Retail Rate Options for Small Customers The California Statewide Pricing Pilot Why? … What? … The Future? The utility industry is renowned for its tendency to conduct pilots. In almost 30 years of utility projects I’ve witnessed and participated in pilots to examine rates, metering, load control technologies, communications, marketing campaigns, customer education, information services - and their respective variations. The California Statewide Pricing Pilot covered all of this and more. Like many consultants and some of the wiser utility representatives I’ve worked with over the years – Pilots are frustrating. As the principal consultant to the CEC I helped structure this pilot. I also suggested it was unnecessary. They did it anyway. Spent lots of money (over $20 million), wasted lots of time (over two years and climbing), collected lots of data and printed thousands of pages of nice reports. One thing was different – Action items and decisions. Levy Associates

Two Objectives 1 Summarize the most significant results from the California Statewide Pricing Pilot. 2 Identify how California is interpreting these results and what they mean for regulatory policy and demand response. I have two objectives today. Summary the most significant results. I won’t be reviewing the experimental design, sampling process, price elasticity's, technology evaluations, or each of the many different rate forms. What I can tell you is that this was very likely the most sophisticated, complex and most expensive demand response / rate pilot ever conducted – anywhere. I can also tell you that the results have been vetted and the results are considered representative and statistically significant. Most significantly, I will describe how and why California is now using the results from this pilot to significantly reshape its energy policies and system operating procedures. One qualifier: While the intent is genuine, we are talking about (1) a regulatory environment and (2) California. Use the results accordingly.

Why ? “ Why was there virtually no reduction in electricity peak demand from the customer side of the market when wholesale prices increased by a factor of five in less than a week in the summer of 2000? “ * How do we solve outage management practices that exempt 50% of the utility customers? How do you provide customers with the capability to better manage their electric bills and tailor reliability to their individual needs? What can we do to turn demand response into a viable resource? What motivated California to do a pricing pilot. While I’ve listed four primary reasons on this slide, it really comes down to one key issue: i)        At best DR programs can be best characterized as patches to compensate for poor and/or ineffective rate design. If you examine the four items on this slide, they all have one they all have one common denominator – Pricing. * CEC Action Plan, October 11, 2002

The Problems DR programs can be best characterized as patches to compensate for poor or ineffective rate design. Without basic price information customers do not have the capability to make rational investment or operating decisions. Price Rates Customers don’t understand their electric rates. PRICE – We present customers with rate forms they don’t understand with metrics that don’t reflect price or value. PROGRAMS – Energy efficiency requires technology or structural changes to facilities and appliances, but is ultimately dependent upon customer behavior to make it work or not work. Demand response is dependent upon customer behavior yet we don’t provide the customer with technologies to support and automate their response. Instead, Efficiency measures actually create barriers to Demand Response and program designs further inhibit customer participation. What is needed is a new view of Demand Response, one that takes a customer perspective. Electrical system problems require that technologies and procedures be in place and instantly available – they weren’t. Programs

Statewide Pricing Pilot Results

Critical Design Features Approximately 2,500 participating customers. CPUC, CEC and CPA cooperative regulatory proceeding. Complex experimental design – statistically representative. SCE, PG&E and SDG&E cooperative joint-venture pilot. Revenue neutral rate designs. Rate treatments: (1) TOU, (2) CPP-F, (3) CPP-V. Technology and information treatments.

Conservation and Peak Load Impacts SPP Conclusions System Impacts Residential CPP rates can, within five years of deployment reduce California’s peak load by 1,500 to over 3,000 mW. Conservation and Peak Load Impacts Dynamic rates encourage greater conservation and peak demand impacts than conventional inverted tier or time-of-use rates. Customer Acceptance Residential and small to medium commercial and industrial customers understand and overwhelmingly prefer dynamic rates to existing inverted tier rates. Source: CEC staff conclusions based on review of collective SPP reports.

Percent Change In Peak Period Energy Use (CPP-F) Residential Load Impacts (Weather) Percent Change In Peak Period Energy Use (CPP-F) -13.4 -12.5 -11.0 -16.0 -12.0 -17.1 -9.6 -8.4 1.0 -20 -15 -10 -5 5 Zone 1 Coastal Zone 2 Inland Zone 3 Inland Hot Zone 4 Desert Hottest Statewide Average % Change In kWh Climate Zone CRA Econometric Model CEC Engineering Method Source: Statewide Pricing Pilot, Summer 2003 Impact Analysis, Charles Rivers Associates, August 9, 2004, Table 5-4

Percent Reduction in Peak Period Usage (CPP-F) Residential Load Impacts (Demographics) Percent Reduction in Peak Period Usage (CPP-F) 2 4 6 8 10 14 16 18 20 Percent Reduction 12 High vs. Low User 200% Average Use 50% Average Use 17.2% 9.79% Central AC Ownership YES NO 12.8% 12.3% Pool Ownership 19.2% 12.1% Income > $100,000 < $40,000 15.1% Single vs. Multi-Family Single Family Multi-family 13.5% 9.8% State-wide Average 12.5% Source: Statewide Pricing Pilot, Summer 2003 Impact Analysis, CRA, August 9, 2004, Table 5-9, p.90

Residential Load Impacts (Technology) Critical Peak Impacts By Rate Treatment Hottest Critical Peak Day * Average Critical Peak Day 50% 47.4% Critical Peak Variable With Automated Controls 40% 34.5% Critical Peak Variable With Automated Controls 30% Peak Load Reduction 20% 12.5% Critical Peak Fixed 10% 4.1% TOU 0% Time of Use TOU CPP-F CPP-V CPP-V Source: Statewide Pricing Pilot Summer 2003 Impact Analysis, Charles Rivers Associates, Table 1-3, 1-4, August 9, 2004.

Residential Load Impacts (Incentives) Residential Response with Automation: Participation Incentive vs. Critical Peak Rate 5.0 CPP Event Critical Peak Rate Participation Incentive Control Group 4.5 4.0 3.5 3.0 kW 2.5 2.0 1.5 1.0 0.5 0.0 Noon 2:30 7:30 Midnight Hot Day, August 15, 2003, Average Peak Temperature 88.50

Residential Load Impacts (Historical Results) Two Tier TOU with Dispatched CPP 47.4% Hottest Critical Peak Day * California Pilot 1 2003 Average Critical Peak Day 0% 10% 20% 30% 40% 50% Peak Load Reduction 41.0% Three Tier TOU with Dispatched CPP 35.0% 34.8% 34.5% Three Tier TOU with Dispatched CPP Three Tier TOU with Dispatched CPP Two Tier TOU with Dispatched CPP AEP Pilot 4 1991 Gulf Power Pilot 3 1992-1993 Midwest Pilot 2 2004 California Pilot 1 2003 Source: Statewide Pricing Pilot Summer 2003 Impact Analysis, Charles River Associates, Table 1-3, 1-4, August 9, 2004. Hottest day impacts on page 105. Private communication, residential TOU pilot study, May 2005. Results of the Pilot Residential Advanced Energy Management System, Gulf Power, November 1994. Levy Associates case study report, July 1994.

Small-Medium Commercial Customer Bill Impacts (Actual) Small-Medium Commercial Residential CPPV CPPF TOU Participants (%) 71.1% 73.7% 70.0% Average Monthly Savings (%) 5.1% 5.5% 4.5% Average Monthly Savings ($) $6.81 $3.89 $3.25 CPPV TOU 80.3% 58.2% 12.2% 9.6% $155.17 $90.65 Customers With Bill Savings Participants (%) 28.9% 26.3% 30.0% Average Monthly Increase (%) 4.0% 6.2% 3.0% Average Monthly Increase ($) $5.03 $4.93 $3.32 19.7% 41.8% 5.0% 10.0% $22.89 $62.52 Customers With Bill Increases summer / winter 2003 Source: Statewide Pricing Pilot, Shadow Bill Results, WG3 report, June 9, 2004.

Change in Average Monthly Customer Bill, July 2003 thru May 2004 Customer Bill Impacts (Actual by Utility) Change in Average Monthly Customer Bill, July 2003 thru May 2004 TOU CPP-F CPP-V SCE PG&E SDG&E 2 -2 -4 -6 -8 $0.18 -$4.15 -$5.86 -$2.60 -$2.10 -$1.76 $1.22 -$3.31 Average Monthly Bill Impact <300 300-500 500-750 > 750 Average Monthly Usage (kWh) Source: Response of Residential Customers to Critical Peak Pricing and Time-of-Use Rates during the Summer of 2003, September 13, 2004, CEC Analysis.

Rates Should be Offered to All Residential Customers Residential Rate Preferences Rates Should be Offered to All Residential Customers Rate Group CPP-V CPP-F TOU 88% 27% 20 10 40 50 60 70 61% 62% 57% Percent Responding 30 80 25% 87% 30% Definitely Probably

Original Inverted Tier Rate Customer Rate Preferences (Old vs. New) Original Inverted Tier Rate Pilot Rates Residential CPP-V CPP-F TOU 80% 81% 20% 23% 19% Commercial 77% 71% 70% 20 40 60 80 30% 29% Percent that Prefer Source: SPP End-of-Summer Survey Report, Momentum Market Intelligence, WG3 Report, January 21, 2004, p23-24.

Translating the Results Regulatory Policy and Improved Operating Practices

Demand Response – A Customer Perspective The Customer Perspective Customer Impact Purpose of DR Valuing DR Advance Notice Time Perspective Full Outage Full Outage Cost Total Loss of Service System Protection 0-6 hrs/yr 5 Entire facility outage. None End-Use Curtailment Grid or System Protection 4 Loss of End-Use Seconds or Less Targeted end-use automatic curtailment brief period 2-10 hrs/yr Reliability Responsive DR Expected Value Partial Outage Cost Voluntary Partial End-Use Curtailment Reduced usage, increase temp settings or curtailments Reliability and Economics 3 Some Comfort Impacts Seconds to Hours 20-40 hrs/yr kW Price Responsive DR Shifting or Rescheduling 2 Load Shifting or Rescheduling No Noticeable Impacts Hours to Days 40-100 hrs/yr Economics kWh Basic Service 1 None None Annual years Efficiency and Conservation Customer Facility Envelope / Equipment Control Systems Interface Customer facility, end-uses and operating practices define the infrastructure that form the foundation for all DR and efficiency options. Efficiency and Demand Response are both part of the same continuum, differing only in time perspectives and valuation factors.

1 2 3 Regulatory Proceedings – Policy Initiatives Statewide implementation of advanced metering. 2 Critical Peak Pricing as the default tariff. 3 Programmable controllable thermostats in the Building and Appliance Standards.

Regulatory Proceedings – Policy Initiatives Measure Purpose Statewide implementation of advanced metering. Facilitate better pricing. Support customer education. 1 Integrates efficiency and demand response on a common financial basis. Demand response becomes a condition of service for all customers. Critical Peak Pricing as the default tariff. 2 There are three elements that define a new customer perspective of DEMAND RESPONSE: Advanced metering – a technology that facilitates data gathering and the ability to provide accountability Pricing – rates that reflect cost in a way that is relatively simple yet responsive to the market and/or system conditions Automation technology – to support customer choice, automate behavior and provide system management with reliability measures less onerous than full rotating outages. When you have pricing that reflects cost, you provide the customer with the capability to establish value. The automation technology allows the customer to translate that value into behavior. The automation technology also provides the system operator with the assurance that when the button is pushed, there will be a response. DEMAND RESPONSE then becomes a resource. Enable / automate customer choice. Economic response (CPP day ahead) Reliability response (CPP day of) Enable system protection and redefine outage management. Programmable controllable thermostats in the Building and Appliance Standards. 3