Benchmarking Your Rate Case

Deck: 

Show the PUC how your filing stacks up against the others.

Fortnightly Magazine - July 2013

Around much of the country, electric rates are trending upwards. But against the current backdrop of a weak economic recovery, state commissions are increasingly reluctant to approve the rate increases requested by the investor-owned utilities they regulate. They fear instigating a ratepayer revolt.1 And publicly owned municipal utilities and rural cooperatives aren’t immune from this trend. They face similar resistance when dealing with their boards of directors. 

If history is any guide, denying these rate requests won’t make the problem go away; it will just compound it. If utility financials are put at risk because rate increases aren’t granted, investor confidence will drop, making utility investments riskier. Once that happens, it will become more expensive for utilities to raise capital, leading to yet another request to hike rates.

Perhaps the best way for utilities to make their case before regulatory bodies is to benchmark their request against those from similarly situated utilities. To assist them in this process, we construct a hypothetical, medium sized investor-owned utility, Smart Power & Light (SP&L), and demonstrate how it goes about benchmarking its rate application.

Constructing a Peer Group

The hypothetical example assumes that SP&L decides to carry out both an historical assessment and a forward-looking assessment using a single metric, such as the average retail rate, defined as total retail revenue divided by total retail sales.

In the historical assessment, SP&L compiles trends in the average retail rate across several samples of utilities.2 It lines up the following samples for examination: all investor-owned utilities; all large utilities (0.5 million to 1.5 million customers); all vertically integrated utilities; all “fast-growing” utilities (>4.4-percent customer growth per year); and the intersection of several of these categories.

Figure 1 - SP&L’s Rate-Case Benchmarks

Next, in the forward-looking assessment, SP&L reviews the assumptions and projections contained in various integrated resource plans (IRP) across a tightly defined “control group” of peer utilities. But because rate filings typically focus only on a test year, which could run one or two years out in the future, they generally don’t contain long-run projections, SP&L decides to use these IRPs to derive implicit rate projections for the utilities in this control group.

SP&L ends up focusing on nine utilities that it calls the “benchmark utilities.” These utilities are geographically dispersed and vary in factors such as size, market structure, customer mix, load growth, fuel mix, and regulatory mandates for renewable energy and energy efficiency. In addition, each utility has published an IRP in the past two years that allows future rates to be inferred (see Figure 1). 

Historically, these nine benchmark utilities represent a fairly wide range of average retail rates, as Figure 2 shows, but they all fall within the wider range shown in Figure 3, representing the overall distribution of average retail rates for all U.S. utilities, as weighted by sales. The median of the distribution is 9.6 cents/kWh and the mode of the distribution is between 9 and 10 cents/kWh. The national average lies at the center of the band created by the benchmark utilities. 

The Historical Assessment 

SP&L’s rate analysts burn the midnight oil going through reams of historical rate data drawn from the various samples of groups of utilities within the industry. From this survey they find two things. First, in 2010, the sales-weighted U.S. retail average rate was 9.8 cents/kWh. Second, over the past decade, the U.S. average rate retail grew at 3.4 percent per year. (See Figure 4.)

Figure 2 - Average Retail Rates of Benchmark Utilities

When the analysts present the results of the historical assessment to management, the inevitable question arises: why do some utilities have higher than average rates? On further analysis of the historical data, four key conclusions emerge. 

First, above-average share of residential load will lead to above-average rates.

Nationally, the residential class accounts for 42 percent of system sales. Having a larger-than-average share for the residential class will drive up the average retail rate, because the residential class is the most costly to serve per unit of electricity consumed. SP&L’s rate analysts do this computation and find that the residential class accounts for 47 percent of system sales. They also find that the company ranks in the 70th percentile of the rate distribution. When they present this result to management, they’re asked: what would happen to our rates if our residential share was equal to the national average? The rate analysts do their homework and find that the company’s average retail rate would go down and contribute to a four-percentile improvement in its rate ranking. 

Second, lower-than-average load factor will lead to above-average rates.

Figure 3 - Distribution of Retail Rates

 SP&L’s load factor lies a few percentage points below the national average, which tends to push rates upwards. A lower-than-average load factor means that fixed costs (such as capital investments in infrastructure) are recovered over a disproportionately smaller amount of sales, resulting in a higher average retail rate. Low load factors are common in hot climates, whether hot or humid, with significant peak demand due to air-conditioning load. Low load factors are often correlated with higher shares of residential load. However, low load factors also indicate that there’s a significant opportunity for demand response programs such as direct load control of air conditioners, time-of-use rates, and, if advanced metering infrastructure is in place, for dynamic pricing. 

Third, rising fossil fuel prices will push rates upwards. 

Like many utilities, SP&L has an automatic adjustment clause that allows the company to pass through rising fuel prices to the customers without going through a rate case. Since fuel prices are a significant share of the utility’s overall costs, year-to-year swings in price can have a substantial effect on the retail rate. SP&L finds that during the second half of the past decade, its rate increases were largely driven by rising natural gas prices. 

Finally, other factors exist that could push rates upwards.

Figure 4 - Retail Price Trend

SP&L is one of those utilities that operate smaller, older, and less efficient power plants with lower levels of generation efficiency as measured by factors such as heat rates, down time, and capacity factors. This is something that could change over time, as older units are retired and replaced with new generation. Additionally, it has a dispersed service territory and, unlike most utilities, it’s experiencing rapid load growth, which puts upward pressure on transmission and distribution costs.

The Forward-Looking Assessment

Of course, history is only good up to a point when making an argument about the future. Rate increases require a well-informed forecast of the factors that drive rates. SP&L’s rate department has made a projection of its future rates, using the best available empirical evidence, forecasting models, planning assumptions, and expert interviews. But to help its regulators place its projected rate increase in perspective, SP&L management now asks the rate analysts to benchmark the company’s projected rate increase so that regulators will find the company’s application to be “just and reasonable.” The rate analysts read through the IRPs of the benchmark utilities, hoping to identify the factors that will drive rates upwards, and hone in on a dozen conclusions, presented below in the form of a Socratic dialogue. 

Is SP&L’s load,  as measured by peak demand and energy sales, expected to grow faster than the benchmark utilities? The answer is yes. Thus SP&L will need to make bigger infrastructure investments than the benchmark utilities to maintain system reliability, providing a rationale as to why its rates will have to go up faster than those for other utilities. Figure 5 provides the projected peak demand and sales growth rates for the benchmark utilities. The utilities are arrayed in order of their average annual growth rate of peak demand.

Is SP&L projecting a decrease in its annual load factor? The answer again is yes. This decline in load factor will put upward pressure on SP&L’s average retail rate, since it will need to install additional peaking capacity that will be run for only a few hundred hours a year. Figure 6 illustrates the projected annual change in load factors across the benchmark utilities. It should be noted that this is the projected change after accounting for the effects of planned demand-side management (DSM) programs.

Figure 5 - Load Growth Forecast

Does SP&L have higher renewable portfolio standards (RPS) than the benchmark utilities? SP&L has an RPS that requires that 15 percent of its electricity sales come from renewable generation by 2025. This puts it in the middle of the benchmark utilities, with three having higher requirements, and five having lower requirements or no RPS at all. Thus, this factor isn’t pushing its rates to be higher than the average of the benchmark utilities, whose RPS requirements are shown in Figure 7. 

Are SP&L’s energy efficiency requirements higher than the benchmark utilities? SP&L has to comply with an energy portfolio standard (EPS) that sets aggressive targets for reducing energy consumption. Even though these programs pass the widely used total resource cost (TRC) test, most of them don’t pass the rate impact measure (RIM) test. Programs that pass the TRC test will lead to reductions in aggregate customer electricity bills, as the value of avoided or deferred resource investment is greater than the cost of implementing the programs. However, a net reduction in costs – and, correspondingly, a reduction in customer electricity bills – doesn’t guarantee a reduction in the average retail rate. If sales are reduced by proportionately more than costs, then SP&L won’t be able to recover all of its fixed costs and rates will need to rise as a result. It’s important to be aware of this distinction between bills and rates when considering energy efficiency programs from a policy perspective. These programs can be cost-effective and societally beneficial even if they don’t reduce the average retail rate. Existing and planned DSM sales reductions for the benchmark utilities are presented in Figure 8.

How soon is new generation capacity needed? After accounting for its energy efficiency and renewable energy programs, SP&L doesn’t project a need for new capacity until 10 years out into the future. A more urgent need for new capacity would translate into higher rates in order to recover the cost of the new capacity investment. This lack of urgency helps control the magnitude of its rate increase. Each benchmark utility’s projected first year of need for new capacity is summarized in Figure 9.

Are SP&L’s capacity additions expected to be larger than the benchmark utilities? As a share of peak demand, the amount of new generation capacity that SP&L plans to add over its forecast horizon is larger than that of any utility in the benchmark group. The additions are expected to be mostly run on natural gas and renewables, which is generally similar to the mix across the benchmark utilities (although there’s some variation, as shown in Figure 10). Given the significant magnitude of SP&L’s capacity additions, and the share of solar capacity in those additions, the overall result is likely to be upward pressure on the average retail rate.

Figure 6 - Changing Load Factors

Are SP&L’s transmission and distribution (T&D) costs rising relative to the benchmark utilities? If SP&L’s T&D costs are well aligned with those of the benchmark utilities (after normalizing for the difference in size), then this factor won’t exert pressure on rates. There’s a fairly significant range of T&D costs across the sample. Some companies, such as utility B, have had to invest significantly to upgrade aging T&D infrastructure due to reliability concerns that have surfaced in the wake of storm-induced power outages. Figure 11 summarizes the 2010 T&D costs.

Is SP&L’s expected annual retail rate growth generally in line with that of the benchmark utilities? Retail rate trajectories are rarely published in the IRPs, which have a long time horizon, while rate applications have a short time horizon. A company’s IRP is largely shaped by its external environment, as expressed by factors such as growth in peak demand and energy sales, load factor, CO2 price, natural gas price, and RPS requirements. 

For the benchmark utilities, SP&L obtains this data from publicly available documents and then traces how these factors are reflected among the benchmark utilities in their resource planning decisions relating to generation capacity additions, fuel mix, DSM programs, and T&D investments. Finally, it derives the projected trend in rates for each utility by dividing the projected revenue requirements by the projected sales. 

In its own IRP, SP&L initially assumes that rates would remain constant in real terms. Then it assesses how the projected sales, capacity costs, fuel costs, CO2 price, and T&D costs would deviate from historical trends, and adjusts the baseline rate trajectory accordingly. The result is a modified rate trajectory that takes into account the impacts of the previously described planning assumptions.

Figure 7 - Renewable Portfolio Standards

Given this analysis, SP&L concludes that all utilities in the benchmark sample are expected to have a rising retail rate over their respective IRP forecast horizons. The range of annual growth rates for the benchmark utilities is between 0.4 percent and 2.5 percent, with a median growth rate of 1.2 percent. In order to allow for a meaningful benchmark across the utilities, it was assumed that all of the utilities would be facing the same future natural gas and CO2 prices that SP&L has projected in its IRP. Also, the average growth rates are estimated over the forecast horizon that’s specific to each IRP, which tends to vary between 10 and 20 years, depending on the utility. In its analysis, SP&L assumed an annual inflation rate of 2 percent to convert from nominal dollars to real dollars. The results are summarized in Figure 12.

Are generation capacity costs a significant driver of SP&L’s rate increase? SP&L finds that nearly half of its projected rate increase will be driven by new generation capacity needs. In particular, relative to historical investment levels, the amount of capital being spent on renewable resources is expected to increase considerably due to the company’s RPS requirement. SP&L’s rapid load growth and declining load factor are additional influences that are driving this investment need. Relative to the other utilities, SP&L has one of the largest shares of its expected rate increase attributable to generation capacity costs. In fact, some utilities are even projecting a reduction in generation capacity costs over the forecast horizon. These tend to be utilities that don’t plan to invest in renewables.

What is SP&L’s CO2 cost exposure? It’s very likely that carbon emissions will carry a price in the future. However, on this score SP&L’s exposure is only moderate. This is due to its well-diversified fuel mix and future capacity expansion plans that are focused primarily on gas-fired and renewable generation. Other utilities that are more heavily reliant on coal-fired generation face more risk.

Will rising fuel costs drive its rates upward? Based on SP&L’s projected natural gas prices and the expected change in its fuel mix, fuel costs will make only a small contribution to the projected rate increase. Utilities with a higher reliance on gas-fired generation would expect this to be a more significant driver of rates. However, it should be noted that fuel prices – particularly gas prices – are one of the most uncertain assumptions in the analysis and have demonstrated some of the most significant historical volatility. SP&L’s well diversified fuel mix will help to hedge the retail rate against unexpected future changes in fuel prices.

Figure 8 - Projected Reduction in Sales Attributable to DSM

The extent to which individual costs are driving each utility’s rate increase is shown in Figure 13.

What is the long view of SP&L’s projected average retail rates? A summary of the historical and projected retail rates for the benchmark utilities is shown in Figure 13. SP&L concludes that both historically and in the future, its rates will lie in the center of the trajectory spanned by the benchmark utilities.

Strengthening the Rate Case

We deal with an issue that’s front and center in the minds of utility executives – how best to file their rate case. We suggest that utilities can strengthen their rate case application by benchmarking it against a peer group of utilities. By shedding light on what factors are driving rates upward, benchmarking makes the rate case more transparent than it otherwise would be. More importantly, it helps place the application in context by showing that nothing unusual is being requested. Of course, care has to be exercised in benchmarking. While it can be done with publicly available data, it has to be done very carefully and contain sufficient documentation that the usual cadre of interveners who participate in rate cases can independently verify its conclusions. 

Through benchmarking, our hypothetical utility, SP&L, was able to infer that its rate increases were being driven by upward pressures on costs and a slowdown in sales growth. Cost pressures were coming from increasingly stringent environmental mandates, and sales growth was slowing down because of a tepid economic recovery, aggressive DSM initiatives, and a pick-up in customer adoption of distributed generation options. It was also able to show that its requested rate increases compared favorably with those of the benchmark utilities.

Figure 9 - First Year of Need for New Capacity

That might not always be the case. So what should utilities do if they find themselves in the unfortunate position of requesting rate increases that are toward the upper end of their peers? It all depends on the specific situation of those utilities. 

Clearly, they can’t do much to improve the pace of economic recovery or change their customer mix. Nor can they change the environmental or energy efficiency standards that are mandated by commissions and legislative bodies. To do so would be to swim against the tide in energy policy that’s sweeping the country.

But these utilities can dampen the upward pressure on their rates by re-designing them so that they provide customers an opportunity to lower their bills while also improving company earnings. Examples include rate designs that recover a greater share of total costs through fixed monthly charges and demand charges and rates that better track the seasonal and diurnal variation in the cost of energy. These rate designs will also improve utility load factors, which for many utilities hover around the 50 percent mark. For a capital intensive industry that’s faced with below 1 percent growth in sales, that really is a sub-optimal number. Utilities should aspire to raise their load factors because that will lower their average costs and relieve the upward pressure on rates.

Endnotes: 

1. The weak economic recovery is one of five factors slowing down sales growth with demand-side management, codes and standards, distributed generation and fuel switching toward gas being the other four. See Ahmad Faruqui and Eric Shultz, “Demand Growth and the New Normal,” Public Utilities Fortnightly, December 2012. Ironically, the slowdown in sales growth is prompting utilities to raise rates, in order to recover their fixed costs which, ipso facto, don’t move proportionately with sales. For example, Xcel Energy in Minnesota is requesting a rate increase partly to recover fixed costs that run the risk of being stranded as sales growth slows down due to an expansion in its Demand-Side Management programs. Especially relevant to this paper are Faruqui’s rebuttal and direct testimonies which can be found here and here.

2. The data used in this analysis is from EIA Form 861, which includes retail revenue and sales data for each utility in the U.S.