Making Peak Electricity Pricing Make Sense
An important change could make the most common dynamic pricing program much more effective.
It’s starting to cool a bit, but today is going to be hot in the Southwest and California, and with the end of the weekend, electricity demand may strain the grid. To ratchet down demand, one of the prime instruments of US utilities is critical peak pricing (CPP). On designated CPP days – which the utility typically declares the day before – retail prices jump by 2 to 4 times during the critical hours. Our utility, PG&E, refers to them as “Smart Days”; from 4 PM to 9 PM on these days (the most constrained hours) our price goes up from 53 cents per kilowatt-hour to $1.13.
(Source)
CPP has been shown to significantly reduce demand during critical hours and shift it to other hours of the day. It’s a valuable tool for days like today; the day-ahead peak demand forecast for the California ISO grid is over 44 gigawatts (GW), getting near levels that have triggered emergency alerts in the past.
But why was PG&E also calling CPP days back on June 30 and July 1, when the peak forecast was only 37 GW and there was no real risk of grid stress? (The actual demand turned out to be a bit lower.) The wholesale prices were also quite tame.
I’m not complaining about the critical peak price increase; we volunteered for this opt-in program and it gives us a discount during most non-critical-peak hours in the summer. Overall, we save money. I have written favorably about CPP programs and argued for expanding use of them.
(Source)
But I do have a problem with many of the times CPP days are called. In virtually all CPP programs, there is a limit on how many CPP days the utility can call each year, typically 10 to 15. Each CPP day creates a revenue boost for the utility above its additional procurement costs, which it typically offsets by a small price discount in all other hours of the summer. [1] Since the discount is set in advance, the utility wants to use all the CPP days allowed in order to balance its revenues, but it doesn’t want to run out of days and then face a hot spell with no CPP days left to call. And, of course, it doesn’t know what the weather will be in the coming months. Thus, if June is ending and it hasn’t called a CPP day yet, it’s time to call one even if demand will be only slightly above normal.
As I’ve written about before, there is a better way to structure CPP programs that does not create these perverse incentives.
That ceiling on the number of CPP days creates a difficult dynamic allocation problem, which results in the utility sometimes calling them on low demand days – if it looks like they will have “extras” – and at other times failing to call them even when there is a real risk of grid stress – because they are uncomfortably close to running out of CPP days. Both will happen, even if the utility is great at forecasting day-ahead demand, because some years are mild and the utility really doesn’t need 15 CPP days, while other years are harsh and 15 isn’t enough.
Still, because calling a CPP day fills the utility coffers, regulators don’t trust the utility to use their judgment in calling them, so the regulators put a cap on CPP days. A better solution would be to stop limiting the CPP calls a utility can make, but at the same time eliminate the utility’s incentive to call unneeded CPP days.[2]
The way to make that work is to neutralize the utility’s incentive to call CPP days by implementing same-day discounts: when a CPP event is called a day ahead, it would increase the price substantially during the critical peak hours, but it would also decrease the price during all other hours of the CPP day, enough to approximately offset the additional revenue. For instance, if PG&E increases the price to $1.13/kWh during 4 PM-9 PM, it would on the same day reduce the price to 20 cents (or whatever number makes it approximately revenue neutral) in all other hours of the same day.
The greatest advantage of same day discounts is that the utility can call more CPP days if it’s a tough year and fewer if there is less grid stress. No more optimizing how to deploy its (artificially) constrained number of CPP annual calls. Calling more days won’t boost the utility’s bottom line and calling fewer days won’t leave it revenue short.
But there are other important advantages.
A big part of reducing usage during the constrained hours of CPP days is shifting demand to other hours: pre-cooling the house, running appliances at other times, changing EV charging times, and shifting hot water heating. Same-day discounts supercharge the incentive to move away from the expensive hours and towards the discounted hours when supply isn’t constrained.
(Source)
Same-day discounts also reduce bill volatility for customers. With a small discount throughout the summer, a customer’s bill is pushed down slightly in every month, but it can jump substantially if the utility calls many CPP days in one billing period. Targeting discounts at the same days as the CPP premium smooths customers’ monthly bills. And, to the extent that some people are home using air conditioning on a CPP day and others are away, the people who are home, and who shoulder more of the CPP premium price, also get the biggest advantage from the same-day discount for their energy use in other hours.
I have heard the response that calling too many CPP days will create customer dissatisfaction. There would probably be more CPP days in some years under this alternative approach, but there would be fewer in other (probably most) years, just not a relatively constant number every year regardless of the weather. Furthermore, if the utility really does face supply constraints on many days during a brutal summer (or winter), the alternative of risking rotating outages is likely to generate far more dissatisfaction. Equally important, with same-day discounts, the utility won’t be incentivized to waste CPP calls on relatively mild days, which confuses customers and undermines the credibility of the program. It’s probably the reason one PG&E customer I know refers to such events as “Stupid Days”.
One blog post is not a business plan and there are implementation questions that would need to be addressed, some of which I discussed in a 2013 paper. But the reality is that we are unlikely to have most small and medium-sized customers on real-time pricing in the near future, and paying for demand reductions is riddled with issues of cost-effectiveness and manipulation. So if demand is going to play a significant role in meeting increasingly constrained supply conditions, critical peak pricing seems to be our best bet for residential and smaller commercial customers. This practical adjustment would allow us to get a lot more benefit from those programs.
Keep up with Energy Institute blogs, research, and events on Twitter @energyathaas.
I’m off Twitter. But you can find me on Mastodon @severinborenstein@econtwitter.net. And maybe soon on BlueSky. And Threads if they ever add a web interface that’s usable from a PC.
Suggested citation: Borenstein, Severin, “Making Peak Electricity Pricing Make Sense”, Energy Institute Blog, UC Berkeley, July 17, 2023, https://energyathaas.wordpress.com/2023/07/17/making-peak-electricity-pricing-make-sense/
—
[1] The difference between revenue and cost increase on CPP days occurs in part because the wholesale price of electricity typically goes up less than the retail price increase on CPP days, but mostly because the utility has bought most of its electricity on long-term power purchase agreements with fixed prices, so they are not paying the wholesale price for most of the electricity they are delivering on those days.
[2] An alternative, most often suggested by economists, is to set objective and verifiable standards for when a CPP day is called, and require the utility to stick to them. That solves the problem of not trusting the utility’s judgment, but it still means that a large number of calls in a year will create a revenue surplus for the utility and higher bills for the customers on CPP, while a smaller number will do the opposite. Same-day discounts stabilize revenues for the utility and bills for the customers.
Categories
Severin Borenstein View All
Severin Borenstein is Professor of the Graduate School in the Economic Analysis and Policy Group at the Haas School of Business and Faculty Director of the Energy Institute at Haas. He received his A.B. from U.C. Berkeley and Ph.D. in Economics from M.I.T. His research focuses on the economics of renewable energy, economic policies for reducing greenhouse gases, and alternative models of retail electricity pricing. Borenstein is also a research associate of the National Bureau of Economic Research in Cambridge, MA. He served on the Board of Governors of the California Power Exchange from 1997 to 2003. During 1999-2000, he was a member of the California Attorney General's Gasoline Price Task Force. In 2012-13, he served on the Emissions Market Assessment Committee, which advised the California Air Resources Board on the operation of California’s Cap and Trade market for greenhouse gases. In 2014, he was appointed to the California Energy Commission’s Petroleum Market Advisory Committee, which he chaired from 2015 until the Committee was dissolved in 2017. From 2015-2020, he served on the Advisory Council of the Bay Area Air Quality Management District. Since 2019, he has been a member of the Governing Board of the California Independent System Operator.
Energy Institute Blog 
Seems like testing whether customer dissatisfaction rises significantly with more calls should be tested empirically rather than trying to speculate about whether it will happen. We certainly are aware of demand response fatigue, so it’s likely to occur with more CPP calls. The issue is figure out what might be the sweet spot for the number of calls.
CPP clearly needs a redo. One issue not mentioned is that the current approach does not play well with battery storage. Those who invested in it did so, at least in part, to protect against power outages. A (real) CPP day is one where there is a higher probability of a power outage and thus the rational consumer would, in anticipation, put their system into “storm watch” mode or at least up their reserve fraction to be quite high. Since there is no special incentive for battery owners to change their behavior a CPP event, they would take a conservative approach. It would make sense to incent them to sell power during this period, but that is not currently done.
Of course, CPP events are going to get worse in the near-term if near-term programs of fuel-switching from gas to electric are pushed on end-uses such as cooking and water-heating.
The idea to reduce off-peak rates on critical days is probably a mistake.
The “critical” periods of the future will be multi-day dunkelflaute (dark doldrums) where there are atmospheric conditions that result in low solar and low wind. On those days, we will be drafting power from storage (hydro is our main source of storage today, but batteries will catch up) and a lower price in the off-peak hours will compound the problem.
Electricite de France had a wonderful rate design, Tempo, for many years. Tempo may have been the ORIGINAL critical peak price rate; it was limited to larger customers, and known as the “mansion rate.” It had three kinds of days (red, white, and blue), and an on-peak and off-peak price for each. On the red (critical) days, BOTH the on-peak AND the off-peak prices went up sharply. That is the kind of rate we’ll need to manage a wind/water/solar (WWS) future.
The SMUD pilot, in 2014, queried customers on what they actually did to reduce load during critical peak hours. People turned off lights and electronics, changed thermostat settings or turned off the AC, avoided doing laundry and baking, and took other practical steps to shift and shed load. Think about it: if you have six pair of socks and underwear, you have an electricity storage battery in the form of delayed laundry.
The solution of “wasting” critical peak days is pretty simple:
1) Make the days AUTOMATICALLY trip at defined day-ahead or real-time prices. Don’t limit the number of days.
2) Run the revenue through a revenue regulation (decoupling) model or other adjustment mechanism, so the utility has no profit incentive up or down.
3) Replace the for-profit utility entirely with a community-owned utility like SMUD, Austin, or Seattle.
The ad hoc definition of “smart days” clearly undermines their value to the point where alternatives should be seriously considered. No one knows in advance how many days each year are going to stress the grid to the breaking point, and $1.13/kWh, set in advance, is not likely related to the actual marginal cost of electricity at the time or to the price that will incentivize load-shifting. These two dimensions of the program are clearly arbitrary and unnecessary. The purpose of CPP should not be enhancing profits of incumbent private monopolists, but in promoting load shifting by consumers tied to real-time conditions on the grid. To that end, CPP prices should be a tool that can be used without an arbitrary limit of days/year or a predetermined price in $/kWh. Surplus revenues can be returned to consumers in the form of fixed monthly credits after the fact to avoid affecting consumption. Finally, software already exists to inform consumers of CPP conditions and provide remote control of HVAC systems. Some problems are indeed “rocket surgery”, but not this one.
An important correction to this blog–the California utilities don’t make higher “profits” when they call more CPP days. The revenues go into balancing accounts where they are redistributed back to other ratepayers in various proceedings based on allocation formulas defined in each General Rate Case. The utilities call more of these in part because it’s politically desirable to shift costs from the general residential ratepayer class to the customers on the CPP rate. The fix to this incentive is to provide a hedging benefit discount to the bills of CPP customers so that those customers are reimbursed by other customers. Ratepayer advocates have resisted that approach.
CPP is one of the many “use it or lose it” incentives that pervade regulatory and budgetary practices. It is not realistic to think that all “perverse” incentives can be eliminated, but they can be minimized. If there’s not a proceeding at the CPUC to address the CPP issues Severin has identified, there surely should be.