Copyright © 2014 The Brattle Group, Inc. A Path Forward for Residential Demand Charges Ryan Hledik Austin, TX November 10, 2015 PRESENTED BY 2015 NASUCA Annual Meeting

| brattle.com1 What explains the new interest in demand charges for residential customers? Existing two-part tariffs do not reflect the underlying cost structure ▀ Fixed service charge ($/month) ▀ Non-time-varying energy charge (cents/kWh) Costs that vary with system peak-coincident demand, a customer’s maximum demand, or with time or location, are rolled into volumetric energy charges This is leading to fairness/equity problems that are likely to become more pronounced in the future Advanced metering infrastructure allows demand charges to be offered without incremental metering costs

| brattle.com2 How would rates change with a demand charge? ▀ The introduction of a demand charge may be coupled with other modifications to the rate’s design ▀ Introducing a demand charge requires the ability to measure demand; nearly half of U.S. households already have a smart meter New Three-Part Rate One illustration among many possible variations:

| brattle.com3 Two important clarifying comments about demand charges… A demand charge is not a fixed service charge ▀ A demand charge can be avoided through load management ▀ It does not automatically increase bills for small customers There are many ways to design a demand charge ▀ Customer’s maximum demand during month ▀ Max demand during peak hours of day (e.g. 2 pm to 6 pm) ▀ Demand during actual hour(s) of system peak ▀ Average of customer’s highest X demand hours of month ▀ Average over interval of 15, 30, 60, even 120 minutes ▀ Etc.

| brattle.com4 Demand charges do not automatically increase bills for small customers With Increased Fixed ChargeWith New Demand Charge  Correlation between bill impact and customer size is stronger with increased fixed charge  Whether small customers are low income customers is another question entirely…

| brattle.com5 Some utilities in the U.S. already offer residential rates with demand charges ▀ 19 utilities offer residential demand charges, 10 of which are IOUs ▀ They were proposed in Kansas, Nevada, and Illinois and are being considered by other utilities as well Summer Demand Charges in Existing RatesComments

| brattle.com6 There are good reasons to introduce a well-designed demand charge Better align prices and costs Reduce inter-class cross-subsidies Regulatory precedent (i.e., commercial & industrial experience) Incentivize smarter load management Bills do not necessarily increase for small customers Potential bill savings for low income customers

| brattle.com7 We reached out to industry stakeholders to get a range of perspectives on demand charges Interviews with 9 consumer advocates Review of public filings, reports, articles Participation in 10 industry events on the topic over the past year Informal correspondence with ▀ Environmental groups ▀ Solar developers ▀ Commission staff ▀ Utilities ▀ Researchers / academics

| brattle.com8 Common stakeholder concerns about demand charges 1.Demand charges will increase bills for low income customers 2.Residential customers will not understand demand charges 3.They will remove the incentive to invest in energy efficiency and rooftop solar PV 4.They will increase monthly bill volatility 5.Demand charges are not cost-based; TOU is a better option 6.They will require investment in expensive metering and billing infrastructure See forthcoming EEI whitepaper for more detailed discussion

| brattle.com9 New initiatives can address stakeholder concerns Quantify bill impacts, particularly for low- and moderate- income customers Assess customer understanding of demand charges through market research and identify the best way to communicate the concept Assess customer response to demand charges through empirical analysis, pilots, and/or a test-and-learn approach Establish a national conversation on residential demand charges

| brattle.com10 Initiatives to address stakeholder concerns (cont’d) Consider innovative variations on conventional demand charge designs Develop a customer education plan Phase in the rate gradually Develop protections for vulnerable customers The transition will have to be tailored to the unique circumstances of each regulatory jurisdiction

| brattle.com11 Further reading Berg, Sanford and Andreas Savvides, “The Theory of Maximum kW Demand Charges for Electricity,” Energy Economics, October Brown, Toby and Ahmad Faruqui, “Structure of Electricity Distribution Network Tariffs: Recovery of Residual Costs,” Australian Energy Market Commission, August Caves, Douglas and Laurits Christensen, “Econometric Analysis of Residential Time-of- Use Electricity Pricing Experiments,” Journal of Econometrics, Caves, Douglas, Laurits Christensen, and Joseph Herriges, “Modelling Alternative Residential Peak-Load Electricity Rate Structures,” Journal of Econometrics, Crew, Michael and Paul Kleindorfer, Public Utility Economics, St. Martin’s Press, NY, Hledik, Ryan. “Rediscovering Residential Demand Charges,” The Electricity Journal, Volume 27, Issue 7, August–September 2014, Pages 82–96.

| brattle.com12 Further reading (concluded) Schwarz, Peter, “The Estimated Effects on Industry of Time-of-Day Demand and Energy Electricity Prices,” The Journal of Industrial Economics, June Stokke, Andreas, Gerard Doorman, and Torgeir Ericson, “An Analysis of a Demand Charge Electricity Grid Tariff in the Residential Sector,” Discussion Paper 574, Statistics Norway Research Department, January Taylor, Thomas N., “Time-of-Day Pricing with a Demand Charge: Three-Year Results for a Summer Peak,” MSU Public Utilities Papers, Taylor, Thomas and Peter Schwartz, “A Residential Demand Charge: Evidence from the Duke Power Time-of-Day Pricing Experiment,” The Energy Journal, April Yakubovich, Valery, Mark Granovetter, and Patrick McGuire, “Electric Charges: The Social Construction of Rate Systems,” Theory and Society, 2005.

| brattle.com13 Appendix

| brattle.com14 Customers don’t need to be electricity experts to understand a demand charge Responding to a demand charge does not require that the customers know exactly when their maximum demand will occur If customers know to avoid the simultaneous use of electricity-intensive appliances, they could easily reduce their maximum demand without ever knowing when it occurs This simple message should be stressed in customer marketing and outreach initiatives associated with the demand rate Examples from utility websites ▀ APS: “Limit the number of appliances you use at once during on-peak hours” ▀ Georgia Power: “Avoid simultaneous use of major appliances. If you can avoid running appliances at the same time, then your peak demand would be lower. This translates to less demand on Georgia Power Company, and savings for you!

| brattle.com15 Staggering the use of a few key appliances could lead to significant demand reductions ▀ Use of some of the appliances is inflexible (1 kW) ▀ Use of other appliances could be easily staggered to reduce demand ▀ Simply delaying use of the dryer until after the oven, stove, and hand iron had been turned off would reduce the customer’s maximum demand by 3.5 kW ▀ This would bring the customer’s maximum demand down to 5 kW, a roughly 40% reduction in demand Avg. Demand Over 15 min Flexible Load (7.5 kW) Inflexible Load (1 kW) Comments

| brattle.com16 Several tools are available to facilitate the rate transition ▀ Gradually escalating the demand charge over time ▀ Temporary bill protection ▀ Tiered demand charges or ceiling on applicable demand ▀ Shadow bills ▀ Enhanced customer outreach and education ▀ Rebates for enabling technologies ▀ Exemption for vulnerable / low income customers

| brattle.com17 Presenter Information RYAN HLEDIK Principal │ San Francisco, CA Mr. Hledik specializes in the economics of policies and technologies that are focused on the energy consumer. He assists clients confronting complex issues related to the recent slowdown in electricity sales growth and the evolution of utility customers from passive consumers to active managers of their energy needs. Mr. Hledik has supported utilities, policymakers, law firms, technology firms, research organizations, and wholesale market operators in matters related to retail rate design, energy efficiency, demand response, distributed generation, and smart grid investments. He has worked with more than 50 clients across 30 states and seven countries. A frequent presenter on the benefits of smarter energy management, Mr. Hledik has spoken at events throughout the United States, as well as in Brazil, Canada, Korea, Saudi Arabia, and Vietnam. He regularly publishes articles on complex retail electricity issues. Mr. Hledik received his M.S. in Management Science and Engineering from Stanford University, with a concentration in Energy Economics and Policy. He received his B.S. in Applied Science from the University of Pennsylvania, with minors in Economics and Mathematics. Prior to joining The Brattle Group, Mr. Hledik was a research assistant with Stanford University’s Energy Modeling Forum and a research analyst at Charles River Associates. The views expressed in this presentation are strictly those of the presenter(s) and do not necessarily state or reflect the views of The Brattle Group.