As power customers throughout the United States clamor for more distributed energy options, regulators and utilities alike are tasked with reforming rate structures to accommodate them.

For years, the typical method for compensating owners of distributed generation like rooftop solar has been through retail rate net energy metering (NEM) — crediting owners for the generation they send back to the grid at the rate at which they buy electricity.

But where rooftop solar and other DER penetration is reaching high levels, regulators are revising NEM structures over concerns about utility finances and cost-shifting to the larger rate base.

Of the 39 states working on solar policy during the first quarter of 2016, 22 considered or enacted reductions to the value of the NEM credit, according to a recent report from the NC Clean Energy Technology Center (CETC).

In place of net metering, many policymakers are looking for a new kind of rate structure that will encourage DER adoption without imposing a burden to the rest of a utility’s rate base. While not mutually exclusive, two central policy options have emerged — residential demand charges and time-of-use rates.

In the first quarter, six of eight utility filings with regulators for changes to residential rate structures were requests to add demand charges. Decisions are coming on utility filings in Arizona, Oklahoma, and Texas, according to the CETC.

Time-of-use rates, while more common in existing utility rate structures, have also been utilized in DER proceedings. Earlier this year, California regulators preserved retail rate NEM until 2019, but also ordered the state’s utilities to reorganize their rate structures to put all consumers on TOU rates by that time.

A new report from the Rocky Mountain Institute, a clean energy think tank, takes a close look at the potential of rate design reforms and what real-world experience can teach utilities and policymakers about both demand charges and time-varying rates.

Despite good rate design principles established in the 20th century, “rate design typically gets reduced in practice to the single objective of cost recovery," said Dan Cross-Call, a manager in RMI’s electricity practice and co-author of "A Review Of Alternative Rate Designs."

“We lose sight of the other objectives and opportunities,” he said. “You can achieve cost recovery while also attending to other social and policy objectives like customer engagement, total load reduction, peak reduction. They are not mutually exclusive.”

Time based rates and demand charges, explained

The problem with traditional utility rate structures is they do not convey the real cost of electricity at any given moment, the report notes, so customers use electricity “indiscriminately.”

To correct this, time-based rates (TBR) — how the RMI report terms TOU and other variable rate structures — “vary by time of day to more accurately reflect costs.” These price signals can motivate customers to alter their usage patterns in ways that can reduce both peak and overall load.  

With a demand charge, customers pay a fee based on their period of highest consumption or their usage during a pre-defined peak period. That sum is added to customers’ volumetric and fixed bill charges.

While programs vary widely by utility, demand charges are typically applied to commercial and industrial customers, while their use for residential ratepayers is a more recent innovation spurred by the growth of distributed resources, especially rooftop solar.  

The demand charge appeals to utilities because they see it as providing a more certain way to cover flat or declining revenues, Cross-Call said. To make it more palatable to consumers, demand charges are typically introduced along with decreases to volumetric and/or fixed fees.

Both demand charges and time-based rates aim to reduce consumer usage and shift it to off-peak hours. While TBRs apply a broad incentive against consumption during periods of high demand on the grid, demand charges more specifically target peak usage of individual customers, motivating them to use practices or technologies to bring it down.

When applied successfully, the rate structures can reduce peak load enough to allow utilities and grid operators to defer or cancel costly investments in grid infrastructure, the RMI report notes.

The researchers offer nine fundamental “design dimensions” for TBRs and eight for demand charges. They are meant to be “a common framework against which utilities as well as regulators can design and evaluate rates,” Cross-Call said.

Time-based rate design

RMI’s review of the utility industry’s experience with time-based rates shows they do “reduce peak consumption and total energy consumption without compromising customer acceptance.”

There is, however, a crucial caveat.

“Evidence shows that well-designed rates can have significant impact, but also that poorly designed rates can have a negligible impact,” researchers noted.

The TBR dimensions examined by RMI begin with cost recovery, extend to how it is done through rates, and examine what role – if any – time periods play.

“Any time based rate has these dimensions,” Cross-Call said. “It is a question of how well they are designed.”

Most of the TBR dimensions are about how they would be structured. These include the ratio of the Peak to Off-Peak Price (POPP), how long the peak lasts (Duration), and when daily and seasonal peaks occur and how they are priced. The last structural question is whether prices or rebates are more effective.

The dimensions that most impact the level of peak demand reduction are the price ratio, the financial mechanism, and the enabling technology. Customer acceptance is most impacted by the duration and frequency of the peak. The enrollment method impacts both.

Ratios used for the POPP have typically ranged from 1:1 to 7:1. “Critical peak period” ratios applied to limit a certain few hours of very high seasonal demand each year may range from 4:1 to 20:1.

“Where the ratio is close to 1:1, it is effectively a meaningless signal,” Cross-Call said.

The paper offers more detail. “A 2:1 ratio tends to produce peak reduction of approximately 5%, vs. approximately 10% from a 5:1 ratio,” it reports.

TBR durations have ranged from 4 hours to 16 hours. The best customer response comes from durations that are “as short as possible while still capturing the necessary peak hours,” RMI reports. Surveys show customers respond best if the duration is no longer that five hours, it adds.

"Some utilities have an on-peak period that is more than half the day, 12 to 18 hours in some cases,” Cross-Call said. “That is an essentially unusable price signal.”

Once those structural questions are answered, the rate design has to define whether participation should be voluntary (opt-in) or default (opt-out). Opt-in rate enrollment attracts “more-engaged participants, but opt-out rates have enrollment rates three times to five times higher,” RMI reports.

“The opt-out rate will typically have a smaller individual average result because it captures a larger set of customers who might not pay as much attention to their electricity bill," Cross-Call said. "The individual impact from an opt-in rate can be higher but the aggregate impact is lower.”

After participation, rate designers must indicate whether TBRs can be implemented through automated technologies or those that require direct customer participation.

One of the report’s most interesting findings is the importance of enabling technologies like smart meters, smart thermostats, and in-home displays. If effective, they can cut customer peak usage up to 20%. But passive technologies have only half that impact. One program that provided passive in-home displays at no cost found up to two-thirds of customers not even using them.

Real world TBR

RMI used time-of-use rate results from Sacramento Municipal Utility District (SMUD) and Oklahoma Gas and Electric (OG&E) to illustrate the interwoven impacts of good design principles.

Both utilities used a peak period duration and POPP ratio that enabled consumption pattern changes. They used marketing and customer education campaigns to drive customer opt-in.

“SMUD’s TOU rate achieved an average peak reduction of 6% despite being a default rate (and therefore including less engaged customers, who are less responsive),” RMI reports.

“Based on the high satisfaction ratings we received from customers participating in our SmartPricing Options pilot, we started offering solar and EV customers the opportunity to shift to time-of-use rates this year,” SMUD Director of Customer Retail Strategy Erik Krause emailed Utility Dive. “We’re seeing high adoption rates with those customers as well, so we plan to offer all of our customers a time-of-use.

OG&E's smart meter-enabled SmartHours opt-in program offers four pricing levels for peak time periods on weekdays from 2 p.m. to 7 p.m. during the months of June through September, according to Senior Communications Specialist Christina Dukeman.

“Those are the times it costs the most to generate electricity and it’s therefore beneficial to engage customers in shifting their energy usage to off-peak times,” she said.

SmartHours customers’ off-peak rate is $0.05/kWh. Their peak period rates, communicated to customers a day in advance, range from the standard $.09/kWh to a high $.18/kWh and a critical $.42/kWh. A free programmable thermostat enables customers to manage usage.

The program was piloted in Norman, Okla., in 2011 and deployed across the OGE service territory in 2012, Dukeman said. Almost 15%, or 115,000, of the utility’s residential and business customers save an average of $200 per summer on the program.

The SmartHours program has produced an approximate peak demand reduction of 150 MW and allowed OG&E to delay building incremental thermal generation, she added. Because of its success, OG&E plans to continue promoting SmartHours and has introduced a marketing effort to increase the use of the programmable thermostat.

Demand charge rate design

As with TBR, demand charge rate design must also define the charge amounts, necessary technologies and how customers are to be enrolled.

Because of limited empirical evidence on demand charges, “most arguments remain speculative,” RMI reports. Despite that, “there is lively debate about how they may affect key desired outcomes, including customers’ peak consumption, total energy consumption, and acceptance,” the paper notes.

Four dimensions are likely to be most important, RMI notes. First, how costs are allocated “directly determines the magnitude of the demand charge price.”

Second, customer acceptance comes from whether customer peak coincides with system peak, whether customer peak is defined by the current billing period or the billing history, and what enabling technology is available.

Cost allocation can be narrowed to “the customer’s service drop and share of the line transformer” or it can be broadened to include “other capacity-related distribution costs,” RMI reports. A more extensive type of allocation includes all costs of system infrastructure built to meet peak demand out to marginal generation and transmission capacity costs.

After that, demand charge rate design should consider the drivers of peak demand and how effective cost recovery can be.

“Costs driven by customer peak demand can include building, operating, and maintaining capacity (e.g., generation, transmission, and distribution), but many other costs are unrelated (e.g., energy, metering, customer connection, etc.),” RMI notes.

The ability of customers to respond may be compromised, limiting cost recovery, if utilities shift variable and fixed costs into the demand charge simply to hedge against uncertainty in volumetric sales, RMI adds. To work for customers, the demand charge should “be restricted to peak capacity-driven costs (broad or extensive demand charges) and/or customer-specific capacity costs (narrow demand charges).”

If that is done, “higher demand charges would yield higher customer peak reduction,” RMI predicts.

Real world demand charges

Demand charges documented by RMI for range from as low as $1.50/kW to $13.50/kW.

Perhaps because they have typically been restricted to C&I customers, “no studies have evaluated the impact of demand charge magnitude on customer adoption or retention,” RMI reports.

The Black Hills Power’s $8.25/kW and Arizona Public Service’s $13.50/kW, both in place for over three decades, “have achieved relatively high enrollment compared to other demand charge rates,” RMI reports. The availability to both their rate bases of customer education and enabling load control technology have been factors.

The APS demand charge,, because it is designed to coincide with the system peak, is more likely to reflect cost causation and its one-hour measurement interval is more likely to reduce peak demand because it allows customers to more readily adjust consumption.

APS is pleased enough with its DC to wants to expand it, according to APS Spokesperson Jim McDonald. Utility data shows 90% of the 120,000 customers using the plan, which includes a TOU schedule, have reduced their rates. There has been a 33% increased opt-in by residential customers since 2010, McDonald reported.

Just this week APS filed a new rate case with the Arizona Corporation Commission that would make demand charges mandatory for nearly all residential customers. Regulators will evaluate that proposal in hearings in the coming months.

RMI: Time-based rates the more secure choice

Given the lack of study of demand charges, especially for residential customers, and the widespread acceptance of time-of-use rate structures, RMI researchers conclude TBR is the more sure choice for utilities in 2016.

TBRs “can reduce customers’ peak consumption and total energy consumption without compromising customer acceptance,” the report argues.

It can be empirically shown, it adds, that well-designed TBRs have the potential to reduce peak load and reduce total energy consumption with significant customer enrollment and retention rates.

On the other hand, “limited empirical evidence is available to provide insight on the efficacy or impact of demand charges on any desired outcome beyond cost recovery,” RMI reports.

The absence of evidence leaves the question of demand charge impacts on peak consumption, total energy consumption, and acceptance “largely speculative.”

RMI’s work clearly shows that residential TBRs can and will help to shift load into lower-cost hours and can ensure that customers with better load shapes will save money, said Jim Lazar, a senior advisor at the Regulatory Assistance Project (RAP), who has also written extensively on evolving rate designs.

Their paper also shows, he added, that demand charge impacts on residential customers "are largely unstudied and unknown but are likely to have significantly less effect on peak demand."

The RMI paper “well identifies the large variation in experiences, as well as the lack of information around impacts of demand charges,” agreed Makena Coffman, associate professor of urban and regional planning at the University of Hawaii and co-author of a recent paper on rates intended to help resolve net metering debates in Hawaii.

The choice between TBRs and demand charges is just the first step in an evolution from traditional rates to more sophisticated structures, RMI concluded.

“It is a complicated and dynamic calculation because rate design does not occur in a vacuum,” Cross-Call said. “Decisions about price levels and how rates are constructed and marketed determine how many people enroll and how they respond to the prices … design dimensions that will result in better customer response is a conversation that must be part of any rate case.”