not Provide Design Demand-Related Costs 45% 20% 35% Demand-related distribution costs should actually be split in two types: Coincident Distribution Peak demand Design or Connected customer demand Fixed-Customer Variable Demand 8

Addressing Cross Subsidies with Rate Design A four-part rate structure follows cost causation: Energy peak/off-peak charges: to recover a forecast of energy and operating reserve market prices by time of day and season As metered on-peak demand charge: peak-period marginal capacity costs (G, T, distribution substations, trunkline feeder) Flat per-contract or design KW charge: monthly marginal cost of local distribution facilities (does not vary by season or time of day) * Fixed monthly customer charge: monthly marginal customer-related costs * May be combined in a single fixed charge * The customer charge and facilities charge will need to be scaled up or down relative to MC to hit the class revenue target at the rate setting stage. 9

On-Peak Demand Charges Need to Be Based on Marginal Cost A demand charge based on the peak-period marginal cost provides the right conservation incentive Demand charge < marginal cost, will not lead to sufficient peak shaving Demand charge > marginal costs, may lead to uneconomic bypass (potential battery storage or cogeneration) A Super Peak per-kwh charge targeting the highest cost hours and peak months might also be efficient if it does not lead to peak chasing 10

Example of Demand Charge - SRP s new Net Metering Rate 1 st Block Charge (< = 3 kw) 2 nd Block Charge (> 3-10kW) Tail Block Charge (> 10kW) Marginal Capacity Cost July - Aug $9.59 $17.82 $34.19 $36.96 Efficient price signal - tail block price is close to underlying marginal capacity costs in the core summer months Blocks were needed (temporarily) for revenue neutrality purposes Unblocked demand charges are preferable to reflect full opportunity cost (incremental value) of demand reductions 11

Per-contract or Design Demand Facilities Charge Appropriate to recover the marginal costs of typical local distribution facility configuration for the class A function of design demand (what is the expected maximum demand of the customers to be served over the service life of the transformer, secondary lines, local primary) Likely to vary by customer mix, region, density (rural vs. urban) A per-contract demand charge may be used, but must be reset if at any point customer demand exceeds contract kw 12

Emerging Alternatives to Base Rate Restructuring Buy-All, Sell-All-arrangement ( Value of Solar ) A per-kwh specific solar compensation for all solar output, which may be fixed for up to 20 to 25 years or updated often The VOS rates adopted to date tend to include externality values These are above the direct REC value or the CO2 policy compliance cost savings a form of targeted subsidy Solar Access fee Reflects the share of costs (grid, social programs) that are not avoidable by the solar customer Has its own limitations since it may be based on assumed solar generation and usage profile for a given nameplate capacity 13

Key Conclusions A detailed and comprehensive marginal cost study is key to: (a) first assess the level of cost-shifting in existing rates and (b) inform the appropriate rate structure decisions Use of multi-part electricity tariffs can largely mitigate net metering concerns, improving efficiency and fairness Absent of optimal residential rates, a separate net metering rate may be next best option, unless well designed buy-all, sell-all agreements are adopted 14