This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Smart Meters Need Effective Electricity Pricing to Deliver Their Full Benefits

EDF  Energy Exchange's picture
Blog, Environmental Defense Fund

EDF's energy experts discuss how to accelerate the transition to a clean, low-carbon energy economy. Guided by science and economics, EDF tackles urgent threats with practical solutions. Founded...

  • Member since 2018
  • 1,023 items added with 953,273 views
  • Dec 20, 2014

By Beia Spiller, Economist, and Kristina Mohlin, Economist

walletSmart meters, which provide detailed electricity use data throughout the day, are a critical piece of a smarter, more resilient 21st century energy system. But they are not a cure-all for modernizing our antiquated power grid.

In Matthew Wald’s recent New York Times article, entitled “Power Savings of Smart Meters Prove Slow to Materialize,” he argues that smart meters have failed to produce measurable savings. And we agree – but not because smart meters themselves have failed. Rather, most customers with smart meters don’t have access to people-powered, or time-variant, electricity pricing, which creates opportunities to save money. This is a missed opportunity for customers, utilities, and the environment.

Time-variant pricing better reflects electricity costs

Throughout most of the country, the price paid for residential electricity is the same regardless of the time of day when it’s consumed. This arrangement is a byproduct of an earlier era, one in which electricity information was difficult to convey and the actions of individual customers was impossible to gauge in real time. In practice, electricity is actually dirtier and more expensive to produce and transmit at certain times of the day, particularly when everybody wants it – for example, at 6pm during a heat wave when customers are cooling their homes. Also, during this high-demand time, energy prices spike and electric utilities flip on expensive and dirty fossil fuel “peaker” power plants to meet energy demand. From an economic point of view, it would be more efficient for electricity used at these peak demand times to have a higher price.

Most industries – from automobiles to food –have fully embraced the benefits of varied pricing for premium items or services. The utility industry is one of the few that has not. Charging the same rate for electricity at all hours is like charging the same price for filet mignon and ground chuck. Of course, we can all agree that pricing meat in this manner is inefficient, resulting in an inevitable shortage of filet mignon. Flat electricity rates work the same way, causing customers to over-consume the most expensive, polluting power. The result: higher bills for everybody.

Time-variant pricing attempts to make the true cost of electricity transparent to customers. Without smart meters, utilities have no idea what time of day people are using electricity and are therefore unable to apply time-variant pricing. Since reducing energy demand helps utilities avoid purchasing expensive peak electricity and investing in extra capacity to meet rising demand, smart meters save utilities money as well. Finally, as far as customers are concerned, there is no incentive to change the way they use electricity without time-variant pricing.

Lost opportunity for customers to save on bills

In many states, even though utilities have the ability to offer time-variant pricing, they simply don’t for a number of reasons – among them are legislative barriers and concerns that this type of pricing might negatively affect low-income customers. But, even in states where time-variant pricing is available, adoption is low due to a lack of education and outreach.

The way in which these programs are structured also affects the adoption rates. As we wrote in an earlier blog post, many of the utilities offering time-variant pricing ask customers to opt in to the program, rather than setting it as a default from which customers can opt out. It turns out that this type of choice structure invariably leads to low adoption rates – not just with time-variant pricing, but with a host of other types of decisions, such as organ donation. Even in California where several utilities offer time-variant pricing and smart meters are widespread, very few customers have signed on to these programs.

This is unfortunate, as NYT’s Wald illustrates with two examples, given that customers who have smart meters alongside time-variant rates can indeed save money. He talks to Karen Taubman, an Illinois customer who knocks nearly 20 percent off her bill every month, and Joe Godinsky, also from Illinois, who’s “convinced he’s saving money and convinced his parents to put their house on a real-time rate, too.”

But very few customers in the country are as lucky as Ms. Taubman and Mr. Godinsky. Some might still have standard metering technology, which only measures electricity use for the entire month, leaving them oblivious to the periodic nuances in electricity pricing that could be saving them money. And, even those who have smart meters will have little incentive to respond to the valuable information their meters provide unless their utility companies offer pricing schemes that vary the price of electricity depending on the time of day it is used.

The rise in smart meter deployment is a great opportunity to realize savings both for the utilities and the customers by avoiding expensive and dirty peak time generation, thereby helping to decrease the average cost of each kilowatt hour of electricity sold and reduce harmful pollution. But this opportunity can only be realized by implementing time-variant pricing; smart meters alone won’t be enough.

Engineer- Poet's picture
Engineer- Poet on Dec 20, 2014

The flip side of saving money by scheduling consumption for times of lower market rates is that generators like grid-tied PV owners would have an incentive to produce when rates were high, not when their systems were able to produce the most.  Producing power when nobody needs it helps no one, and today’s system of net metering at flat rates offers no incentive to time-shift generation to actually coincide with demand.

Bill Hannahan's picture
Bill Hannahan on Dec 20, 2014

The most reliable grid would consist of many small reliable dispatchable power plants located close to population centers in an interconnected grid that can easily absorb the occasional loss of a generator or transmission line. There would be no central control system. Components would be designed to automatically detect the failure of adjacent components and to compensate for those failures without external control inputs.

The proposed smart grid is the exact opposite of this approach. It is a massively complex grid with a huge number of components transmitting large flows of energy over great distance, under the control of a very complex interconnected control system. If one were asked to design a grid that would be most vulnerable to cyber attack, it would be the smart grid.

The attack could be designed like the STUXNET virus to inflict maximum damage to key components like massive transformers, switches and capacitor banks. It would take a long time to manufacture and replace those components in large numbers.

A well designed attack could bring down a U.S. smart grid for many days, perhaps weeks. If the attack should come during a severe cold or heat wave the death toll could be in the millions.

We should implement a massive APOLO moon shot like program to develop small modular nuclear plants that are very safe and can be factory mass produced so cheaply that they can be profitable even when load following at modest capacity factors.

Hops Gegangen's picture
Hops Gegangen on Dec 21, 2014


My power provider just put in an iTron OpenWay “smart meter” that supposedly has a ZigBee wireless capability for a home-area network connection, but I can’t find anything that would talk to it, other than to buy the bare-bones electronics and try to hack into it.

The LCD display on the front of the thing just shows the total KWh. It would at least be good for something if it showed the current load.

Engineer- Poet's picture
Engineer- Poet on Dec 21, 2014

something really dangerous

There’s dialectic, and then there’s Greenpeace sound-bite appeals to emotion.  If you were principled you’d use the former instead of the latter. Nuclear power plants are extremely robust (containment buildings are designed to take direct impacts from aircraft) and nuclear has the lowest per-TWh fatality rate of any source of electricity, roughly 1/4 the figure for second-place wind.  And yes, that includes Chernobyl.

Solar is much easier and safer to distribute.

It’s total dark outside here at 45 degrees north, and well below freezing.  Distributing solar to me right now, when I need it most, would be staggeringly expensive.  If you didn’t think in sound-bite appeals to emotion you’d realize that some “solutions” cannot work, and may even be intended not to.  The well-heeled interests who finance “environmental” organizations like Greenpeace and NRDC really are cynical enough to do this.

Jeffrey Miller's picture
Jeffrey Miller on Dec 21, 2014

Finally. An EDF article that I agree with.

The corollary to TOU pricing on the consumption side is TOU (or time of production)  pricing on the production side. Following exactly the same economic reasoning that you present above, it follows that metering should be considered as two separate transactions: a bill for electricity purchased from the grid and a credit for any excess production sold to the grid.

The bill for energy purchased from the grid should be based on the full retail rate (to pay for the grid) using TOU prices. The credit for any excess power sold to the grid should be based on TOU wholesale energy prices.

Treating consumption and production on an equal TOU footing sends the correct economic signals to residential solar installers, something which is necessary if the market is to efficiently allocate resources to this sector. 

Do you agree?



Robert Bernal's picture
Robert Bernal on Dec 21, 2014

At the global level, we would have to green the Sahara desert, in order to provide the biofuels to back up a world powered by diffuse and (very) intermittant sources (that would be good for sequestering our excess CO2!). Guess how much more clean energy that requires, however…

Engineer- Poet's picture
Engineer- Poet on Dec 22, 2014

Compared to the stuff I cut my teeth on, EVERYTHING these days is “extremely capable” unless it comes in a 4-pin package, and maybe even then.

Engineer- Poet's picture
Engineer- Poet on Dec 22, 2014

NREL said we could cost effectively go solar to like 20-30% without any significant additional costs.

NREL has been totally wrong in the past, and it’s totally wrong there too.  If by “we” you mean California, any penetration close to the capacity factor (20%, tops) starts imposing large costs on the parts of the system tasked with balancing its variability.  Nationwide the PV capacity factor is closer to 12%, so the big costs would be hitting by 6-8%.

For most of the rest of us, solar is right during the middle of peak demand.

Both the daily and seasonal demand peaks come after peak solar generation.  The daily demand peak is usually afternoon (A/C) or evening (dinner), while solar peaks at noon; the seasonal demand peak is usually August (southern) or winter (northern), while solar usually peaks in June.

There isn’t any reason to load shift.

If you are going to run your A/C on solar, you have to save some of your noon-peaking generation to run cooling into the evening hours when the generation isn’t there.  Either you use batteries, or you over-size the chiller to absorb your peak power and store ice instead.

In a way “baseload” is fairly fictiuous as well.

The old RMI canard about “no such thing as base load”.  Hint:  it’s a canard.  The base load is the load that’s always there.  How you serve it is a matter of economics.

Now storage and load shifting becomes extremely important, not because of the “duck curve” but because they can make their systems much more efficiently. [sic]

More like, storage and load-shifting become essential because major sources of RE can go AWOL for days, even weeks.  Are you ready to install a battery to serve your minimum needs for a week?  Can you even calculate what such a battery would cost?

If you instead serve most of your needs with “base-load” generators like nuclear, you only need a small amount of storage.  With the base generation increased to fill storage overnight, you have a reduced gap between generation and peak demand for storage to supply.  Given the storage, you can use RE to serve some of that peak load if you want but sans subsidy the economics are not going to be there; OTOH, sans subsidy the economics aren’t there today either.  Storage can be batteries, pumped storage, or conversion of electricity into heat, ice or other forms for later use.  100 gallons of hot water in a well-insulated tank would store a lot of energy; a ΔT of 60 C would store 26 kWh of heat.  If you topped up such a tank many nights and every weekend, and only used an in-line gas heater when storage ran out, you’d seldom burn fuel.  That is the sort of thing that we need to do:  displace 90% or more of fossil fuel.

Engineer- Poet's picture
Engineer- Poet on Dec 22, 2014

Germany has a LOT of solar and they are between 47 and 55 degrees north.

German sources tout their record fractions of PV generation, always on sunny Sundays around noon.  The total share of PV generation even on the peak days is just a fraction of those records, and Wikipedia notes that total PV generation was just 7% of the total for the first half of 2014.

When you start exceeding 100% of the net load after the must-run generators are subtracted, you have to spill power more and more often.  Achieving 10% PV generation in Germany is probably too costly to be politically acceptable.

Whenever I go up beyond the 45th parallel, we also use a lot of biomass.

In other words, deforestation.  I’ve laid in a bunch of dead wood too (ash trees are dying by the millions due to borers).  I don’t expect to run my house on it, as much as I’d love to.

Robert Bernal's picture
Robert Bernal on Dec 22, 2014

I’m still trying to learn (about) how to get all the different leds to flash, lol. Will eventually buy a chip and make a fan that blows the heat from the ceiling back down using a temp sensor like the TMP36 to ADC to auto turn on only when needed. Not sure whether to go with BASIC and picaxe or Arduino and C.

Robert Bernal's picture
Robert Bernal on Dec 22, 2014

I never thought about how “mere” computers could damage equipment. That’s a good reason to go with passive save reactors.

Bill Hannahan's picture
Bill Hannahan on Dec 22, 2014

Good point Robert.  The pre Model T reactors that generate almost 20% of our electricity were designed before the internet. The important safety related pumps, actuators and valves are wired (through power relays as required) to switches in the control room. It’s impossible to hack into a system like that.

If I were designing the Model T of Nuclear power plants it would be a simple Molten Salt reactor of the sort David LeBlanc proposes. The primary reactor controls would be hardwired to the control room. Instrumentation would be a mix of hardwired indicators for the key parameters and computer driven displays to make it easy for operators to have the big picture.

Given the strong negative feedback and stability inherent in the MSR, and inherent safety features in that type of reactor, it would be easier to control than today’s reactors.

I cannot think of any way a rogue operator could cause a serious accident with a large release of fission products, perhaps Sean will tell us how he would do it if he were an operator.

Engineer- Poet's picture
Engineer- Poet on Dec 23, 2014

Frequency regulation is a key part of any electrical system. The gained efficiency, by the better regulatiion, actually balances out the costs.

Storage only helps this if it has capacity to offer.  It can’t down-regulate if it’s full, it can’t up-regulate if it’s empty, and almost all storage has limits of power handling.  The problem is that wind and PV place such large demands on regulation resources that these resources can easily be overwhelmed, making the problem worse.  This is just a part of why NREL is wrong.  We can easily prove that NREL is wrong by looking at Denmark, which managed to produce 33.8% of its electricity from wind last year but had to use 48% coal plus the massive hydro resources of Sweden and Norway to balance it.  NREL has not explained Denmark’s inability to do what it claims is possible, almost certainly because the failure cannot be explained in any other way than that the goal cannot be achieved with current or projected technology.  To admit this is to admit that NREL’s mandated purpose is a wild-goose chase, so they don’t.

We should be doing this anyway, instead of idling peaker plants

My suggestion for the off hours is to tap steam from nuclear plants to drive other processes, such as hydrolyzing cellulose and distilling alcohol.  The turbine output power will track very rapidly with the changes in extracted steam flow.  There’s your off-peak regulation.

During peak hours I’d use a number of different resources, but I’d add all available vehicle batteries in conventional hybrid vehicles for down-regulation.  Drivers would be encouraged to plug into (retrofitted) chargers when parked, and the chargers would dump power into the hybrid batteries when down-regulation was required until the batteries reached their maximum state of charge.  This would substitute grid electricity for up to maybe 1 mile’s worth of petroleum per parking session; nothing huge individually, but at a thousand or more miles per vehicle per year, big in aggregate.  There are now more than 3 million hybrid vehicles in the USA, and at 1.8 kW apiece they could absorb close to 6 GW for a few minutes.  That is a major resource.

Peak hours do run 10am through the middle of the day, and do end after dark. Yeah not wiping it all out.

That’s where the “duck belly” problem comes from.  Major dependence on PV to serve such load curves requires storage that can be filled from mid-morning to late afternoon and dumped in the evening.  Most batteries fare poorly under such deep cycling; I am partial to flywheels.

The “canard” is that a lot of the “baseload” is merely overgeneation to keep the coal or nuclear plants running since they don’t follow load or cycle very easily.

To refute a canard, you create another?  Baseload plants do not dump power; if they could do that, they wouldn’t have issues with load-following.  What they generate goes to the grid, because a loss of load without throttling back steam would overspeed the turbine and rapidly lead to failure.  Last, it’s not important that they cycle because they’re intended to serve load that never goes off-line.  Or rather, it wasn’t until some people decided that the grid should be forced to take all the output from a new set of unreliable generators and everything else forced to make up the difference.

They don’t go AWOL acrossed the whole US for weeks.

They went AWOL for the entire Bonneville Power Administration service area for almost 2 weeks early this year, and that wasn’t the only such event even this year.  If you think the rest of the country can make up for this, you’re deluded; the necessary transmission lines do not exist and it would be a legal nightmare lasting decades just to establish rights-of-way to build them.

I don’t think measuring current battery prices and making an estimation that they will be the same price in 10-20 years when we get to this point is a fair calculation.

In other words, you evade the question.

we can guestimate that storage will be like 5-10 dollars /kwh

Current lithium batteries require about 1.7 kg of lithium per kWh of capacity.  Even if you can get this down to 1 kg, you are still severely limited by lithium supplies.  Current resources are not sufficient to even give every new vehicle a Leaf-class battery pack (25 kWh); 1000 kWh per household is half a metric ton of elemental Li.  Identified world lithium reserves are about 40 million tons, so a whole 80 million households could get batteries before it was all gone.  Your figure of $10/kWh is also delusional; lithium carbonate is running in excess of $10/kg, or in excess of $50/kg of elemental Li.

you are looking about about 5k-10k for a months worth of storage capacity which is about the price of a whole house backup genset.

Try $50K minimum, then add all the hardware needed to make certain the battery is full when you need it (a battery that starts flat when your dry spell hits is useless).  Don’t forget to add the cost of the spilled power during periods of surfeit to your LCOE.

Not being at the mercy of the weather and the seasons is why nuclear power is so great; all the “renewables” become very un-renewable under such conditions, down to burning jet fuel in New England last winter.  With the closure of Vermont Yankee in 7 days, it’s going to get worse there.  Much worse.

Nathan Wilson's picture
Nathan Wilson on Dec 23, 2014

I supposed the question of whether smart meters are valuable depends on one’s opinion of time-of-use electricity pricing.  A very important feature of the electricity we buy is that it is available whenever we want it.  With time-of-use pricing, end-users can presumably help accomodate fluctuating energy sources by varying their usage.  

This works fine for electric vehicle owners (who can charge at night), but for everyone else, it strikes me as impracticle and undesirable.  

As a way to help solar power, I think time-of-use metering is over-sold, since with even modest penetration (an average of few percent solar), the peak demand period will shift to after sunset and no peaking plants can be eliminated with further solar installations. In most of the US, even cheap batteries would not change this, since clouding days (which prevent battery charging) would force the peaking plants to be built anyway.

The other problem with time-of-use billing is that it makes electricity bills have even more seasonal variation than demand (i.e. we use more electricity during peak times, plus that electricity is more expensive).  This is bad for the poor, who don’t generally get paid more when electricity is more expensive.

Mark Heslep's picture
Mark Heslep on Dec 25, 2014

“…Current lithium batteries require about 1.7 kg of lithium per kWh of capacity.  Even if you can get this down to 1 kg, you are still severely limited by lithium supplies.”

The mass percentage of actual lithium in a Li-ion cell is in the single digits.  Most of the cell mass is aluminum, oxygen, carbon, i.e. the usuals. So at 4 kg for the entire cell, per kWh, for the best li-ion cells, the lithium content is more like 0.3 kg per kWh, and the ratio is improving with time. The current ~40 million tons of global reserves then produces 141 million Tesla Model S sized 85 kWh packs.  I expect at scale lithium would be extensively recycled, as is, for example, 75% of steel in the US.

Also, reported reserves of lithium have grown over time, as the reserves of actively used minerals do, quadrupled for lithium since ’76.  After all, lithium is known to be more abundant in the earth’s crust than copper, and there are a half billion tons of copper reserves in the US alone. 

Mark Heslep's picture
Mark Heslep on Dec 26, 2014

“Germany has a LOT of solar and they are between 47 and 55 degrees north.”

In the German winter PV panels are so much dead weight on the roof.  See daily German solar output data here:

Couple days ago, the 22nd, *peak* power never exceeded 7% of namplate, and generation is actually zero before 9AM and after 3:30PM local time.  That works out to daily capacity factor of about 1%. 


EDF  Energy Exchange's picture
Thank EDF for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »