Skip to content

Billing Tweaks Don’t Make Net Metering Good Policy

Just before Christmas, an administrative law judge at the California Public Utilities Commission put a shiny package with a big bow under the tree for distributed solar PV companies, and gave the state’s regulated utilities a lump of coal. That ALJ’s proposed decision, which will next go to the CPUC commissioners, rejected the utilities’ attempts to end net energy metering (NEM) for new residential solar customers.  Around the same time, the Nevada Public Utilities Commission came to the opposite conclusion.

The immense challenge we face from climate change, and the limited resources the political landscape is willing to allocate to it, means it is critical to design energy policy that gets the greatest long-term greenhouse gas reduction bang for the buck.  In previous blogs, I’ve argued that net metering is an inefficient and opaque way to support the growth of low-greenhouse gas technologies (see here, here, and here), and should be replaced with more direct and transparent subsidies.

MinimumBillsFig1Rather than rehash that argument, however, I thought I would start the new year discussing two rate structures that are being proposed as alternatives to repealing NEM: minimum bills (mostly proposed by rooftop solar supporters) and demand charges (mostly proposed by those on the opposite side).


Minimum Bills

Here’s the simple math: a minimum bill is identical to a  fixed charge plus giving every customer a set quantity of free electricity.

For instance, let’s say that electricity costs $0.10 per kilowatt-hour (kWh) and there is a minimum bill of $8 per month.  This is identical to a fixed charge of $8 per month and each customer receiving the first 80 kWh for free each month. Either way, it will make no difference to the bill of anyone who consumes more than 80 kWh. Either way, any customer consuming less than 80 kWh per month will face an increased bill and an incremental cost of zero for consuming one more kWh of electricity.  The same math holds whether retail rates are constant or tiered (though there is an extra step to calculating the number of free kilowatt hours if it goes beyond the first tier).  A minimum bill yields exactly the same bill for every single customer every single month as the equivalent fixed charge and free quantity of electricity.

MinimumBillsFig2If the minimum bill is set at a low level, such as $10 per month, then nearly all customers will exceed the minimum level and it will make no difference to either customer bills or utility finances. If the minimum bill is set at a higher level, such as $50 per month, then it will raise the bills of a significant number of low-use customers. And it will also give those customers an incentive to consume more, because until they hit $50, their incremental cost of consuming one more kWh is zero. If you are already a small user who is nonetheless forced to pay $50 every month, why bother turning off the lights when you go out this evening?

Then why in the world are some environmental groups among the proponents of minimum bills?  Perhaps because they will effectively maintain the status quo of recovering all revenues through volumetric charges. No one is seriously talking about a minimum bill of $50. Both NRDC and some solar companies have endorsed $10.  The Regulatory Assistance Project is also advocating very low minimum bills, proclaiming that their virtue is that almost no customers would be affected.

That would be just fine if volumetric charges reflected the social marginal cost of a kWh, that is, the utility’s marginal cost plus all emissions externalities from greenhouse gases, NOx, SO2, and particulates.    But in California, and many other states with high electricity rates, the prices exceed any realistic estimate of a kWh’s social marginal cost, as I wrote about a year ago.

Maintaining a high volumetric electricity price is not a worthy goal if that price greatly exceeds the full cost to society. It not only prevents customers from getting as much value from electricity as they can, it also discourages efficient switching from other energy sources, such as natural gas for heating and gasoline for transportation.  I recognize that fixed charges are no panacea either, creating a difficult trade-off between efficiency and equity/distributional concerns.

In any case, rather than hiding behind a minimum bill, regulators should just admit that they are making no changes and continuing to put all costs into the volumetric price of electricity.  A minimum bill either does nothing or takes rate design in the wrong direction.


Demand Charges

A demand charge is a fee based on the customer’s highest usage in any one hour (or shorter timeframe) during the billing period, regardless of whether that individual customer peak occurred when the system was stressed or during a time with plenty of spare capacity.  With the technology of the mid-20th century, it was extremely expensive to meter usage in every hour, so a simpler device was used, which captured peak consumption, but not the time at which it occurred.

Back then, a demand charge may have been the best available approximation to a customer’s usage during system peak periods, but it was never a very good approximation as the customer’s peak may not be coincident with the system peak.  Furthermore, the customer’s single highest consumption hour during each billing period is not the only, and may not even be the primary, determinant of the customer’s overall contribution to the need for generation or transmission capacity.

In any case, the value of such approximations has been mostly eliminated with smart meters that record usage in hourly or shorter intervals.  Smart meters permit time-varying price schedules that can easily be designed to more effectively capture the time-varying costs that a customer imposes on the system, including their contribution to the need for generation, transmission, and distribution capacity.

Interestingly, some utilities and others are proposing a fee based on the customer’s usage during system peaks and calling it a “demand charge.” That gets a lot closer to Critical Peak Pricing, a form of time-varying pricing, though the details of such “demand charges” still make them a poor alternative. They tend to focus on one or two hours of highest system demand within the billing period, regardless of whether that demand level actually stresses supply capability.  In one year, the hottest day in June may be a real scorcher and may occur when some major generation is off-line, while in another year June might be mild and have plenty of available capacity every day.

Sure, such a nouveau demand charge may be a small step in the right direction towards time-varying pricing, but it requires all the same infrastructure as pricing alternatives that are much bigger steps and are likely to seem much less risky to customers, because they are not driven entirely by just one or two hours.

An additional explanation for demand charges is that they capture the customer-specific fixed cost of providing a certain level of service capacity to the customer’s site. Such capacity, however, is established by making up-front and largely sunk investments in the local distribution network and the final connection to the customer.  These may constitute substantial fixed costs, but those costs are determined by the attributes of the connection, not by the customer’s peak usage after the connection is established. A monthly fixed charge based on the customer’s service capacity would more appropriately capture these costs.

The use of demand charges has also created a large market of consultants advising customers on how to reduce their peak demand (see here, here, and here for examples). Those strategies yield little benefit to the utility or society.  Customers faced with demand charges place high private value on reducing their very highest hour of usage even if there are other hours in which usage is nearly as high, and even if none of those hours are coincident with system peak times.

At their very best, demand charges capture some customer-specific fixed costs and may very imperfectly reflect the time-varying costs of the customer’s consumption.  But customer-specific fixed charges that reflect service levels and time-varying pricing accomplish these goals much more effectively, so why would one use demand charges?


So What Should We Be Doing About Net Energy Metering?

The reality is that a customer who consumes 300 kWh in a month is imposing very different costs on the system than a customer who consumes 1500 kWh over some hours and also injects 1200 kWh into the grid during other hours. NEM treats them the same. That may have been a convenient benign fiction back when solar PV barely existed.  But today it is a costly distortion that has the potential to create huge economic inefficiencies and unfairly shift billions of dollars in costs among customers.

The CPUC needs to dig into the studies and take a stand on the real value that distributed generation brings to the grid.  And to be transparent about any additional subsidies based on learning or other technology development arguments.  Then the Commission needs to craft a tariff that accurately accounts for these values. Continuing NEM is just politically ducking the hard questions. If California is going to be a leader in alternative energy and combating climate change it has to be willing to make serious and defensible policy choices.


I’m still tweeting energy news articles and new research papers @BorensteinS 

Severin Borenstein View All

Severin Borenstein is E.T. Grether Professor of Business Administration and Public Policy at the Haas School of Business and Faculty Director of the Energy Institute at Haas. He has published extensively on the oil and gasoline industries, electricity markets and pricing greenhouse gases. His current research projects include the economics of renewable energy, economic policies for reducing greenhouse gases, and alternative models of retail electricity pricing. In 2012-13, he served on the Emissions Market Assessment Committee that advised the California Air Resources Board on the operation of California’s Cap and Trade market for greenhouse gases. He chaired the California Energy Commission's Petroleum Market Advisory Committee from 2015 until its completion in 2017. Currently, he is a member of the Bay Area Air Quality Management District's Advisory Council and a member of the Board of Governors of the California Independent System Operator.

25 thoughts on “Billing Tweaks Don’t Make Net Metering Good Policy Leave a comment

  1. In our paper Smart Rate Design for a Smart Future, RAP examines many alternative rate design options. We identify three guiding principles for rate design:
    1) Customers should be allowed to connect to the grid for no more than the cost of connecting to the grid (meaning customer-specific costs like the service drop, billing, and collection).
    2) Customers should pay for power supply and grid services based on how much they use and when they use it.
    3) Customers providing power to the grid should receive full and fair compensation, no more and no less.

    We come down to a recommendation with the following characteristics:
    1) A customer charge to cover customer specific costs, such as billing and collection; no more than a few dollars per month (actually, the only reason we read meters and render bills every month is because people use a lot of power, making MONTHLY billing costs a volumetric cost, not a customer-related cost.)
    2) A demand charge of about $1.00/kW/month to recover customer-specific capacity costs, primarily the final line transformer and secondary lines needed to connect to the grid (even a PV customer needs enough capacity here to meet their maximum use or their maximum export level).
    3) A time-varying energy charge, including a critical peak pricing component to recover shared grid costs (poles, wires, transformers, substations) and power supply costs.

    We did consider minimum bills, but found they were less desirable than a low fixed charge / time-varying rate design. And for low-cost utilities, with overall revenue requirements lower than total system long-run incremental cost (TSLRIC), we recommend adding an inclining block element to the rate design, in order to better align incremental prices to incremental costs.

    Our rate design results in net-metering being generally a compensatory approach. A solar customer that exports to the grid during on-peak periods, and imports from the grid during off-peak periods, will greatly benefit by selling at on-peak rates, and buying at off-peak rates. Conversely, a solar owner in Hawaii, where 10 – 2 PM is “off-peak” and 6 – 10 PM is “on-peak” will pay more for the power they import from the grid than they recieve for power they export to the grid, effectively paying the grid to store power for them and re-deliver it during on-peak hours.

%d bloggers like this: