Latest Developments in Dynamic Pricing and Advanced Metering Adoption Restructuring Roundtable #99 February 9, 2007

1. Latest Developments in Dynamic Pricing and Advanced Metering AdoptionRestructuring Roundtable #99February 9, 2007 Stephen S. George, Ph.D. Principal Consultant Freeman, Sullivan & Co.

2. There is so much going on, it’s hard to know where to start • Pricing experiments continue to show that customers can and will respond to dynamic prices • Combining enabling technology with price signals significantly increases demand response • Getting customers to try time-varying rates remains a challenge • Information feedback is showing promise • Unfortunately, it’s still true that if you ask two people what AMI is, you’ll get three answers • Open standards (at least half open) are the newest “in” thing • EPAct continues to generate a lot of heat for DR and AMI (but how much light remains to be seen) • PPL takes the plunge into the world of MDM • California continues to build DR infrastructure

3. As the results of each new pricing pilot become available, a common story is developing Price Ratio: 4.1:1 4:1 4.6:1

4. More experimental results are on the way? • Analysis from PSE&G’s pricing pilot should be available within the next few weeks • Xcel Energy’s pricing pilot is complete and a regulatory filing is due in April • The regulatory log-jam that stalled the Pepco (SMPPI) pricing pilot finally broke earlier this month • Will examine CPP and Peak Time Rebates side-by-side • Results are a year away • Similar results are being seen across a wide spectrum of pricing concepts • CPP, RTP and PTR (rebates) There is 30 year’s of research indicating that mass-market customers can and will respond to time-varying price signals—is it time to stop doing experiments and just get on with it?

5. Combining price signals with control technology significantly increases demand response • California’s Statewide Pricing Pilot showed • That peak-period impacts on critical days roughly doubled for residential customers when PCTs were used • For C&I customers with demands less than 20 kW, there was no reduction on critical days without PCTs but a 13% reduction with PCTs • For C&I customers with demands between 20 and 200 kW, there was a 5% reduction without PCTs and a 10% reduction with PCTs • In the Ameren pilot, reductions doubled with PCTs for residential consumers • The Gulf Power program is the “poster child” for enabling technology combined with dynamic pricing • But Gulf Power’s population characteristics differ significantly from consumers in most other parts of the country

6. The real challenge is getting customers to try time-varying prices • Market research and experience indicate that customers are reluctant to sign up for CPP tariffs due to risk aversion • Customers like time-varying rates once they try them, but when considering whether or not to sign up, customers focus more on down-side risk than upside potential • There is no political will to place customers on default dynamic rates • Massachusetts’ examination of this is encouraging and CA keeps talking about it • SDG&E’s proposal for a carrot-only peak time rebate offering addresses the downside risk and may prove popular • UC Energy Institute has proposed a new Incentive Preserving Rebate tariff that addresses psychological barriers to acceptance • PG&E’s AMI application proposed an aggressive marketing campaign to attract 30% of it’s target population for CPP rates • CA’s aggressive approach with enabling technology could help

7. Information feedback is gaining attention • Hydro One’s experiment was the first statistically rigorous study of the impact of real-time information feedback on energy use • More about energy conservation than demand response • Roughly 400 customers monitored over 2.5 years • Provided with usage, dollar and CO2 emissions per hour, total to date and predicted • 6.5% average reduction across all households • Reductions were lowest for households with electric space heating (1.2%) and highest for households with electric water heating but no electric space heating (16.7%) • Salt River Project has conducted a similar pilot

8. What is advanced metering infrastructure? • Advanced metering is a metering system that records customer consumption [and possibly other parameters] hourly or more frequently and that provides for daily or more frequent transmittal of measurements over a communication network to a central collection point. (FERC, Assessment of Demand Response & Advanced Metering, August 2006) • Most people would agree that this definition identifies the minimum functionality of AMI—interval data daily • Others (DRAM, CPUC) include much greater functionality, including two-way communication, remote connect/disconnect, information provision to consumers, and interface with end-use control devices

9. What advanced metering is not! • Mobile AMR ≠ AMI • And a movement to legislate accelerated depreciation for “AMI” that includes mobile AMR for monthly collection of TOU data will hinder demand response penetration • Indeed, any investment in mobile AMR will hinder DR penetration for years to come • Fixed-network AMR ≠ AMI • While defining AMI correctly, FERC’s survey counted every fixed-network AMR system as AMI • Substantial investments and/or additional outsourcing payments would be required to obtain interval data on a large number of customers from many legacy fixed network AMR systems • PPL’s system is the ONLY one that currently gathers hourly data on all customers (once installed, PG&E’s system will do so, as will Edison’s and SDG&E’s if approved)

10. Open standards are all the buzz • OpenAMI, Grid Wise, Intelligrid, SCE’s functional specification, etc. • There are many organizations touting the benefits of open communication protocols between an AMI system and end-use devices • Eventually, this could allow consumers to purchase compatible control and display devices at Home Depot for example and foster competition among suppliers • SCE is requiring AMI suppliers to have Zigbee compatible meter communication capability • Itron and DCSI (and perhaps others) are going to market with offerings that meet this requirement

11. EPAct is generating a lot of heat (but how much light?)

12. PPL takes the next step • PPL is the only utility that currently collects hourly data on all of its customers • But until recently, that data was being discarded • PPL’s AMI system was estimated to be essentially breakeven based solely on operational savings—DR was not directly considered • PPL recently completed initial installation of Nexus’ MDM • Energy Vision—enterprise-facing software that manages usage data in order to improve business processes such as billing, forecasting, settlement, revenue protection and distribution planning • Energy Prism—a customer-facing application that provides customers with more useful information on bills or through web-based interfaces. • PPL has high hopes that the self-service aspects of a web portal will lower costs, improve customer service and foster better consumer decisions

13. Web portals can provide valuable information Links to additional content and tools Peak energy is expensive- Use it Wisely Save money by using less peak energy, especially on Critical Peak days How do I spend my energy dollars during Peak Periods? Take action, especially on Critical Peak days

14. A load shift calculator can help customers decide what actions to take to shift load

15. CA is developing the most DR friendly infrastructure in North America • The CPUC approved PG&E’s AMI application in June 2006 • PVRR of $2.2b to replace 5+ million electric meters (PLC by DCSI) and 4+ million gas meters (RF by Hexagram) • Deployment will be complete by the end of 2010 • SDG&E’s AMI application is being reviewed and a decision is expected this quarter • Settlement discussions currently underway with interveners • SCE’s RFP is currently on the street • The CEC has proposed to modify building standards so that all new construction and all HVAC retrofits will have PCTs • PCTs will have one-way radio or paging capability (with no override) to be used for reliability and a port through which Zigbee or other protocols can be incorporated for use by utilities for economic dispatch

16. At PG&E, operational savings covered 90% of costs; demand response easily covered the gap $460 Demand Response 2500 $2,227 $200 Gap $2.026 2000 1500 Operational Benefits (PVRR) Millions of Dollars Costs (PVRR) 1000 500 0

17. 1000 $360 $741 $300 750 Demand Response Gap Millions of Dollars $441 500 Costs (PVRR) Operational Benefits (PVRR) 250 0 At SDG&E, operational savings only covered 60% of costs. An aggressive DR strategy covered the gap

18. Key Conclusions • Customers can and will respond to time-varying prices • Responsiveness seems to be similar across alternative rate concepts (CPP, RTP, PTR) and across utilities (after adjusting for population characteristics) • Getting customers to try time-varying rates can be difficult • But once they do, they often embrace these options • Risk aversion is a significant barrier that carrot-only approaches and enabling technology can help address • Enabling technology such as programmable communicating thermostats can significantly increase demand response • And may be essential for small C&I customers

19. Key conclusions (continued) • AMI functionality continues to evolve rapidly • Open standards for beyond-the meter control and information technology will help drive down costs and increase penetration • MDM functionality is also evolving • In-home displays, web portals and enhanced bill presentment can improve customer decisions • Demand response is rapidly becoming “the” hot topic at both the state and federal level and a key driver of interest in AMI • The combination of pricing, enabling technology and enhanced information, built on an AMI platform, can produce substantial demand response

20. For more information, contact Dr. Stephen S. George Principal Consultant Freeman, Sullivan & Co. 415 777-0707 StephenGeorge@FSCGroup.com