What’s the best way to price electricity?

RMI Outlet

Electricity stakeholders have long debated what makes for good rate design. There is a persistent drumbeat from many corners for wider adoption of time-based rates, and more recently, proposals are popping up for incorporating demand charges for residential and small commercial customers. Numerous pilots have been run, and many utilities now offer customers the option of time-of-use or other alternative rates.

Unfortunately, many of these offerings are seriously flawed and, as a result, severely underperform against policy goals for system cost reduction, customer engagement, and environmental improvement. Not helping matters, utilities, their regulators, and other stakeholders often talk past each other in rate design discussions, using different language and frequently approaching the issue with starkly different motivations.

RMI has released a new report, A Review of Alternative Rate Designs, to help address these challenges. The report provides a meta-analysis of numerous rate design programs and pilot studies, as well as dozens of reports and expert analyses conducted in recent years, to identify the key design dimensions of time-based and demand charge rates for mass-market customers. We cut through the noise and competing claims of what works and what does not in order to support more constructive rate design dialogue across the industry. And where there is simply not enough evidence to support definitive conclusions, we identify the gaps where future research is needed.

What makes alternative rate designs work?

To structure our review, we identify the key dimensions for each rate type—nine dimensions for time-based rates and eight for demand charges—then evaluate available research against those. Based on the identified dimensions, and measured against key objectives, including peak reduction, total energy consumption, and customer acceptance, we find impacts ranging from negligible to significant, as well as many cases where a lack of empirical results makes definitive conclusions difficult to reach. Although the results may be mixed, this is for a reason: our research points to the criticality of specific design choices to make rates meet objectives. By designing rates according to our identified dimensions, utilities, regulators, and the industry can go a long way toward doing rate design better.

Time-Based Rates: Lots of Opportunity, but Design Is Key

Utilities have experimented with time-based rates for mass-market customers for decades, and the results are in: time-based rates can be very effective when designed well. Our review found clear evidence that these rates can lead customers to reduce their peak consumption without sacrificing customer enrollment or retention, and may also lead customers to consume less energy overall.

There are many approaches to time-based rates, but all share one essential characteristic: their prices vary by time of day to more accurately reflect costs. But how often do those prices change, what is the difference between base and peak prices, and how much notice does the customer have? These fundamental design decisions are determined in part by a rate’s base structure, and by any modification options that are used.

To simplify matters, we identify three base structures that underlay all time-based rates:

• Flat volumetric rates, in which the price is the same in all hours;
• Time-of-use (TOU) rates, where predetermined prices differ across set time periods; or
• Real-time pricing (RTP), where prices vary hourly (or thereabouts) to reflect system costs.

Although many might view RTP as something of a holy grail, it is largely regarded as impractical for most customers in the near future. Instead, flat and TOU base structures can be modified to achieve some of the dynamism of RTP by sending targeted price signals when grid costs are especially high. Modification options include:

• Critical peak (CP) mechanisms that apply a predetermined price to a limited number of days each year, where events are finalized a few hours to a day in advance;
• Variable peak (VP) mechanisms that use a predetermined peak period but can vary the peak price, for example according to wholesale prices or reliability needs; and
• Flexible duration (FD) mechanisms, where peak periods are not predetermined, allowing a closer match between the actual timing of peak system costs and the signal sent to customers.

Source: A Review of Alternative Rate Designs

Whether a rate will be effective in achieving desired outcomes depends on more-nuanced design decisions along several critical dimensions. These include, in particular, structural decisions for the ratio of peak to off-peak prices (or POPP ratio) and the length and frequency of on-peak periods, as well as implementation choices, including whether the rate is opt-in or opt-out, and what enabling technology is available to automate customers’ response. These choices are key to determining whether the rate accurately reflects time-varying system costs, and whether the rate will be acted upon.

Despite their importance, designers of time-based rates have overlooked many of these choices. For example, you don’t have to look very hard to find TOU rates where the on-peak period lasts more than half the day. A choice like this may make sense on paper from a cost recovery standpoint, but the resulting rate is practically unusable for most customers.

Demand Charges: Better Understanding of Impacts Is Needed

Demand charge rates are another hot topic in the utility industry. They can provide a mechanism to directly recover marginal costs associated with peak load on the grid, and provide a signal to customers that reducing their demand is economically efficient. However, while the opportunities for demand charges are interesting, our review finds that there is comparatively little industry experience with them relative to time-based rates. Consequently, there is also less empirical evidence available to understand the efficacy of demand charges on key objectives.

While claims about the impact on outcomes remain speculative, the demand charge debate still offers insight into which design choices will be meaningful to make a rate effective. First and foremost is the extent of the demand charge—does it include most costs or only a few? At one end of the spectrum, “narrow” demand charges include only costs for things that are sized to meet a customer’s individual peak demand, such as the line transformer serving their building. “Broad” demand charges expand to include other capacity-related distribution costs, whereas “extensive” demand charges go further to include all infrastructure built to meet peak demand.

 

* In some situations, a portion of advanced metering infrastructure (and other smart-grid infrastructure) costs may be appropriately recovered through energy or demand charges. 
Source: A Review of Alternative Rate Designs

As with time-based rates, more-nuanced design choices are very important. In particular, whether charges are for customer load that is coincident with peak will determine whether or not the rate sends accurate price signals about system costs. Meanwhile, other structural choices likely influence whether customers can respond to the rate—specifically, the length of the period over which peak demand is measured, how frequently it’s measured, and whether the charge ratchets up based on past demand. It is also essential to design demand charges that are actionable by customers. At a minimum, demand charges should be communicated to customers in a straightforward manner and send a price signal that they can reasonably respond to—if customers can’t actually respond, the demand charge is effectively a fixed charge.

But before advocating for wider adoption, the industry needs to acknowledge what is known and unknown on demand charges, and where further research will be most useful. Our report attempts to make progress toward those ends.

Utilities and regulators can move forward with improved rates

Rate design is evolving, and the rates reviewed in this report are incremental steps toward more sophisticated, full-value price signals on the grid. Moving forward, new approaches are needed to achieve better pricing that reflects system costs and fulfills societal needs. This includes developing new rate options with greater temporal, locational, and attribute-based granularity, as well as improved compensation mechanisms that recognize the value provided by customer-sited technology, including distributed generation.

Still, it is important to understand the design choices involved in alternative rates being offered today. To that end, our report is instructive. First, more effective time-based rates can be developed, building on the wealth of industry experience summarized in our report, and can be widely deployed through new default rates or improved opt-in programs. On the other hand, before wider rollout of demand charges can be considered, a better understanding of their impacts is needed. That experience can be gained using demonstration and evaluation projects using robust experimental designs.

Rate designers and regulators can use this new report to help structure their work in order to more clearly and consciously address the range of outcomes that electricity rates need to achieve. Similarly, we hope that the report supports the broader electricity industry in having a more honest conversation about rate design, and spurs new ideas about how to manage the tension between maintaining a minimally complex customer experience and continuing to increase rate sophistication.

This article was originally published on Rocky Mountain Institute’s Outlet blog. Reproduced here with permission