Dynamic pricing: the facts are in, Part II

Data over perception = billions of $$ in savings

Phil Carson | Aug 06, 2012


Say you're a regulator and dynamic pricing conjures fears of high prices on the consumers you're charged with protecting and treating fairly.

That topic and emotion might well have been on your mind in Portland two weeks ago when the National Association of Regulatory Utility Commissioners (NARUC) met for summer committee meetings. Enter Ahmad Faruqui, a principal at The Brattle Group and pricing expert, who addressed the gathering on the topic of dynamic pricing and advocated facts over fears. (Faruqui's presentation to NARUC is here.)

For the initial installment in this three-part series, please read "Dynamic Pricing: The Facts Are In," which ran yesterday. In that article, Faruqui explained the implications of the power industry's low load factor and how dynamic pricing could improve it. The third installment will run tomorrow. 

Faruqui presented calculations to his NARUC audience that document that the nation's electricity customers collectively are currently over-paying by about $7 billion each year. (That estimate is based on a Federal Energy Regulatory Commission staff estimate that 92 gigawatts would be saved under universal dynamic pricing, plus demand response valued at $75 kW-year.) 

Also consider that customers with flat loads subsidize those with "peaky" loads to the tune of $3 billion each year. And that 80 percent of low-income customers would actually save on their bills under dynamic pricing. 

Those are the facts. And those numbers do not take into account potential changes in customer behavior to actively take advantage of dynamic rates. 

Of course, the burning question remains: "But will customers respond?" 

A decade of experiments (more than 134) have tested dynamic or TOU pricing in the U.S., Canada, Europe and New Zealand, according to Faruqui. He plotted more than half of those studies, which have been published. 

You'll need to examine the data for yourself. But Faruqui found that customers do react to price signals and that the advantage of in-home, enabling technology is measurable but relatively slim. Price is the fundamental driver.

"You can get a lot (of response) with price alone," Faruqui told me in a recent interview. "That is not commonly recognized."

"Some argue that the only way dynamic pricing will work is to provide customers with enabling technology inside the home, that customers cannot respond without a smart thermostat, smart appliance or in-home display," Faruqui added. "The data reveal that you get most of the response based on price alone. But you do get a boost from adding enabling technology."

This is a critical point, he argued, because many regulators and utilities consider enabling technology too expensive for customers to purchase or for the utility to subsidize.  

"People say we have to do `prices to devices,'" Faruqui said. "Those devices remain expensive. So regulators and others say `we can't do it then.' Because enabling technology costs would be on top of smart meter costs."

On the other hand, dynamic pricing cannot be accomplished without smart meters, because their load profile must be measured on an hourly basis to see how much to charge them during the different hours and rates, Faruqui said. 

New billing systems are needed as well, but in most cases, those come with advanced metering infrastructure (AMI) systems, according to Faruqui.

Faruqui even looked back to studies from 30 years ago and concluded that several decades have elapsed, but "consumer behavior is remarkably consistent."

"The upshot is that people in the utility industry have remained skeptical for 30 years about a scientific fact that consumers respond to price," Faruqui said. "Some look at newer experiments and say that's just a `flash in the pan.' It won't last. In reality, we've measured consumer response over three decades and found similar results. It's a robust finding."  

The good news, in this view, is that smart meters on residential customer premises will cover half the nation by 2017, he said. Dynamic pricing has been a success in Arizona. Arizona Public Service has 51 percent of its customers on various time-of-use rates and Salt River Project has about 30 percent. And dynamic pricing is expected to rollout in the Mid-Atlantic region with Baltimore Gas & Electric and Pepco Maryland, followed by the Midwest (Illinois) and California. 

"There is some momentum building," Faruqui observed. "But 33 percent of all meters today are smart. How many of the prices are smart? One percent. That's a huge gap. That's the regulatory hesitation, with consumer advocates raising the fear factor. That's causing the momentum to stop. Those perceptions, those fears, carry more weight than they should."

So where's the sweet spot for balancing risk and reward among dynamic rate programs? 

Faruqui plotted eight types of pricing programs, from a flat rate to an inclining block rate, a seasonal rate, TOU, "super peak" TOU (2-3 hour peak), critical peak pricing (CPP), variable peak pricing (VPP) and real-time pricing (RTP). That's the gamut from low risk/low reward to high risk/high reward. (A separate category to itself is peak time rebate (PTR), which carries zero risk and a pays a rebate for load reductions relative to a baseline .) 

Faruqui's conclusion: "CPP represents a good hybrid, a sweet spot between a flat rate and a real-time price."

"Let's aim towards a future in which we change the paradigm," Faruqui argued. "Go from a flat rate, the default rate for 99 percent of the country, to a situation in which critical peak pricing becomes the default rate. It wouldn't be mandatory. A customer could opt-out and select other options. They could pick a flat rate or a real-time rate if they want to be aggressive. But the anchor to these choices rests on critical peak pricing."

The logic: opt-in programs get one-fifth the involvement of opt-out programs.

Whoa! Opt-out over opt-in? Isn't that another major sticking point with regulators? 

Opt-out doesn't imply a mandatory program, Faruqui argued. Opt-out simply establishes CPP as the new default rate, but other options, including a flat rate or real-time prices might be offered as alternatives. You're simply forcing customers to make an active choice and providing them with tools to calculate which choice is best for them. And this can be accomplished in steps, using customer segmentation to find a receptive, initial audience and working outward to other customer segments. 

Tomorrow: incremental steps for regulators and utilities on the path to dynamic pricing. 

Phil Carson
Intelligent Utility Daily



Related Topics


Less Well Off Could be Better Off

A couple of thoughts:

First, THE cause of low and declining load factors is air conditioning use. Daily load factors are typically in the low 80% range.  It's the summer spikes due to air conditioning (or in a few places, winter spikes due to extreme cold) that drive the need for electric generating capacity.

Second, one way to shift demand is by cooling structures at night using air conditioning or ventilation.  This is the same principle upon which adobe homes are built in places like Arizona and New Mexico.  Research shows it can work equally well in California.  Of course, having lower electricity prices at night than during the day would make this strategy a lot more economically attractive.

Third, the whole point behind dynamic pricing is to match up what people pay for electricity with when they use it.  If we didn't have more than 100 studies that show consumer benefits, I'd have to agree with the naysayers, but in fact, study after study (in other words, hard data rather than speculation) shows that consumers not only adapt, they're generally more satisfied.

Dr. Faruqui isn't the only one who has proposed some form of dynamic pricing.  Others have suggested variations that allow consumers to try dynamic pricing out with shadow bills or bill protection or other mechanisms that protect them in case they're adversely affected.

I haven't had an air conditioner since I left the south 30 years ago and for a long time, it was a family irritant (my wife hates the heat).  In all that time, I resented the fact that I was subsidizing residences and businesses who kept their premises absolutely frigid while we used fans at night to stay very comfortable during the day.  So even though I probably can't take any better advantage of dynamic pricing than less well off folks because my load is pretty flat, I'm all for it.  The sooner, the better.

Jack Ellis, Tahoe City, CA

Customers Will Surprise Utilities

Best recent story I heard was from a Texas utility with a residential pilot program.  Residential volunteers were notified a day ahead of the next day's pricing.   

A retired couple on a fixed income made a different choice than the utility expected.  It was going to be a critical peak day and the price during that period was moved to reflect it.  
The couple only had an device signaling price amounts.   They chosen to  handle CPP periods by turning off their electric at the main breaker in their home and spending the afternoon at the movies or the mall. 
The utility was curious how they were saving so much on energy without any in-home-control devices. Needless to say it was much different than expected but worked perfectly well for the couple.  
All those associated with the industry when they hear the story have the same reaction I did.  It was a "Deer In The Headlight" moment.
Richard Pate, Pate & Associates

Is it really worth the trouble?

Wow, that is huge $3 billion and another $7 billion, 10 billion and only 300 million in the US that is a whopping $33.33 each one in my family of three it could a $100/year.  Will that even cover the higher failure rate of smart meters?  I know I want another worry for a potential $33.33 a year.  I suggest running the numbers out over a hundred years to make them really big.


It is nice that everyone is worried about the poor,  I have seen several articles touting no adverse impact on the poor from TOU or CPP. Unlike the statisticians, I have been a poor southerner.  The poor often cannot afford comfort cooling so they have a flatter load profile.  I guess the logic is they cannot afford comfort cooling now, so let us enact rules to keep them that way.  Suffering with the heat is not adverse impact; the poor just need to get accustomed to it. They can buy another Icy with the $3 a month saving they just have to drink it real slow.


Consider the risk profile, a small saving or profit is offered but it is coupled with the large downside of dynamic price spikes risk, in power trading I believe we called that a loser.


Sounds like Dr. Faruqui made a persuasive sales pitch at the regulators conference, but here are few things to consider.  Is it really this simple or is someone trying to sell a bill of goods?  Over simplification is a common aide in justification of large expensive undertakings. In 2000 the flaws in the well intended California market design cost ratepayers and utilities tens of millions and were attributable to oversimplification and premise that the market itself would answer all.  Simple, obvious, easy fixes or improvements in complex systems should send up a red flag.


This does little to shift loads it is peak shaving only, it does not increase off peak loads.  Is the good doctor suggesting this will make a fundamental change in typical on to off peak ratios which is the major contributor to low capacity factor?


The capacity factor was stated as 50 to 60% let’s call it 55% plus the 15% NERC required reserves is 62- 70% committed capacity.  Is CPP or TOU rates going to eliminate the need for reserves?  This still leaves out maintenance, maintenance and maintenance requirements vary by technology, but an 85% EAF (availability) is a reasonable number.  It does look like there is 15-23% of underutilized capacity, which is far stretch from the 45% touted. The potential upside is smaller.  Ignoring maintenance and the reserve requirements definitely makes the justification easier.  What difference will this “Solution” make, how much is it changing the ratio of peak to off peak loads?


What creates price spikes?  Should ratepayers pick up the tab and bare the risk for poor or poorly planned maintenance?  CPP will shift this risk towards ratepayers.  It incentivizes IOUs to use whatever the limit is on critical peak periods as high profit days.  Is that the right price signal?

I liked Faruqui's phrase...

"The plural of anecdote is not 'data.'" 

Regards, Phil Carson

Great Response

You totally corrected all my errors and misconceptions.  I do not understand why you bothered. Is the $33.33 per year per capita incorrect?  Is the ignoring of reserves and maintenance incorrect?  Is my risk profile incorrect?  Is my belief on why the poor's load profiles are flatter incorrect?

You obviously intended to be snide.  I enjoy commenting in this form, but just say the word and I will stop.  

A bit of difficulty agreeing with Farugui's observation

I understand that he used some utility data to do his calculations but I am not so sure about how it applies to most residential customers and a lot would depend upon his definition of low-income.

Here are my not-so-statistically derived thoughts.  I cannot answer for New York, but in the South and especially in warmer months, demand is lowest in the midnight to 0400 hours timeframe.  As people awake and get ready for work, demand climbs and continues its upward trend until it peaks around 1700 or 1800 hours.  It is during that 1700 to 1800 hour time that people have just left or are leaving work with a number of them stopping off at the grocery store, autoparts store or some other retailer to pickup something or other before going home.  The day's temperature just finished peaking around 1500 or 1600 hours so there is plenty of residual heat and evaporation of surface waters has bumped up humidity.  The electricity demand in the retail locations is increasing due to the influx of customers, more than making up for the decreasing demand from the other business sectors, and adding to the cost of business that gets passed on to his customers or bites into his bottom line.  People are arriving home hoping to get a break from the heat and therefore turning down the thermostat or turning on the ac that may have been off all day.

Those are some of the people who will pay for dynamic pricing--the retailers and the working class.  Add to that, the stay-at-home parents with infants and toddlers and a one-worker income.  And also, the elderly retired person or the disabled without a job but enough money to be above the low-income bracket.

The electron going to manufacturing, to a bank or corporate headquarters, to a mom-and-pop retailer, to your local grocery market, or to your house does not really differentiate what generator it came from--a baseload nuclear or coal plant, a highly efficient combined cycle plant, a simple cycle gas turbine, or a solar panel.  The demand at the time is the demand.

Many industrial users have power purchase agreements, which are generally flat rate and still considerably below the cost of residential or business retail rates.  The only way time-of-use pricing will affect demand is to cause the customers, in this case mostly working-class, middle-income families to have to chose between a little relief and relaxation after a work day and paying more for electricity because someone has decided the average working Joe/Jane, or the retiree, or the stay-at-home parent with infants and toddlers, or the ill or disabled home-bound person are the sole cause of power demand peaks rather than the whole system of users being responsible.


How do dynamic pricing and tiers mix?

Many places in the country have their pricing based on "tiers" -- your price is based on your total cumulative electric use for the month. Go over a certain amount, and you get bumped into the next highest tier. This is basically the only price incentive structure available when the utility doesn't have smart meters. Such a cost structure does not really reflect the utilities cost to produce electricity when it's actually consumed.

With smart meters, price can be based on the utilities actual cost to produce the electricity at the time of consumption. I hope utilities don't plan to mix dynamic pricing with tiers. With dynamic pricing, tiers should be eliminated as an artifact of a more primitive technology.

     Milton Scritsmier

     Boulder, CO