Residential Demand Charges: Bad Choice

An increasing number of utilities propose to add demand charges to rooftop solar customers’ tariffs. Other jurisdictions are considering applying demand charges to all residential customers.

This discussion is limited to the effect of residential demand charges for rooftop solar. Nevertheless, any residential demand charges would sharply reduce the economic incentive for all customers to purchase energy efficient appliances and devices for the home.

Demand charges would significantly increase the amount rooftop solar customers pay. 

Existing rooftop customers would pay twice for capacity. First, for the cost of installing renewable electricity generation. Second, for a new assessment built into the tariff to collect capacity fees for the utility.

Utilities argue that because capacity costs are rolled into the basic energy charge, rooftop solar customers avoid paying their fair share of capacity costs when they reduce the amount of electricity they purchase from the utilities. But the utilities go too far in their proposals to recover capacity costs from rooftop solar customers who self-generate. Customers who self-generate cause the utility to buy less peak period energy or postpone new generation investments to meet future peak demand.

This is unfair and inefficient. It introduces a rate design that ignores the cost causality and user pays principles of regulation. Rooftop solar customers mostly do not use utility capacity during peak demand periods. The customer is self-generating part of her electricity requirement during those costly high utility demand periods.

The first conceptual flaw and problem with demand charges for rooftop solar customers is that the utilities propose a demand charge based on a customer’s total household consumption. This unreasonably ignores the amount of peak period energy and related capacity that rooftop solar customers purchase from the utility.

In effect, the utilities’ proposed demand charge forces rooftop solar customers to pay for both capacity they no longer use and for capacity the utility no longer must provide to rooftop solar customers for an extended number of years, likely forever.

Regulators should reject the proposed residential demand charges for rooftop solar because such charges do not recognize the difference between a rooftop solar customer’s maximum household consumption, which includes the electricity the customer self-generates, and the same customer’s maximum purchase from the utility, particularly during peak demand periods.

A second problem with requiring demand charges for rooftop solar customers can be readily understood with an example based on an assumed monthly electricity bill. Current tariffs roll the cost of capacity into the basic energy charge.

For example, suppose a customer consumes a thousand kilowatt-hours each month, and pays ten cents per kilowatt-hour to the utility, for a monthly electricity bill of a hundred dollars. Potential rooftop solar customers under prevailing residential tariffs compare the all-in solar costs per kilowatt-hour that they produce to a utility average rolled-in price of ten cents per kilowatt-hour, which represents average energy and capacity costs bundled together.

Continuing the example, suppose the utility imposes an eight dollar per kilowatt demand charge on rooftop solar customers. And, for simplicity, assume the customer in question is an average residential user, who has maximum peak consumption of eight kilowatts.

This would result in a new tariff for rooftop solar customers that collected sixty-four dollars per month in demand charges. The remainder of the customer’s one hundred dollars per month, or thirty-six dollars per month, would be collected in an energy charge.

The new energy price would decrease from 10 cents per kilowatt-hour to 3.6 cents per kilowatt-hour.

The customer would already be paying for the rooftop solar system she uses to self-generate, and she would pay the additional sixty-four dollars per month in demand charges. Her only savings would be restricted to any electricity that she purchases when her solar unit is not operating, and when any electricity she has stored is not sufficient to satisfy her needs.

It is inconceivable the rooftop solar customer could ever come close to saving enough on her remaining utility purchases to make any significant dent in the sixty-four dollars per month hole she would find herself in, after paying both a demand charge to the utility for capacity she no longer needs, and also for the rooftop solar system she added to her home.

This makes the decision to go solar very uneconomic for prospective new solar customers. If demand charges were required retroactively, her investment in rooftop solar made under prior tariffs would become under-water.

Incentives to add rooftop solar would be sharply diminished. Potential rooftop solar customers would compare the reduced utility energy price of 3.6 cents per kilowatt-hour, not the previous 10 cents per kilowatt-hour, to her expected all-in solar costs.

Quite obviously, there would be fewer rooftop solar installations if the utility energy prices were reduced to 3.6 cents per kilowatt-hour. And, no matter how much she self-generates for her own use during the utility’s peak demand periods, she would pay a monthly demand charge of sixty-four dollars based on her use of eight kilowatts of maximum demand and a utility demand charge of eight dollars per kilowatt.

The household’s economic incentives for installing rooftop solar would virtually disappear. This would happen just as the installed cost of solar panels is decreasing, and their performance is on the upswing.

A third reason for rejecting utility proposals to introduce demand charges is that there is a better tariff reform that regulators should consider. This alternative is Time-of-Use tariffs, where the residential customer would pay higher than average energy prices during peak periods of utility system demand, likely during daylight hours on weekdays in the summer.

The customer would also pay less than peak prices for electricity during intermediate hours and days when system demand was less. And even less during off-peak periods with relatively low system demand.

Both Time-of-Use and adding demand charges require new metering technology and billing changes. The costs of implementing them are about the same for both the utility and residential customers.

Time-of-Use tariffs would be based on the utilities’ marginal or avoided costs that reflect how utility engineers and competitive wholesale markets actually and conceptually dispatch electricity generation. Peak demand periods, both hourly and days of the week, have higher utility marginal costs because as utility load or customers’ demand increases, electric companies or the wholesale market dispatch increasingly more expensive-to-operate generation.

Time-of-Use pricing would signal varying electricity marginal costs to consumers, in either real-time or by relying on objectively-determined projections of electricity prices for different time periods.

Under Time-of-Use pricing, the ten cents per kilowatt-hour average residential energy price might, for example, become twenty cents per kilowatt-hour during peak hours and days. Off-peak prices might become five cents per kilowatt-hour. There could be interim period prices of nine cents per kilowatt-hour.

Time-of-Use tariffs introduce incentives for rooftop solar that are dramatically different than either rolled-in residential tariffs or tariffs that add demand charges. Much of the electricity that rooftop solar produces would come during peak and interim periods. Therefore, potential rooftop solar customers would compare their expected all-in rooftop solar self-generation costs mostly to the utility’s peak and, perhaps to a lesser extent, interim prices.

Time-of-Use tariffs would increase the expected savings for potential rooftop solar customers because at the times rooftop solar would typically be generating electricity, the customer would save twenty cents per kilowatt-hour, not the ten cents per kilowatt-hour under the previous rolled-in residential tariffs.

More importantly, the incentives are very different than the much reduced incentive for self-generation based on the 3.6 cents per kilowatt-hour energy component for the demand charge tariff. The potential rooftop solar residential customers would also pay less for the electricity they purchase from the utility during off-peak nights and weekends when their rooftop solar facilities were not producing electricity. These prices would likely be less than the pre-existing average or rolled-in price of ten cents per kilowatt-hour.

The affirmative case for Time-of-Use tariffs that reflect marginal costs is strong for all customers. The tariffs follow widely accepted engineering economic decisions using least cost principles of economic efficiency.

Regulators should not be misled by pitches that utility companies are making to fix the rooftop solar problem by requiring customers to pay demand charges. Time-of-Use tariffs are fully in accord with the public interest. This is in contrast to a demand charge tariff that is based on the mistaken notion that customers should pay the utility for capacity they no longer purchase from the utility. The new demand charge tariffs quash competition from a burgeoning rooftop solar industry.

The conclusion that Time-of-Use tariffs are the preferred fit for rooftop solar customers stands on its own. It increases in strength when regulators consider other rooftop solar benefits, including environmental and climate change advantages, local jobs, and economic security.