11. Demand Response

When considering PG&E's AMI deployment we must examine and adopt a forecast of demand response - reduced energy consumption by customers34 - and we must value that reduction for its contribution to AMI's overall cost effectiveness. As discussed below, we find PG&E presented the most comprehensive and persuasive demand response forecast of 448 MW in 2011 onward (following full-deployment).35 We find PG&E's range of forecasts for total resource cost benefits to be the most persuasive: a range of $510 million with a $52/kW avoided cost in the base case and $338 million in benefits using an $85/kW avoided cost for Scenario 1(e) .36 We note as discussed below that we will rely on PG&E's avoided cost method for the very limited scope of this proceeding, but in no way does our finding prejudging our pending R.04-04-02537 where the Commission will adopt a comprehensive policy and method for determining avoided costs.38 The Commission ordered that it would "... consider any potential revisions to the [adopted interim] methodology in Phase 3 of [the] rulemaking. At that time, we will also consider the potential application of the [adopted interim] methodology to other resource options, such as distributed generation and demand response programs." (D.05-04-024, mimeo, p. 3.) We will do nothing here to otherwise disturb that Rulemaking.

PG&E's expected demand response by 2011, with full deployment of AMI and an aggressive marketing campaign, ranges from 206 to 448 MWs for the proposed CPP rate. This estimate is based on estimated price elasticities of demand for the proposed rates which were derived from the econometric energy demand models that were developed in the SPP research project39, customer participation level, and customer characteristics (i.e., customer consumption and air conditioning saturation in each zone, etc.). The level of customer participation relies on the customer preference market research40 (CPMR) results from the SPP and PG&E's customer population characteristics. The CPMR demonstrated that more customers are likely to sign-up for a time-differentiated rate (CPP rate) if there is a significant opportunity to save money on the rate. The research results also showed that acceptance rates increase as customer awareness increase.

PG&E will conduct focused marketing of the CPP rate to customers with the greatest demand response potential. (Ex 4, pp 2-2.) This is consistent with PG&E's AMI deployment strategy to begin deployment in the hot inland areas which have the greatest demand response potential. PG&E proposes two phases for its communication and marketing strategy. Phase 1 focuses on AMI deployment introducing the concept of time-differentiated rate options, and educating customers about price responsive behaviors. Phase 2 focuses on customer recruitment and marketing of the CPP program. PG&E requests $18 million in funding for phase 1 for the duration of the AMI project deployment.

PG&E proposed a voluntary (opt-in) tariff (for E-1 customers) with a higher rate for CPP periods, a lower rate in non-CPP summer hours, a participation credit for Tiers 3, 4, and 5 in non-CPP summer hours, and first year bill protection - a guarantee that the customer pays no more under the CPP tariff than under the default rate. PG&E also includes in the program an aggressive CPP marketing campaign to entice and educate customers.

PG&E's CPP rate design provides customers an opportunity to save money by making reasonable reductions in consumption during critical peak periods. The demand response estimates by 2011 are based on an assumption that 10% to 35% of residential customers with central air conditioning will participate and 5% of those without air conditioning will participate. (Ex. 4, p. 2-8, Table 2-2, and, attached herein, Table of Demand Response Forecasts and Benefits.) PG&E's estimate also assumes a targeted and aggressive marketing campaign. DRA on the other hand sees these forecasts as overly optimistic and its own optimistic forecast is 30% and its pessimistic forecast is that only 9% will participate. (DRA's Opening Brief, p. 23.)

DRA introduced a study of the experience with a program called "GoodCents" by Florida's Gulf Power. We agree with PG&E's rebuttal testimony that the program is too different to reliably apply to the PG&E situation. For example, GoodCents was focused only the largest-load customers and required that customers have in-home automated energy management systems as well as large electric loads such as pool pumping, electric water and space heating. (Ex. 12-6W, p. 6-2). DRA did not persuade us that the GoodCents program bore a sufficient likeness to PG&E's situation that we should apply any of its experience to this AMI project.

TURN also questioned PG&E's forecasts and proposed significantly lower estimates. TURN asserts that the California Solar Initiative would significantly impact PG&E's targeted reduction of air conditioning load. (Ex. 201, Ch. 3, p. 57.) PG&E responded that the solar installations will not be made by the CPP's targeted population (Transcript p. 306) and TURN applies the full solar target of 176,000 homes in 2001 (AMI's fully-installed date) instead of in 2017 the fully-installed date for solar. TURN compounds this number by annually escalating solar installations after 2011. (Ex. 12. p. 1-5, figure 2-1.) PG&E disagrees with that compounding.

We agree with PG&E that the likely benefits from CPP are different than the solar program benefits. Solar energy tends to displace non-solar generation rather than reduce consumption - it is a form of fuel switching which is comparable to using a hybrid car instead of a gasoline-only car without reducing the miles driven. Here, PG&E forecasts much of the demand response to come from a specific reduction in usage, most especially air conditioning. While it may be true that a customer willing to install a solar device would also tend to be aware of and concerned about their overall energy consumption, the CPP program would still provide them with a direct means to participate in demand reduction in critical peak periods and it would still provide them a rate incentive for their net consumption.41

PG&E persuasively illustrated this demand reduction effect in Ex. 6: in a hot zone, a moderate-usage residential customer with air conditioning who uses 700 kWh in the summer would have 180 kWh in tier 3 (beyond the AB-1X fixed rates) and consumption during a critical peak period would likely range from a low of 21 kWh (3% of all consumption) to a high of 42 kWh (6%). If such a customer reduces its load by 25% during the critical peak period, PG&E's proposed rate design would save the customer as much as $12.72 (13.7% of the bill under the default rates) to $2.64 (2.8%).42 Other examples show that, except for very-high users, customers should generally see a reduced bill. For example, very-high users (1,500 kWh) with 8% of their consumption in a critical peak, and who reduce by 25%, will adversely see a bill increase of $2.80.

34 Demand response impact refers to the change in customer specific peak demand and energy use, by rate period, resulting from time-varying rate.

35 This forecast applies to both the base case and PG&E's scenario 1(e), as discussed elsewhere in this decision.

36 Ex. 4-1S, p. 1-2, revised Table 1-1.

37 See the Rulemaking's December 27, 2005 Scoping Memo: "Recognizing that `[t]he proper valuation of peak load reductions...is needed whether such reductions are achieved through energy efficiency measures, distributed generation or demand response,' [D.05-09-043, p. 141] the Commission directed that consideration of these issues be carefully coordinated and addressed in this generic avoided cost rulemaking. " (Mimeo, p. 2.)

38 Recently in D.05-04-024, dated April 7, 2005 the Commission adopted "... a new avoided cost forecast methodology described in a report prepared by the consulting firm E3. This report, Methodology and Forecast of Long-Term Avoided Cost(s) for the Evaluation of California Energy Efficiency Programs, (E3 report) [footnote omitted.] and associated spreadsheet models, describe and generate 20-year forecasts of (1) hourly wholesale electricity costs, and (2) monthly wholesale natural gas costs. These wholesale energy cost forecasts represent the total avoided cost of power that a utility would otherwise have to generate or procure in the absence of other resource options like energy efficiency programs." (Mimeo, p. 1.)

39 "Impact Evaluation of the California Statewide Pricing Pilot" prepared by Charles River Associates, filed on October 31, 2005, by PG&E, in R. 02-06-001. This report is received into evidence pursuant to Rule 72 and we waive the requirement to file an additional copy in this proceeding.

40 Customer Preferences Market Research (CPMR): A Market Assessment of Time-Differentiated Rates Among Residential Customers in California, Momentum Market Intelligence, December 2003.

41 Residential customers with solar installations are billed at tariff rates for their net consumption - if they produce 2kWh and consume 5kWh, the bill would reflect the net 3kWh.

42 Ex. 6, p. 1-11. This is illustrated at PG&E's rate at the time testimony was filed. Actual results on current rates would be slightly different.

Previous PageTop Of PageNext PageGo To First Page