Skip to content

Why Aren’t We Talking About Net Energy Metering for LEDs?

The fights over net energy metering have gotten loud and heated. For those of you who have missed the drama, here, in a nutshell, is what “net metering” means. Say I install enough solar panels on my roof to provide about half of my electricity over the course of a year. On a sunny afternoon, if I’ve turned off my Tivo and my refrigerator and dryer aren’t running, my system might be generating more electricity than my house is consuming.

Net metering means that my utility will credit me for the “extra” power my system generates at times like this and charge me based on the difference between my total consumption and my total solar production, i.e. my net consumption. I will be selling back to the grid during the sunny afternoons when my own consumption is low. (See Severin’s previous post on this.)

Will you neighbors admire your solar panels?
Will you neighbors admire your solar panels?

So, how does this apply to LEDs? From a purely technical perspective, it doesn’t. An LED would never bring a house below zero consumption so that it’s selling back to the grid. But the zero consumption threshold isn’t what all the fights are about. If my consumption without the solar panels puts me on the fourth tier, where my rates are as high as 36 cents per kWh, my solar system helps me avoid paying for some pretty expensive electricity. Even if I never go below zero, a solar system will keep me down on the lower tiers, paying only 13 cents per kWh. (On a side note, the solar installers who have approached me understand this and size their systems to avoid the expensive power but not the cheaper stuff.)

But, PG&E has fixed costs, which won’t disappear just because I install solar – it still has to run distribution wires to my house and pay for my meter. It’s paying property taxes and financing costs for its power plants no matter how much electricity they produce.

The basic problem is that utilities are collecting fixed costs – which by definition do not vary as a function of how many kWh customers consume – on a volumetric basis. So, every time someone installs  solar panels, the remaining ratepayers have to pay slightly more to cover those costs. Yes, my neighbors (and all of PG&E’s residential customers) would have to pay for my meter and my “share” of the distribution lines running down our street if I got a large solar system.

The opponents of net metering argue that it provides unfair incentives for people to install solar, which leave the rest of the users on the hook to cover the utilities’ fixed costs. They cite the “death spiral,” meaning that rates get higher for non-solar customers, which induces more of them to switch to solar. The more sophisticated opponents note that solar installations are generally on rich people’s houses, so net metering is regressively subsidizing the rich.

This brings us to the question posed in the title. Why isn’t anyone complaining (at least very loudly) about unfair cost shifting or the death spiral when I buy an LED bulb? Just like a solar system, my LED bulb will help me avoid the 36 cent power, and, given that some of that is collecting fixed costs, my neighbors will be left paying a tiny bit more – if not tomorrow, after PG&E’s next rate case.

My first guess was that we’re not talking about an energy efficiency death spiral since we’re still talking about the energy efficiency gap, which implies that customers are not investing in seemingly cost-effective energy efficiency measures. In other words, customers have been leaving the proverbial $20 bills on the sidewalk and bypassing energy efficiency opportunities, like replacing their incandescent bulbs with LEDs, even though the switch could save them money.

But, I did some rough calculations and concluded that annual energy savings from LEDs could be on par with or even larger than distributed solar, especially when you bring in commercial lighting. Very roughly, California added approximately 1,000 MW of solar in 2013. At an estimated capacity factor of 17%, that’s roughly 1,500 GWh of annual solar production. This report estimates nearly 1,000 GWh of annual savings from light bulb standards in California, climbing to 11,000 GWh by 2018.

Lots of potential?
Lots of potential?

And, we’re talking about potentially large amounts of money shifted across customers, as a large share of the typical utility’s costs is fixed. The California Public Utilities Commission put out a study on net energy metering that calculated that the typical solar customer paid bills that were 54% higher than the utility’s incremental cost of serving them before they installed solar and 12% less than the incremental cost of serving them after they installed solar. This is suggestive of just how much fixed costs are collected on a volumetric basis, particularly on the higher tiers.

So, I predict many more years of heated discussions about rate restructuring. I’d guess that we will look back with amusement at the contentious debates about whether to add a monthly fixed fee to PG&E rates, because the typical customer will be paying a much larger share of their bill as a fixed charge.



Catherine Wolfram View All

Catherine Wolfram is Associate Dean for Academic Affairs and the Cora Jane Flood Professor of Business Administration at the Haas School of Business, University of California, Berkeley. ​She is the Program Director of the National Bureau of Economic Research's Environment and Energy Economics Program, Faculty Director of The E2e Project, a research organization focused on energy efficiency and a research affiliate at the Energy Institute at Haas. She is also an affiliated faculty member of in the Agriculture and Resource Economics department and the Energy and Resources Group at Berkeley.

Wolfram has published extensively on the economics of energy markets. Her work has analyzed rural electrification programs in the developing world, energy efficiency programs in the US, the effects of environmental regulation on energy markets and the impact of privatization and restructuring in the US and UK. She is currently implementing several randomized controlled trials to evaluate energy programs in the U.S., Ghana, and Kenya.

She received a PhD in Economics from MIT in 1996 and an AB from Harvard in 1989. Before joining the faculty at UC Berkeley, she was an Assistant Professor of Economics at Harvard.

48 thoughts on “Why Aren’t We Talking About Net Energy Metering for LEDs? Leave a comment

  1. Hi Catherine

    One thing to note. The CA IOUs are much more protected when it comes to EE then say rooftop solar. The protections come from two distinct but related places. First, in ratemaking, CA electricity and natural gas IOUs have decoupled. Which means that achieving the revenue requirements need to run the system is not dependant on volumentric sales – we also allow for recovery of avoided sales which are attributed to things like installing the LED. So there is no death spiral because of how the rates themselves are designed. Secondly, the CA IOUs actually receive a shareholder incentive for acheiving EE goals and good management. This reward, termed the Efficiency Savings and Performance Incentive (ESPI) is a critical pathway for senior management to actually recognize and put top talent into the deployment of efficiency, which is the state’s top priority resource. So the death spiral does not occur because there are no stranded costs and because there is a potential to earn a substantial amount of profit. In fact, PG&E in it’s Q4 report to investors specifically called out this incentive earning as a critical source of profit.

    While there are many reasons why fixed charges do and do not make sense for the CA IOUs, I would aruge that the efficiency model in the state has adapted to the current system. If the state elects to adopt a fixed charge model for ratepayers, we will have to think through the implications of decoupling revenue from sales and how to adjust the ESPI to continue to send strong signals to IOU senior management to aggressively pursue our state’s cleanest and cheapest energy resource.

    • Isn’t this all more or less the same as paying farmers not to plant crops? The folks who pay, are the folks using electricity (some use less, everyone pays more).

      • Glenn

        Statute requires that on a portfolio basis that all EE programs be cost effective. We are not “growing extra crops” just as a feel good – we are deploying as much efficiency as possible to curtail load growth to avoid new, more expensive power plant development.

        Decoupling is a ratemaking mechanism designed to separate revenue requirements from sales. If not for decoupling, then the utility would need to sell as much as possible to make their full rate of return. If we ask the utility to not sell as much of the product, then we should ensure that we are not compromising the safe and reliable service they do provide by having them come up short in terms of meeting revenue requirements.


      • I don’t agree with your summation Michael. Its amazing that in the deregulated states the going residential rate per kWh is $0.20 to $0.30 (or more at various times) while in the states wise enough to not regulate to destruction the electric utility industry and take a more measured approach, the going rate is nearer to $0.10. We should sell California to Mexico at the earliest opportunity because the decades long typically and traditionally rampant screwing up the other 49 states has been going on for far too long… Please, bring home Nancy Pelosi and Dianne Feinstein at the earliest opportunity and keep them and their corruption within your borders where they can do the least harm to the rest of us non-Californians.

    • Hi Michael-

      Thanks for the comment. Doesn’t decoupling only help for savings attributed to utility programs, so if LEDs diffuse because of standards, e.g., decoupling won’t help?


      • Catherine

        Two quick points back. First, LEDs are not yet code and are a part of the IOUs portfolios. You might not see a physical check, but much of the IOUs activities are either upstream (with the manufacturer ) or midstream (with the retailer). LEDs and its dissemination (for now) fall into those camps. Before LEDs can become code, far more standards need to be developed. Right now, we are trying to buy down the top half of the market (in terms of quality of bulb) to avoid the mistakes we made with CFLs.

        Second, the IOUs get savings credit for advocacy work for a more stringent code, at least at the state level with Title 20 and Title 24. There’s some consideration about having code advocacy at the Federal level, but attribution is much more difficult in that arena.

        At any rate, the question you posed is about if decoupling would work with items placed ‘in code’. I genuinely don’t know. I think that since total load growth has increased (but has stayed flat on a per capita basis) that it has not been an issue, yet.

        In your original question, you ask why isn’t net energy metering an option for EE. Some folks are trying to figure out if a PPA model is viable for EE. It would be a far better thing to finance around. We are talking about the permanent installation of an asset, which seams parallel to a solar panel on a roof. However, so much of the savings associated with EE is dependent on behavior of the occupant. I would posit that in order for this model to work you need far better commissioning data and the ability to compare to the degradation of solar production. (Solar panels don’t produce as much over time; does that decay rate outpace the EE savings degrade???)


  2. Mark Lively has it in a nutshell. I agree that difference between solar and the LED is that the LED reduces demand or the amount of distribution system that is needed to support that customer. I would add that NEM customers avoid paying for some utility fixed costs and that the costs they avoid paying vary from utility to utility as the benefits of solar generation also vary. For example there are distribution system benefits resulting from solar generation when a utility peaks with air conditioning load. There are no distribution system benefits from solar generation for a night-time, winter peaking utility. So the discussion pertaining to NEM is not the same for all utilities. What is common is that when NEM customers avoid paying for their share of the distribution system and billing costs they leave the non-NEM customers to bear that cost burden.

    Restructuring rates to recover fixed costs independent of usage is a way forward however this will result in a reduction in the bundled rate which in turn will reduce the incentive for energy efficiency and solar generation. To counteract this effect it is important to develop alternative, innovative programs to continue to incent EE while instituting rate structures that value solar and storage according to each individual utility’s cost structure. A well thought out Feed In Tariff can accomplish these goals. There are several examples that break down the utility’s costs based on a Cost of Service Analysis and establishes a value for each component to which distributed generation contributes (such as Austin’s). The result serves to encourage cost effective distributed generation that does not burden non-generating customers. The solar installer industry is not generally supportive of this approach I think in part because it requires they become more thoughtfull and their pricing more competitive.

  3. Several years ago I was discussing those $20 bills lying on the sidewalk with the late Lester Lave, who told me, based on his own research, that they weren’t 20s, but 50s or maybe even 100s, but they were glued to the sidewalk and the job of the EE experts was to design tools to scrape them off without destroying them. Rate design is probably one of those scrapers.

  4. A small rural electric cooperative (REA) in rural Colorado recently implemented a $29/month service charge – details at . So I now pay $29/month plus ~10.2 cents/kWh for residential service. REA’s aren’t seen as leading-edge innovators, however perhaps this is the start of a shift towards higher fixed, and lower variable/volumetric, pricing in residential electricity.

  5. I once worked for a PhD economist who said that some analyses such as the concern about LEDs should be looked at after one waves a magic wand and puts into place a new, optimized system. With a new optimized system, we would still need the same system for the roof top solar because the utility has the obligation to back them up. But the LED situation wouldn’t need the same system because the demand would be permanently reduced. That explains why LEDs aren’t an issue in the cross-subsidy discussion.

    The real issue is that we are trying to collect demand costs using an energy billing determinant. The energy billing determinant is going down while the demand and the demand costs are staying the same. The solution is to change the determinants used in billing electricity, using a demand charge instead of an energy charge. The huge metering infrastructure that has been put into place has 15 minute interval metering, which can be used for demand billing to recover the wires costs of the distribution systems.

    Residential demand charges, though rare, are not new. I saw a residential demand charge 35 years ago in Burbank. That residential demand charge had a twelve month ratchet. Because of the backup nature of serving a customer with PV, perhaps a five year ratchet would be appropriate. Further, the growth in computational speed means that there could be separate demand charges for customer demand, contribution to neighborhood peak demand, coincident demand, . . .

    I talk a little about this in my blog entry, “Electric Demand Charges: A Lesson from the Telephone Industry,” 2014 February 10,

    • You seem to be talking about contribution to coincident demand on various types of equipment, which are much better than charges based on the customer’s non-coincident demand. Utilities tend to prefer the latter, since they are very difficult to avoid. Unfortunately, traditional demand charges provide no useful signal about optimal timing of power usage, since every customer is being charged for usage at a different time. Some percentage of customers will be rewarded for shifting peak load onto the transformer, feeder, substation, transmission line and generation peaks.

      Coincident peaks also oversimplify the cost drivers, which include any high-load hour that contributes to LOLP for generation capacity, total energy usage for investment in intermediate and baseload generation, usage on high-load days that contributes to overheating and premature aging of transformers and lines, and the numerous reasons that transmission lines are built (economic transfers in off-peak periods, exports, contigencies…). Time-dependent energy rates do a better job at conveying appropriate signals in most cases.

      And ratchets make the problems of demand charges worse, making cost-control even harder and reducing incentives to conserve or shift load (unless the shifting is absolutely perfect).

  6. There is probably some historical justification here. Regulatory and rate treatment of energy efficiency was instituted when demand was growing quickly enough that energy efficiency investments were expected to slow that growth down, not reverse it. Hence current rates could still cover current rate base. The concern with net metering appears to be that it could reverse load growth — total load, not just the load served by utility supply. The reduction in peak load growth may also be considered an unreliable basis on which to reduce capital investment — if you get a hot overcast day (the Southwest Monsoon?) could the net metered PV go away? LEDs probably won’t fail & be replaced by incandescents all at once.

  7. Hi Catherine,

    That’s a nice reminder that with volumetric pricing of fixed distribution costs–back in the old telecom days, they were called “non-traffic sensitive” costs– the incentives for energy efficiency are distorted as well as incentives for distributed generation. I suspect the incentives for DG are probably more distorted, as not only does one see per kWh prices that are too high with volumetric pricing, but DG installers also get the benefit of subsidized power insurance in the form when they can maintain a connection to the grid without paying for it most of the time. Still, your point is very well taken; thanks for writing it up.


  8. By extension, a person who flips the master breaker off, and sits in the dark, is imposing a cost on all of the other rate payers who don’t flip the master breaker. A homeless person imposes even greater costs, because they don’t consume electricity in a home but still utilize public facilities which do consume electricity.

    I believe this argument is misleading because the problem with net metering really boils down to everyone paying a fair share of the transmission capacity when they pump electricity onto the system, not when they take it off. In Texas, the electric bill has two parts: the cost of the electricity per kwh and the cost of the transmission capacity per kwh. I think the transmission capacity runs about four cents where I live. On the bill it appears that the consumer is paying the transmission factor, but in fact, the producer is paying the factor and passing it on to the end user.

    If you think of transmission capacity as analogous to a tollway where everyone pays to use it. as they enter, then the solution is very simple: Charge the net producer of electricity, the full transmission factor of four cents, or the like, as she pumps her electricity onto the grid. Yes, it appears that the the solar producer pays for electricity, both coming and going, but the net solar producer, who didn’t build the grid, is both a producer and a consumer. When she is a net consumer, she pays the passed on cost of transmission. When she is a net producer, she is charged for the capacity factor and other net producers are unharmed, economically. If the net producer wants to use that grid, for profit, i.e. net metering; then she should pay her fair share of the capital cost.

    • Mike, you’re partly right but you should have stopped while you were ahead. A homeless person imposes no cost on the utility because he is not connected to the system. Public facilities, like light street lighting, are “public goods,” i.e, one person’s consumption does not limit another person’s consumption and it is essentially impossible to stop someone from consuming the good for free. Anyone consuming a public good imposes no costs on others.

      The separate buy-sell scheme that you propose in your last paragraph is essentially what Austin Energy (City of Austin Texas) does. It charges customers with solar based on the amount of electric energy they consume (including that which they produce on-site) and separately credits them for their production at a price based on Austin Energy’s avoided cost.

      • Thank you for the example about Austin Energy. I wonder if there are any other such examples, nationally. I know that net metering has become a very contentious issue in places like AZ and CA. This seems like a very commonsense compromise with which to placate all stakeholders.

    • I wonder what percentage of fixed / annual / historic costs are tied up in implementing / piloting / removing / re-installing about 6 generations of smart meters by the millions under order from the California Public Service Commission and related (problem causing) poltiticains?

  9. I was under the impression that under California deregulation PG&E and SCE were required to sell off generation base load plants (except a nuke or two no one will buy). Which of course leaves them as nothing but middle men / distribution companies subject to volatility of the wholesale power markets, hardly fixed cost for supply. Has something changed?

%d bloggers like this: