Vitriol is to exploit the public’s inclination to favor solar to make utilities pay an exorbitant political price to have policy decided on the merits.
David Raskin is a partner in the Washington, DC office of Steptoe & Johnson LLP. He has been representing clients in the electric power industry for over 35 years.
Charles Cicchetti and John Wellinghoff have authored a recent article in this publication titled "Solar Battle Lines," (Public Utilities Fortnightly, December 2015) which condemns utility proposals to include demand charges in their retail rate designs to redress problems created by net energy metering. "Solar Battle Lines" avoids the real issues and contributes to a serious misunderstanding of electricity pricing. Populist rhetoric in favor of rooftop solar energy and against utility monopolies is almost certain to have broad emotional appeal, but net energy metering is the prototypical wolf in sheep's clothing. Messrs. Cicchetti and Wellinghoff are doing the public no favors by evading a rational discussion of this ratemaking practice.
The Economics of Net Metering
The economics of net energy metering are straightforward. If a utility uses a volumetric per kilowatt-hour (kWh) retail rate, and rooftop solar owners are allowed a full offset to their retail charges for every kWh of energy produced behind the retail meter, every kWh of rooftop solar energy produced is being valued at the full retail rate. This is true for both the kWhs consumed by the retail customer behind the meter and, in most net metered states, all of the excess kWhs delivered to the host utility.
But, the solar panels are producing only non-firm energy. The retail rate being credited to net metered customers is a "bundled" rate that includes, in addition to the cost of energy, the utility's costs for the transmission and distribution facilities used to deliver energy to consumers and the costs to achieve reliability of service 24 hours a day. These services comprise more than half the total cost of providing retail electric service that are included in the derivation of the retail rate and are primarily fixed costs.
Since rates are designed to recover the utility's total costs, the fixed costs embedded in the bundled rate that net metered customers avoid paying are allocated to the remaining classes of customers who don't own rooftop solar panels. A California PUC staff study illustrates the cost-shifting effect. It predicts that the cost shift in California will exceed $1 billion per year in a few years.1 In Nevada, the Public Utility Commission recently found that net metered customers were receiving a subsidy of between $471 and $623 per year from net metering by shifting costs to others. Studies also show that the people who have installed solar panels are much wealthier on average than those who don't, so the reallocation of costs caused by net metering is from wealthier to poorer people. The effect is equivalent to charging lower income tax rates to higher earners.2
If solar panels could be shown to reduce the utility's fixed costs, an argument could be made that customers with solar panels are entitled to at least some reduction in the fixed charges included in the retail rate. In other words, net energy metering overdoes it, but the compensation should be above the market price of energy alone. The evidence indicates that solar panels do not reduce fixed costs for several reasons discussed below.
Fixed costs are traditionally recovered through demand charges, which Messrs. Wellinghoff and Cicchetti violently oppose. A traditional demand charge would apply to the extent a customer imposes a demand on the utility's system, and it is hard to argue rationally that customers with solar panels do not. For example, most utilities peak in the evening, when the sun is setting or has set. So, solar panels do not typically cause material reductions in utilities' peak demand or a customer's use of the system during the system peak.
Indeed, the loss of solar output over the peak period means it is necessary for utilities not only to have sufficient generation to meet peak demands without a solar contribution, but also to have additional "ramping" capability to meet the peak load as solar energy is rapidly disappearing. This additional ramping capability adds to the utility's fixed generation costs. As discussed later, distributed generation also increases distribution system costs.
The shift of utility fixed costs to customers that do not own solar panels is not the only public policy concern. The developers of larger solar and wind facilities, and other generation connected on the utility side of the retail meter, receive only the unbundled wholesale price for their energy (the Locational Marginal Price or LMP), which is typically less than half the net metered price. Net energy metering in combination with volumetric rates therefore discriminates in favor of the rooftop solar alternative, distorting competition for the supply of energy. This should concern the Federal Energy Regulatory Commission (FERC) very much, since FERC has concluded that competition must be non-discriminatory in order to establish just and reasonable wholesale rates under the Federal Power Act.
The much higher effective price paid to the owners of rooftop solar panels makes rooftop solar a very lucrative business and drives capital investment toward rooftop solar and away from other forms of renewable energy that get paid less. However, recent studies show these other forms of renewable energy are much more efficient than rooftop solar.3 So, net energy metering (combined with volumetric rates) is distorting capital investment by moving capital from more efficient methods of clean electricity production to a less efficient one. As a consequence, more dollars will have to be spent to achieve any given level of clean energy use and associated carbon dioxide emission reductions.
In sum, net energy metering (combined with volumetric rates) is a bad idea if you care about consumer equity, fair competition, efficient capital allocation, and achievement of environmental goals at a reasonable cost.
To partially overcome these effects, a few utilities have proposed to include a demand charge or other fixed fee in the retail rate design. The recovery of fixed costs through a demand charge is consistent with ratemaking theory and replicates the decades-old policy of FERC. It has always required that electric requirements rates include demand charges to recover fixed costs and volumetric charges to recover variable costs, including energy. Under standard ratemaking theory, this two-part rate design ensures that fixed costs will be allocated appropriately so that customers pay for the costs they cause. The primary reason the two-part rate design has not been followed historically for most residential retail rates is that residential meters did not measure customer demand, only total energy. That is rapidly changing.
Significantly, those utilities that have proposed demand charges or other fixed fees in response to net energy metering have not done what FERC typically requires, which is to put all fixed costs in the demand charge. The retail demand charges proposed thus far have been much lower and only partially ameliorate the cost shift produced by net energy metering. So, when you read "Solar Battle Lines," which lauds FERC ratemaking, you need to understand that FERC's ratemaking rules would produce demand charges much higher than those the authors are complaining about.4
To eliminate fully the adverse effects created by net energy metering, utilities could propose to do what FERC has done to promote wholesale competition and fully unbundle the retail rate so that the component services can be accurately priced and made available for competition. With unbundled retail rates, when a retail customer reduces the amount of energy it acquires from its host utility by producing energy behind the meter, it would be credited for the unbundled energy component of retail service, at the LMP. Other component services would be priced to allow competitive entry as well, and a proper rate design would offer additional credit to retail customers that supply these other component services.
Alternatively, utilities could propose to meter the bundled retail service they supply separately from the non-firm energy supplied from solar panels behind the retail meter. So, customers would pay the bundled rate for all of their retail service and then be paid the LMP for the non-firm energy they produce.5
The Authors' Arguments
The authors do not address these issues. Most of their arguments are irrelevant to net energy metering.
The authors first claim that transmission costs are irrelevant because they have been unbundled in most jurisdictions and are subject to FERC jurisdiction. But, transmission costs are in the bundled retail rate whether the price level has been established pursuant to federal or state regulation. If net energy metering allows a consumer to avoid these costs, there is a subsidy that is picked up by other customers. There was no legitimate justification for the authors to disregard these costs, which typically are at least 10 percent of the bundled retail rate.
Although obscured by a barrage of anti-utility rhetoric, the authors appear to be making the following argument regarding the costs of the distribution system. They claim the distribution system is sufficiently robust to handle large amounts of distributed generation. They argue that because the distribution grid is a "common" resource with the capability to handle different uses, it should be made available to everyone for whatever purpose they wish to use it. In addition, they say utilities should be required to make investments in the latest technologies to enable these different uses.6 The authors conclude that absent "negative externalities, economic efficiency and public interest increase with unfettered use" of common networks like the electric distribution system.
These arguments are, first of all, irrelevant to the debate over net energy metering. No one claims that distributed generation should be denied access to the distribution network. The issue here is that the combination of net energy metering and volumetric retail rates permits a particular class of customers to use the distribution system without paying for this use, shifting the costs of the distribution system to other customers. In addition, "unfettered access" is impossible on any network that is capacity limited. If a favored class of customers can avoid paying for its share of the distribution system costs, it is effectively claiming not only a preferential price, but preferential access to constrained capacity.
In any case, the authors are wrong in contending that distribution networks are sufficiently robust to handle large amounts of distributed generation. Distribution networks have been designed to move power to customers.7 Distributed generation creates power flows in the other direction and requires the installation of equipment to manage the variability of distributed resources. The Electric Power Research Institute has published a report on this issue, which shows that adding material amounts of distributed generation will require significant additional distribution investments.8 In California, utilities have recently submitted distributed resource plans to the California Public Utilities Commission that call for several billion dollars of additional distribution investment between now and 2020 to meet distributed generation goals set by the state.9 So, the integration of consequential amounts of distributed generation substantially increases distribution costs, while net energy metering shifts these costs from those causing them to be incurred to others.
Next, the authors turn to the excess energy produced from solar panels and injected onto the utility grid, which they believe should be priced at the bundled retail rate. The authors argue that when a solar panel owner injects surplus energy onto the host utility system, and gets compensated at the bundled retail rate, there is no pricing problem because this is "banking" rather than a sale. But, the authors know that electric energy is not stored on the grid for later use. Supply and demand must be perfectly balanced in real time, so every kWh injected onto the utility's system from behind the retail meter is used instantaneously to serve load, just like every kWh supplied by a generator connected on the other side of the retail meter. Energy from solar panels displaces energy that would otherwise be supplied by such other generators, and solar panels are therefore competing with generation connected on the other side of the meter to supply energy for resale.
Of course, if a solar panel owner can sell its excess kWhs at the bundled retail rate, while generators located on the other side of the meter are paid the LMP, it is not hard to guess who will be the successful competitor. This is unquestionably discriminatory pricing, which is prohibited by the Federal Power Act and most state laws. The size of the discrimination is huge. Bundled retail rates are typically at least two to four times the market price for energy. Cleverly referring to this as a "banking" transaction does not change the physics or the commercial reality of what is occurring. A rose by any other name is still a rose.10
The authors are entitled to argue that rooftop solar panels offer such an unalloyed public benefit relative to other energy sources that a huge subsidy over and above the one Congress has already provided is warranted. They did not attempt to make that case. What is not appropriate is distracting decision-makers and the public with contrived arguments that hide the existence of the subsidy and regressive cost shift, packaged with gratuitous bashing of utilities that have raised a serious issue in open regulatory proceedings. The vitriol aimed at the utility industry on this subject is not about the merits. It is part of a strategy to exploit the public's natural inclination to favor solar energy in order to make utilities (and regulators who may agree with them) pay an exorbitant political price for seeking to have an important energy policy issue decided on the merits in a public forum.
1. Cal. Pub. Utils. Comm'n, California Net Energy Metering (NEM) Draft Cost-Effectiveness Evaluation (Sept. 26, 2013).
2. In fact, a significant portion of the economic benefit from net energy metering goes to the suppliers of solar panel equipment and not to residential consumers.
3. Bruce Tsuchida et al., The Brattle Group, Comparative Generation Costs of Utility-Scale and Residential-Scale PV in Xcel Energy Colorado's Service Area 1 (July 13, 2015) (The analysis finds that "projected 2019 utility-scale PV power costs in Xcel Energy Colorado['s]" service territory will "range from $66/MWh to $117/MWh (6.6¢/kWh to 11.7¢/kWh) across" all scenarios, while projected power costs for a typical, customer-owned PV system will "range from $123/MWh to $193/MWh (12.3¢/kWh to 19.3¢/kWh)"); Eric Lindeman, Studies: Rooftop Solar Far More Expensive Than Utility-Scale Solar, IHS The Energy Daily, Sept. 29, 2014.
4. The authors claim that solar panel owners have been mislabeled as "partial requirements customers." Although I do not understand the relevance of this claim, they are wrong. Partial requirements customers are defined as customers who supply a portion of their own electric requirements and rely on the utility to supply the remainder. This perfectly describes a solar panel owner that remains connected to the utility grid in order to obtain all of the electricity it needs over and above what it self-supplies.
5. If either unbundling or separate metering were used, proponents of distributed generation would have the right to argue in rate proceedings that the benefits afforded by this alternative justify paying an increment above the LMP. But they would have to make their case on the record in public proceedings.
6. The authors suggest, without evidence, that utilities are not investing in their distribution networks to provide new distributed services. That is simply untrue. Utilities make money by investing in assets used to provide service and recovering a return on that investment. Regulators dictate what investments will be made. The regulatory issue raised by distributed energy is whether such investments are warranted from the standpoint of overall consumer welfare when only a small number of customers benefit from the investments.
7. One of FERC's standards for distinguishing between transmission and distribution facilities is that on the latter, power flows in only one direction, toward the customer. As FERC recognizes, that is how distribution systems have historically been designed and operated.
8. Elec. Power Research Inst., The Integrated Grid: Realizing the Full Value of Central and Distributed Energy Resources 7 (Feb. 10, 2014) ("We estimate that the cost of providing grid services for customers with distributed energy systems is about $51/month on average in the typical current configuration of the grid in the United States; in residential PV systems, for example, providing that same service completely independent of the grid would be four to eight times more expensive.")
9. Herman Trabish, "How California's Biggest Utilities Plan to Integrate Distributed Resources," Utility Dive, July 7, 2015.
10. In addition, the authors are unclear about what is being "banked" or what the proper compensation is for using a utility's system to bank energy. They appear to believe that solar panel owners should be able to "bank" energy but receive a credit for bundled retail service. They also seem to believe that solar panel owners are entitled to bank energy with the utility for free.
Lead image © Mpfaff | Dreamstime.com - Solar Photo