Smart Meters but Dumb Pricing? Not in Sacramento

Smart meters are supposed to give you more control over your electricity bills. As a PG&E customer, I can log onto a website and look at our household smart meter data, which shows our consumption hour-by-hour. That’s how we learned just how much electricity our porch and kitchen lights were using, until we switched them to LEDs.Smart Meter

Smart meters can also help you avoid consuming when power prices are high—wholesale prices fluctuate throughout the day—and instead run your dryer when power prices are low. But, there’s a hitch: you’re only rewarded for moving your drying around if you’re on a rate plan that reflects some of the wholesale price variation.

Most U.S. households won’t cash in on the benefits of smart meters anytime soon. According to recent statistics, 40% of U.S. utility customers have smart meters. Yet fewer than 1% pay prices that reflect wholesale price variation, high when demand soars, as on hot afternoons in the summer, and low when it slackens.

These facts make no sense in light of the daily ups and downs in wholesale electricity prices. (Roughly, the wholesale price is what power plants are paid for the electricity they produce.) The blue line in the figure below depicts wholesale prices in the Middle Atlantic during a week in Summer 2013. Prices varied from a low near $0/MWh early Saturday morning to a high over $450/MWh on Thursday afternoon.

Roller-Coasting Wholesale Power Prices (in Blue)

Roller-Coasting Wholesale Power Prices (in Blue)

Yet, in the face of these roller-coasting wholesale prices, the vast majority of residential customers in the Middle Atlantic were paying retail rates that average around $160 per MWh no matter what time of day or night they used electricity. This disconnect between wholesale and retail pricing violates what we teach our students in introductory economics—and leads to significant economic losses. Customers are consuming too much power at times of peak demand and too little at off-peak times. As a result, utility companies need to build too many power plants that only produce in peak periods.

A handful of utilities are introducing pricing plans that let customers pay retail prices that vary over time. Their customers can save money by consuming electricity when it’s cheap and conserving when it’s costly. One of these is Sacramento Municipal Utility District (SMUD). Under SMUD’s aptly named “Smart Pricing Plan,” some customers saw prices that varied dramatically by time of day. Some of them, for instance, watched prices spike to $750 per MWh on 12 summer afternoons. The rest of the time, these folks paid roughly $100 MWh, a slightly lower rate than the standard flat rate.

SMUD’s rollout of the plan was also smart.** It implemented a “randomized control trial,” which is the technical term for an experiment in which some randomly selected customers stayed on flat pricing while others had the time-varying option. SMUD could then compare their usage.

Customers on the time-varying plan cut consumption relative to their counterparts on flat rates when the price rose to $750 per MWh – not shocking since that’s a big price hike. Perhaps more interestingly, customers on the time-varying plan cut consumption on afternoons outside the 12 highest-cost days. Maybe these people tuned up their air conditioners or figured out that they should turn up their thermostats while they were out of the house.

SMUD very cleverly took two approaches to introducing customers to the new pricing. In one group, people could opt in, while in the other, they had to opt out (i.e., they were put on the plan unless they took several simple steps to leave). Both groups were randomly chosen, and the customer communications were nearly identical. Customers in the opt-in group got a letter that said, “Sign up today and you could save on your electric bills next summer!” while the letter for opt-outs said, “You’re now on a new pricing plan that can help you save on your summer electricity bills!”

Social scientists have documented many instances of the “default bias”: when confronted by a choice in which one option is viewed as the default, people stick to that option. SMUD’s customers showed this in spades: 95% of them stayed with time-varying pricing when it was the default, but only 18% chose to opt in.

Presumably, the 18% would have stuck with time-varying pricing if that had been their default. The real question is what was happening with the 77% of the people who didn’t opt out and presumably wouldn’t have opted in if given that choice. Maybe they weren’t paying attention to any communication from the utility and didn’t realize they were on the new pricing plan? I was fascinated to see, however, that as a whole the opt-out group reduced consumption by a lot more than the opt-in group during the 12 priciest days. More than double. This suggests that some of the 77% were paying attention to the prices enough to reduce their consumption.

SMUD has decided that time-varying pricing makes sense. It saves the utility money because it doesn’t have to buy as much expensive wholesale power during peak periods. And, it can pass these savings on to customers. It thus has charted a plan to move most customers onto time-varying pricing as the default rate by 2018.

But California’s investor-owned utilities are blocked by recent legislation from introducing default dynamic pricing until 2018. This is foolishness masquerading as consumer protection. Flat-rate pricing creates economic waste and hurts consumers. Plus, as I’ve blogged about earlier, adopting time-varying pricing could help integrate renewables onto the electric grid.

And, as SMUD has shown, even if people aren’t clamoring to sign up, the vast majority of them embrace the option when it’s offered.

If only the legislators in Sacramento had as much foresight as their local utility.

* The title is based on one of the many creative turns of phrase I have heard Bill Hogan use.
** The Department of Energy encouraged recipients of Recovery Act funding to engage in pricing experiments. Nine recipients, representing eleven separate studies, did so. Study descriptions as well as their evaluation reports can be found here. Meredith Fowlie, fellow blogger, and I were part of the advisory group that oversaw SMUD’s study.

About Catherine Wolfram

Catherine Wolfram is the Cora Jane Flood Professor of Business Administration at the Haas School of Business, Co-Director of the Energy Institute at Haas, and a Faculty Director of The E2e Project. Her research analyzes the impact of environmental regulation on energy markets and the effects of electricity industry privatization and restructuring around the world. She is currently implementing several randomized control trials to evaluate energy efficiency programs.
This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

10 Responses to Smart Meters but Dumb Pricing? Not in Sacramento

  1. Jack Ellis, Tahoe City, CA says:

    Too bad legislators aren’t motivated by what’s best for their constituents. TURN often refers to rate plans like SMUD’s as “heat wave pricing”. What they seem to miss is other research by LBL that points the way toward methods for keeping residential customers comfortable on hot days without having to run their air conditioners all the time. The real shame is the extent to which that research still seems to be gathering dust on a shelf.

  2. Robert Borlick says:

    Great article! Well written.

    However, it only focuses on utility rates that are regulated. In much of the country retail customers are free to choose among competitive providers who have the freedom to offer dynamic pricing but have not to any significant extent because the customers are not demanding such rates. So we have a classic chicken-egg conundrum.

    Also, the way to make dynamic rates more acceptable and more effective is to combine them with enabling technology that automatically reduces the customer’s loads in response to the set points the customer programs into the device in advance. (I am not familiar with the SMUD program but I suspect it incorporates enabling technology,) Retail choice suppliers have a disincentive to invest in enabling devices because they have no idea how long the customer will be with them.

    Lastly, I don’t know when you heard Bill Hogan use the term, but he may have borrowed the phrase from an article by former DC PSC Commissioner, Rick Morgan, published in Public Utilities Fortnightly: “Rethinking Dumb Rates,” March 2009.

  3. Karen Herter says:

    While SMUD’s Smart Pricing Options pilot described here didn’t make use of enabling control technologies, its predecessors at SMUD – the Residential (2011-12) and Small Business (2008) Summer Solutions studies – did just as Robert says, providing thermostats that participants could program to automatically respond to the OpenADR initiated price events, with the option to change those settings at any time before or during events.

    Both SMUD and the IOUs are moving toward default TOU by 2018, so it seems they are on the same timeline. SMUD has contributed a dozen or more relevant studies that provide critical insights on how to how to do this well, but don’t forget to also give PG&E credit for slowly-but-surely building a large-scale voluntary dynamic pricing program of their own (SmartRate). If they move forward with OpenADR thermostats for residential and small commercial customers as they’ve said they plan to do, then both SMUD and PG&E will be ahead of the game. I’m not sure where that leaves southern California, but I have faith that they will catch up.

  4. Mark Lively says:

    Smart pricing is very necessary to take advantage of smart meters. But the wholesale price of electricity, as shown in the graph from PJM should be only one portion of the driver for smart pricing. We also need smart pricing of the distribution wires.

    At a smart grid conference sponsored by IEEE and the National Institute of Standards and Technology four years ago, someone wished for the day that all 500 of the participants arrived and parked their electric vehicles in the parking lot. At 12 KW per vehicle, that is 6 MW load on the distribution plant, perhaps an overload, at least until the short haul cars had topped off. There needs to be a way to charge for the strains on the distribution grid, as I subsequently wrote in “Dynamic Pricing: Using Smart Meters to Solve Electric Vehicles Related Distribution Overloads,” Metering International, Issue 3, 2010. More frightening are claims I heard in October 2013 that some distribution feeders in Hawaii have distributed generation that is equal to three times the load on the feeder. The line losses and voltage extremes scream for a market price that moderates the problems, as I wrote in “Dynamic Distribution Grid Pricing”, which is in the draft section of the library on my web page, LivelyUtility.com.

    As for the PJM graph, it clearly demonstrates the problem with time of use prices. The prices had a squared error sum (used for variances) of 733,261. A two period pricing period would reduce the squared error sum by only 38% to 457,695, and that had a Swiss cheese type of on-peak pricing periods. Obviously real time pricing was the only way for pricing to improve economic efficiency.

  5. Pingback: Time-Varying Pricing For Electricity – Glimpse From a Height

  6. I honestly believe that smart meters can only be a positive thing. It gives the consumer the option to make more educated decisions on how to better save money and save energy!

  7. Pingback: Money for Nothing? | Energy Economics Exchange

  8. Pingback: Peak-Time Rebates: Money for Nothing? | My Website

  9. Pingback: Money for Nothing by EU Energy Policy Blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s