Smart Non-Residential Rate Design; Optimizing Rates for Equity, Integration, and DER Deployment
Carl Linvill, PhD, Jim Lazar, Max Dupuy, Jessica Shipley, and Donna Brutkoski
Executive Summary
According to the Energy Information Administration, electricity use by non-residential customers accounts for nearly 66% of California’s total consumption. Many of these customers are interested in adopting distributed energy resource (DER) technologies and many have access to sophisticated energy management and load control technologies, which means that these customers can be an important grid support resource. All utility customers stand to benefit if non-residential customers support a reliable, clean, and least-cost grid.
Current non-residential rate design, however, does not adequately encourage the deployment and use of non-residential customer resources in support of grid needs. Instead, current rate design encourages customers to control their own bills without synchronizing their consumption and production with the situation on the grid. Getting rate design right will ensure that price signals conveyed to the customer reflect what the power system needs. In other words, non-residential customer resources will become an important resource for integrating renewables and ensuring grid reliability. Well-designed price signals will induce cost-effective use of energy efficiency, selfgeneration, and demand response for the benefit of both the non-residential customer and all customers.
Effective price signals will increase supply, decrease demand, and thus decrease market clearing prices for energy, capacity, and services.
Many California businesses, educational institutions, and city and county governments have commitments to the state’s decarbonization goals, some have made commitments that go beyond state-level mandates. With well-designed rates, these leaders will have an economic incentive to make private investments that serve the public interest. These non-residential customers could become a large and beneficial contributor to least-cost, reliable decarbonization in California, but aligning their private choices with the public interest requires rate design reform.
Ascending clean energy technologies and aggressive California policy are changing the power system from one where we focused on ensuring adequate supply to meet anticipated demand, to one where active supply and active demand are optimized to ensure balance. Smart grid innovations allow utilities and customers to make more granular decisions about their energy use, while new storage technologies (including thermal storage) offer unprecedented opportunities to absorb variable energy production and shift usage. Wholesale markets are increasingly open to DER aggregators, introducing new value streams to customers who invest in DERs. Load control technologies and new end uses for electricity, especially, add new opportunities for system flexibility. For the past 125 years, the electricity industry has focused on controlling resources to match varying loads. In this new landscape, the challenge is increasingly to ensure that the power system is able to use demand and supply resources together to ensure reliability at least cost.
Rate design needs to embrace these changes—ensuring that customers have incentives to shift or control load and DER production when it benefits the system. Time-varying pricing (TVP) rate designs are necessary to better align private choice with the public interest.1 Dynamic pricing options such as critical peak pricing (CPP) further refine price signals and are easy for customers to understand. More complicated dynamic rates like real-time pricing (RTP) can further refine price signals but require more sophisticated energy management, so are likely to be of interest to those organizations that have or hire sophisticated energy managers.
California’s existing rate design evolved over decades, and transmission and distribution rates in particular have not been generally been updated to reflect the profound changes in customer loads, metering technology, and DER technologies. California is making changes like adapting the timeof-use (TOU) peak periods to match solar impacts, establishing default TOU rates, and encouraging movement toward coincident demand rates from non-coincident demand rates. Despite these interesting steps forward, California non-residential rate design has room for further improvement.
RAP’s Smart Rate Design for a Smart Future2 undertook an extensive discussion of residential and small commercial rate design, and identified three principles that should, in our opinion, apply to all customer classes:
• Principle 1: A customer should be able to connect to the grid for no more than the cost of connecting to the grid.
• Principle 2: Customers should pay for grid services and power supply in proportion to how much they use and when they use it.
• Principle 3: Customers who provide services to the grid should be fairly compensated for the value of what they supply.
In this paper, we propose smart non-residential rate principles that build off of these three. We propose:
• Non-Residential (NR) Principle 1: The service drop, metering, and billing costs should be recovered in a customer fixed charge, but the cost of the proximate transformer most directly affected by the non-coincident usage of the customer, along with any dedicated facilities installed specifically to accommodate the customer, should be recovered in a NCP demand charge.
• NR Principle 2.1: De-emphasize NCP demand charges except as noted in NR Principle 1. All shared generation and transmission capacity costs should be reflected in system-wide timevarying rates so that diversity benefits are equitably rewarded.
• NR Principle 2.2: Shift shared distribution network revenue requirements into regional or nodal time-varying rates. This recognizes that some costs are required to provide service at all hours, and that higher costs are incurred to size the system for peak demands.3
• NR Principle 2.3: Consider short-run marginal cost pricing signals and long-run marginal cost pricing signals together in establishing time-varying rates for system resources.
• NR Principle 2.4: Time-varying rates should provide pricing signals that are helpful in aligning controllable load, customer generation, and storage dispatch with electric system needs.
• NR Principle 2.5: Non-residential rate design options should exist that provide all customers with an easy-to-understand default tariff that does not require sophisticated energy management, along with more complex optional tariffs that present more refined price signals but require active management by the customer or the customer’s aggregator.
• NR Principle 2.6: Optimal non-residential rate design will evolve as technology and system operations matures, so opportunities to revisit rate design should occur regularly.
RAP applied these principles to evaluate existing commercial rate designs at each of California’s investor owned utilities. We found that if rate design is not changed to better align with these principles, California will continue to see underinvestment in DER resources and under-utilization of DER resources toward meeting California’s policy goals.
RAP searched for rate design examples that better comport with these principles in California and elsewhere. The non-residential rate design we found that best comports with the principles and elements we have described above is that of the Sacramento Municipal Utility District. SMUD’s non-commercial rate has a fixed charge to recovery customer-specific costs of billing, collection, and customer service; a site infrastructure cost ($/kW) to recover location-specific capacity costs; a super-peak demand charge ($/kW) to recover marginal T&D capacity costs associated with oversizing the system for extreme hours; and a TOU energy cost to recover all generation costs and remaining T&D costs. SMUD’s rate sets it apart as an industry pace-setter, but we believe their rate design can be improved further.
One important goal for revision of non-residential rate design should be to better adapt to the incorporation of customer resources, such as thermal or electrical storage, customer provision of ancillary services through smart inverters, and customer load control for peak load management. The general framework of the rate design we propose directly compensates many of these through simple, clear, and compensatory TOU rate elements:
This design is generally similar to SMUD’s, with three important differences. First, it is unbundled between generation, transmission, and distribution to enable more granular application. Second, rather than have a super-on-peak demand charge, those costs are reflected in a critical peak price for up to 50 hours per year. The amount recovered is similar to that for SMUD’s super-peak demand charge, but converted to an hourly rate to directly track high-cost hours and to enable better customer response as system conditions change. Third, we have introduced a super off-peak rate, consistent with the recommendation of CAISO. We have intentionally left the definition of time periods unstated, as these will be specific to particular utilities and particular nodes within each service territory, and will change over time as loads and resources evolve.
RAP also reviewed a number of real time pricing tariffs and, while we did not identify one in particular that we would classify as best practice, we did identify lessons learned from Texas, Illinois, Georgia, and Maryland that will be useful to the CPUC as it considers RTP optional tariffs. We suggest designing an RTP option that builds from our TOU plus CPP recommendation, and propose the following simple initial design:
• A wholesale energy cost component, charged on a per kWh basis, that fluctuates hourly. This would be based on the relevant CAISO zonal locational marginal price and would replace the “production cost” component of our recommendation above.
• Transmission costs and distribution costs would be collected in the same way that they are collected under our recommendation above, as would any generation capacity costs that aren’t accounted for in wholesale rates.
Note that this design would not achieve the full benefits of an ideal RTP approach. In particular, this would not include comprehensive price signals reflecting conditions on the local distribution network. Instead, the hourly pricing innovation here is increased exposure of end users to existing CAISO wholesale prices. Over time, as California introduces new approaches that animate the value stack for resources at the distribution level, new rate designs will be able to incorporate more complex and comprehensive RTP components…
Introduction…Rate Design Foundations, Ascending Technologies, and California Policy…Principles for Smart Non-Residential Rate Design…Non-Residential Rate Design in California Today…What Can We Learn from Others About Effective Rate Design…Concluding Recommendations…Appendix A: Some Important Rate Design History…Appendix B: Traditional Cost of Service Methods and Their Application to Rate Design…