Revisiting Lost Revenues and Cost-Effectiveness Testing in a World of Expanding DERs: Lessons from the Frontlines of Benefit-Cost Analysis
by Karl R. Rábago, Rábago Energy
There are few topics in modern rate regulation more controversial than the ongoing debate about the growth in deployment and operation of distributed energy resources, or “DERs” and the associated potential of lost and/or gained utility revenues, and how to treat them in rates, program design, and program evaluation. DERs are energy resources deployed at the distribution level of the utility grid and include distributed generation, distributed storage, energy efficiency, demand response, electric vehicles, and energy management. Lost revenues represent the money that a utility expected to collect from customers under approved rates that it will not collect—at least until, and if, a regulator approves changes in rates or charges—because the DERs resulted in a reduction in use of energy. The reverse applies in cases where a utility may “gain” revenues from, e.g., building or transportation electrification.
As brief and tidy as these definitions are, the issue of lost or gained revenues impacts associated with DERs has become a hotbed of argument, confusion, frustration, and not just a few controversial regulatory pronouncements and decisions. To this frequent trainer, debater, and expert witness on the topic, it seems time to address the topic again, and with hope that more of us will approach the issue from a position of more common understanding.
Framing the issue – the basics
First, it is important to frame the issue and not just define the terms. Regulated utilities live on the recovery of revenue requirements. Revenue requirements in turn reflect the return of and on investments made by utilities to provide utility service. Revenue requirements are determined primarily in rate cases, though sometimes also in associated or alternative regulatory proceedings, such as formula rate adjustments, tariff adjustment mechanisms, riders, and performance-based regulation decisions.
For example, if a utility invests $1,000,000,000 to build a new substation, and that substation is used and useful in the provision of electric service, and that spending is approved by a regulatory authority, then the utility will be entitled to recover the money spent plus a return on or profit on the investment. That return will be spread out over the expected useful life of the substation—say 40 years. If the utility spends $100,000 paying for miscellaneous consulting services and the spending is likewise approved, the utility will be entitled to recover that cost, though generally not with a profit, because consulting fees are an operating expense and not a capital investment. Both the annual share of the cost of the substation and the consulting fees will be included in the revenue requirement for the utility. And the utility is entitled—as a matter of U.S. Constitutional law, no less—to a reasonable opportunity to earn its revenue requirement. These and other amounts are the cost of electric service, and the approach I just described is known as cost-of-service regulation.
How the utility recovers or collects its cost of service is referred to as “rate making,” that is, the apportionment of the costs to the customers that use or benefit from the service provided. And the mechanism for cost recovery is dividing approved costs—approved revenue requirement—by the number of energy units the utility expects to sell in the time following the approval of rates.
So, if our example utility has a $27,350, 000 annual revenue requirement (1/40th of $1,000,000,000 for the substation, plus, say, 9% profit on the annual amount, plus $100,000 for the consulting fees), and it expects to sell 273,500,000 kWh of electricity at a single uniform rate, then it will be authorized to charge its customers 10 cents per kWh by the regulators. And if the utility sells exactly 273,500,000 kWh of electricity, it will recover its revenue requirement exactly.
Of course, this is a highly simplified example. There are many kinds of costs that the utility incurs to provide electric service, and there are many kinds of customers rates and charges that a utility uses to recover its cost of service. And there are many reasons why the actual level of sales that occur after rates are approved will not match the forecasts—the weather and economic conditions being chief among them. It is also possible that the forecast itself was simply wrong, for whatever reason.
These deviations from planned revenue recovery seldom move in the same direction. A particularly hot summer could drive sales higher due to increased use of air conditioning but could also motivate some customers to make significant changes in how much energy they use, such as by adding insulation or getting a more efficient air conditioner.
Generally speaking, utilities take the ups with the downs and get pretty close to meeting their approved revenue requirement. And if their earned revenue is consistently lower, or higher, than the approved revenue requirement, many adjustments/fixes are available.
Enter Distributed Generation (DG) – increased costs vs. lost revenues
Just like energy efficiency, personal conservation habits, and/or changes in household makeup—which can persistently change usage levels and patterns—customer-sited distributed generation (DG) can permanently change the usage in a particular customer home or business. Generally, sales to customers with onsite generation go down, especially during the sunny hours of the day if the generation is solar-powered. Sales go down with efficiency and conservation, but could go up due to demand response programs, building electrification (e.g., air source heat pumps), EV adoption, and use of distributed storage.
In the case of DG—which is the focus hereafter—if the utility did not forecast, or underestimated, the adoption of customer-sited DG in its last rate case, sales will fall below the forecast level (all other things being equal). When those sales fall, and if nothing else happens, the utility will face a revenue shortfall as compared to revenues assumed in that last rate case.
So, if the utility does not earn as much as it expected due to the unexpected or underestimated installation of distributed generation, why does that matter? And is that a cost?
DG can provide benefits to the utility and non-DG customers. Overall, reduced use of the grid and utility generation as a result of DG operation can save everyone money over both the short and long term. Transformers wear out faster on hot sunny days when they operate at or above the upper limits of their design. Hot summer weather causes peak demand for energy across entire regions, which in turn requires provision of the most expensive generation and increases overall system stress and can be counteracted by local distributed generation. These are beneficial impacts of distributed generation that, among others, should be accounted for in evaluating whether the utility system and society benefit from distributed generation.
But what about the cost to the utility of distributed generation? Utilities can incur costs related to distributed generation. High concentrations in particular locations can require spending on system upgrades. These added, incremental costs of distributed generation are typically charged to customers seeking to interconnect their generation systems to the grid. But if the utility incurs costs that it does not pass back to the generator, then it faces a cost. And if the utility did not plan for that cost in its last rate case, it could have an earnings shortfall.
The reduction in sales, however, is not by itself a cost. A cost is an added financial burden associated with a particular event or action. Not selling as much electricity as planned is not an added cost of serving customers, even for the customer who installs the distributed generation. The reduction in sales does not increase the cost of service.
“Not selling as much electricity as planned is not an added cost of serving customers, even for the customer who installs the distributed generation. The reduction in sales does not increase the cost of service.
The operation of an interconnected distributed generation does not add a cost to utility operations just because expected sales are reduced. So, in the business of performing a benefit-cost analysis, lost sales have no place. Lost revenues, by themselves, are not a cost or a benefit of distributed generation.”
The operation of an interconnected distributed generation does not add a cost to utility operations just because expected sales are reduced. So, in the business of performing a benefit-cost analysis, lost sales have no place. Lost revenues, by themselves, are not a cost or a benefit of distributed generation.
If lost revenues are not a cost or a benefit for including in a BCA, what are they?
Lost revenues due to sales reductions could have an impact on the utility and other customers. The utility could have a revenue shortfall and be forced to reduce spending or find new sources of revenue to do all the things it planned. One of the sources of new revenues could be other customers who do not have generators of their own, or to seek increases in rates—in the next rate case—for all customers, including those with generators.
This raises the oft cited and inaccurately named concept of a “cost shift.” The notion behind cost shift concerns is that if a utility does not earn its planned and approved level of earnings, even after trying to forecast for such impacts in the last rate case, due to self-generation, then the recovery of the costs that the utility would otherwise recover from customer-generators will be shifted to other customers without generation.
The cost shift argument, therefore, expressly assumes that the utility is entitled to all the money it would have made on a customer, even if that customer pays less due to self-generation. Often, the argument speaks in terms of customer-generators paying their “fair share” of utility sunk costs. And far too often, the argument does not include an evaluation of the net benefits and costs of distributed generation.
If a utility is subsequently allowed to recover lost revenues, and the distributed generation customer continues to use less of utility services, three important questions arise.
- First, did the self-generator reduce utility costs by using less electricity, and to what extent?
- Second, does a redistribution of those reduced costs to other customers amount to fair, just, and reasonable rate making for non-generator customers?
- Third, does the redistribution of costs to non-generators have distributional impacts that unjustly burden non-generator customers?
Answering these questions is the stuff of rate impact analysis, whereas BCAs answer the question: Which resources have benefits that exceed costs and therefore merit utility acquisition or support on behalf of their customers?
“Taken together, we can see that the best course is to fully evaluate the costs and benefits of distributed generation in terms of impacts on the utility and other customers, and on society as a whole [. . .] Then, and in parallel, undertake a rate impact analysis to determine whether an actual cost shift will arise, whether the utility should be authorized to collect lost revenues, and whether that collection will have unacceptable impacts, net of benefits, on non-generators.”
Taken together, we can see that the best course is to fully evaluate the costs and benefits of distributed generation in terms of impacts on the utility and other customers, and on society as a whole (to extent applicable to a jurisdiction’s articulated societal policy goals). Then, and in parallel, undertake a rate impact analysis to determine whether an actual cost shift will arise, whether the utility should be authorized to collect lost revenues, and whether that collection will have unacceptable impacts, net of benefits, on non-generators.
Doing a rate impact test without a BCA ignores the long-term and system-wide benefits that distributed generation can bring. And as distributed generation markets grow, evaluating rate and distributional impacts is essential to ensuring just and reasonable utility service rates for all. This evaluation should align with NSPM principles, especially in accounting for the full range of utility system impacts, and, where consistent with jurisdictional policy goals, also in accounting for host customer and societal impacts.
What if rates increase? If the potential level or distribution of rate increases for non-generators is unacceptable, the regulator and the utility have options. They can limit distributed generation with system-wide caps or by imposing charges on distributed generation that makes customer investments in generation less attractive. They can also work to expand the pool of customers that can benefit from self-generation, such as with special incentives for target customers (which can add a cost) or with support for community or shared generation programs. They can seek to maximize the value of installed self-generation with “Value of DER” rates that encourage siting and operation to maximize locational and temporal benefits of distributed generation. And they can include customer-hosted distributed generation in non-wires solutions programs aimed at avoiding or deferring specific and imminent infrastructure costs. Finally, because distributed generation and other distributed resources enjoy economies of deployment scale—they get less expensive with higher deployment rates—they can target programs to accelerate distributed generation deployment. Though this last option sounds counterintuitive, more affordable distribution generation means deployment rates won’t suffer if rates for self-generators are adjusted to increase the amount that these customers pay to the utility even after self-generation.
Despite all-too-common misperception, utilities are not guaranteed their income or their profits for shareholders. Well-established law provides that they are only entitled to a reasonable opportunity to recover their investments and money spent providing utility service, and to earn an allowed level of profit on some of that spending associated with capital—long-lived investments. So, the regulator and the utility are faced with a series of questions when customer-sited / customer-owned generation is installed and operated. These questions include:
- What does jurisdictional policy say about distributed generation? Is distributed generation part of the state’s plan for decarbonization?
- With all the other things that impact revenues, is dealing with distributed generation impacts on revenues a policy priority?
- Did the utility anticipate the installation and operation of the distributed generation? (if the generation is in the forecast, the lost sales are already accounted for)
- Were the sales and earnings impacts material to the utility’s ability to recover its revenue requirements? (hot day – increased sales and solar generation)
- What else happened with sales and earnings to offset or compound the earnings impacts the utility has seen?
- Should distributed generation revenue impacts be treated similarly or differently from all the other things that affect sales and earnings?
- Are there already mechanisms in place, like revenue decoupling, which will adjust earnings for unanticipated shortfalls?
- Will limits on distributed generation, or changes in distributed generation rates reduce or preclude the system-wide benefits that distributed generation can provide?
The issue and treatment of lost revenues versus benefit-cost analysis is critically important to understand, not only for regulators and utilities, but also the stakeholders interested in the merits and impacts of utility DER investments.
As the deployment and operation of DERs continues to grow, with many customers deploying more than one DER at a time, a BCA framework that aligns with the National Standard Practice Manual principles, coupled with rate impact analysis (Appendix A of the NSPM), is essential to avoid rate and revenue surprises and unfair burdens. And a clear understanding of the different purposes and methods of benefit-cost analysis and rate impact analysis is essential.