Michael Giberson
A study of consumers’ responses to real time pricing, by Hunt Allcott, examines 2003 and 2006 data from the Center for Neighborhood Technology’s Energy Smart Pricing Plan. Allcot observes:
This is a particularly interesting time to be studying real time pricing. Most US households currently have electricity meters that simply record the total consumption of electricity since installation, meaning that the consumer cannot be charged prices that vary from hour to hour. Furthermore, the only way for the electric utility to observe households’ ?consumption is to actually send a worker to read the meter, a costly and potentially error-prone process. The “Smart Grid” is a set of emerging electric power information technologies that include, among other things, household energy management devices and technologies that facilitate communication between electricity retailers and consumers. From the utility?s perspective, improvements in these technologies offer reduced meter reading and administrative costs and the potential for real time metering of electricity use. Furthermore, by allowing households to more easily observe prices and consumption, and even to automate how air conditioners and other appliances turn on and off in response to real time prices, Smart Grid technologies can increase consumers? price elasticity of demand.
Allcot offers three key conclusions from his data analysis: households participating in the program were price elastic, but just a little; the typical response is conservation rather than load shifting; and energy management technology can signi?cantly increase households?’ price elasticity. He calls this final result “fundamental but perhaps unsurprising.” Perhaps unsurprising, but frequently overlooked: how many times have economists, utility executives, or policymakers claimed that power consumers are not responsive to electricity prices based upon observation of consumers lacking both the information and incentive to respond to prices?
Allcot then assessed the estimated welfare affects of real time pricing policies. In general he concluded that efficiency gains from consumer responsiveness to prices were not likely large enough to overcome the costs associated with conservation and metering infrastructure.
In weighing this conclusion, a couple of issues must be considered. As Allcott notes, as the CNT is a voluntary, “opt in” program, we may expect participants to be more likely to be price sensitive than other consumers. Switching all consumers to a real-time price may produce even smaller benefits. On the other hand, larger programs may induce more complementary offerings, enhancing responsiveness. For example, in the CenterPoint and Oncor service areas in Texas, millions of consumers will have smart metering, which may induce, say, local appliance retailers to promote “price aware” appliances and competitive energy retailers to offer real-time pricing contracts. (In fact, as noted by Lynne yesterday, energy retailer Direct Energy is partnering with an energy management systems developer, a electronics retailer, and two appliance makers to offer a consumer-friendly home energy management system. See Lynne’s post for more along these lines.)
Allcot observes in his paper that, prior to his own study, there was “no empirical evidence on how households would respond to hourly real time prices.” (Is this true? Sounds incredible. Surely somewhere … Lynne, do you know of anything?) Whether or not his report on the literature is correct, clearly this is, as Allcot says, “a particularly interesting time to be studying real time pricing.” More study is needed.
As I recall, SDG&E exposed all of their customers to real time pricing just before the CA market went crazy. One of the daftest things CA did was ban hedging, so they had little choice about it. I’m sure you can imagine what sort of newspaper stories started to appear when the price spikes hit.
Is anyone still claiming that consumers won’t respond to prices? The claim that there’s “no empirical evidence on how households would respond to hourly real time prices” might be strictly true, but there’s lots of evidence on residential response to peak pricing and peak hour rebate models.
From what I’ve seen in PJM areas, there’s pretty strong consensus that the combination of understandable pricing and enabling technologies can drive fairly good residental response. I also seem to remember some studies done by LBNL and PBNL that showed similar results. I don’t follow the West coast closely, but I think there have been substantial studies done by CA and WA utilities.
If you’re interested, some fairly successful PJM utility programs reviewed at:
BGE: http://www.pjm.com/~/media/committees-groups/stakeholder-meetings/dr/drs-III/20091109-ami-pilots-panel-bge-program-gulhar.ashx
ComEd: http://www.pjm.com/~/media/committees-groups/stakeholder-meetings/dr/drs-III/20091109-ami-pilots-panel-comed-program-eber.ashx
PepCo: http://www.pjm.com/~/media/committees-groups/stakeholder-meetings/dr/drs-III/20091109-ami-pilots-panel-pepco-pilot-sunderhauf.ashx
FERC’s recent DR study provides some empirical data about price elasticity under various pricing pilot programs. The FERC study’s authors provided further analysis in Public Utilities Fortnightly’s January cover story here:
http://www.fortnightly.com/article.cfm?p_id=395
I’m doubtful that all SDG&E customers were exposed to *real time* prices, but I’m quite sure that they were exposed *at least* to market-based average rates when the Transition Charge ended. Nevertheless, the price response was strong while it lasted. SDG&E rates were re-frozen after some months, and the response dissipated.
I’m a little perplexed by that statement about no empirical evidence. I work for CNT on this project (and have since prior to its beginning), and we had Summit Blue Consulting do evaluations of the program from 2003 through 2006. One of those studies is even cited in the paper. So I’m not sure why those evaluations don’t rise to the level of being studies of “empirical evidence on how households would respond to hourly real time prices.”
There has also been a lot more water under the bridge since the ESPP pilot. So while the data we collected and Hunt analyzed is interesting, it’s arguably quite out of date by now. Since 2007 larger residential RTP programs are now underway at both ComEd and Ameren. They aren’t experiments in the same sense, they are real rates and we don’t try to control variables for research, we let consumers decide what to do (or not do). As a result the current programs have fewer limitations in their rate designs as compared to the ESPP pilot. These rates line up better with actual market conditions.
See http://www.cntenergy.org/real-time-electricity-pricing.php
for the Summit Blue evaluations of ESPP and their more recent work evaluating the Ameren Power Smart Pricing program (and see: http://www.powersmartpricing.org for more info on that program). They evaluated 2008 so far and will have an evaluation of 2009 this spring.
The big issue does remain the metering and infrastructure costs. We are still operating in a situation in Illinois where we don’t have a wide-scale deployment of AMI so meters are replaced as needed as people sign up for RTP. A full scale AMI deployment would make running an RTP program less expensive, but the cost benefit analysis of that deployment is going to have to rely on a lot of other factors than just the value of RTP.
I’m a ComEd consumer. In my town, there is no choice in electric providers, and the IL Energy Commission controls the prices. They offer real time pricing and their website http://www.theWattSpot.com says average customers save 6% on their monthly bill. Last month’s bill shows the energy charge at $.06435/kwh. The additional charges on my bill (for distribution, transmission, environmental cost recovery, energy efficiency programs) plus the flat rate charges for being a customer and metering, plus taxes adds up to $.1286+/kwh. So, the energy cost is less than 1/2 of the bill this month. In addition, the meter needed will be rented to me at $2.25/month. Seems like less of a good deal.
Now, lets assume I am average and save 6%. This month’s bill is $67.52. 6% of that is $4.05. Adding the $2.25 for the meter reduces my benefit to almost nothing. Soon, I plan on getting a new heat pump that will save gobs of electricity. Then, I’ll save a lot less!