Smart Meters: an Electrifying Idea

Commentary, Energy, Frontier Centre, Worth A Look

Ontario Energy Minister Dwight Duncan recently spelled out the elements of a new energy policy that he said would put the province “back on solid footing by taking a balanced approach.” Included among his proposals was establishing a pricing plan for small consumers that would ensure they “can take advantage of time-of-use rates so that they would have the opportunity and incentive to shift consumption from periods of high demand and prices to periods of lower demand and prices.”

Premier Dalton McGuinty indicated that the Ontario Energy Board will be charged with developing “a plan to install a smart electricity meter in 800,000 Ontario homes by 2007 . . . and in each and every Ontario home by 2010.” These smart meters, “combined with more flexible pricing,” would provide an economic incentive for consumers to avoid energy use during peak periods when generating costs are much higher.

By shifting electricity use to off-peak periods, many consumers can reduce their energy bills. More importantly, the collective response of many consumers can reduce the need for new power plants and transmission lines, dampen price volatility in energy markets and improve the environment.

These are good ideas, supported by a 30-year history of electricity-pricing studies that indicate without a doubt that consumers can and will shift load in response to time-varying electricity prices. The most recent example is from California, where supply shortages similar to those in Ontario have led regulators to implement the largest pricing experiment ever conducted among homeowners and small businesses to estimate the impact of time-varying prices on energy-use patterns.

The California Statewide Pricing Pilot (SPP) is testing a variety of pricing options, including traditional time-of-use rates, as well as critical-peak pricing. Critical-peak prices incorporate a dynamic component into the tariff by signalling relatively high peak-period prices to consumers only on a small number of days when market conditions are particularly dire. These days are not known until the day before they occur, at which time consumers are notified that the next day is a critical day when electricity will be more expensive than normal. Analysis of data from the first summer of the California experiment indicates that these “rifle shot” price signals can be both popular and very effective at reducing peak demand, thus dampening wholesale prices and helping to avoid the need to construct relatively inefficient peaking generators.

While evidence from the SPP (and many other pricing experiments that preceded it) supports the Ontario government’s interest in smart meters and more economically rational pricing, it also identifies a number of pitfalls that the government must avoid if the benefits are to exceed the substantial investment and operating costs associated with wide-scale deployment of the meters.

One of the most important lessons from the SPP and other pricing experiments was perhaps summed up best by Tom Adams, executive director of Energy Probe, when he was recently quoted as saying, “If we’ve got smart meters but dumb prices, we’re not going to get the benefit of these better meters.”

We have found that a relatively high price signal during peak periods is generally needed to induce sufficient load reduction to offset the cost of the meters. A muted price signal offers neither sufficient opportunity for consumers to reduce energy bills, nor sufficient reduction in peak demand to defer power-plant construction.

Another critical lesson is that, if asked to volunteer for a time-varying rate, the vast majority of consumers refuse. However, if they are placed on a time-varying rate as a default and given the opportunity to choose a flat-rate option, the vast majority stay with the time-varying option, and many adjust their behaviour to take advantage of the bill-reduction opportunities that time-varying price signals provide.

In other words, the Ontario government cannot simply require that smart meters be installed in homes and small businesses and expect that enough consumers will voluntarily sign up for time-varying rate options to offset the meters’ cost. The government will need to boldly change default prices to time-varying rates, while providing consumers with the choice of a flat-rate option that fully reflects the cost of providing this option.

A third issue that the government and the Ontario Energy Board must examine is how to implement the rollout of smart meters in such a way that the investment and operating costs will not exceed the benefits of the pricing options that the meters allow.

The Ontario Energy Ministry says the meters now cost $400 a unit, but it estimates the cost will be closer to $100 when more companies start producing them. We have examined the cost-effectiveness of smart meters combined with innovative pricing for several U.S. utilities, as well as for Victoria, Australia (where the government is contemplating a policy like Ontario’s), and Singapore. Based on this work, it is likely the metering costs will exceed the benefits if the cost is anywhere near $400. Done correctly, however, the cost per customer based on a mass deployment of smart meters could be closer to $150. At that price, it is much more likely benefits will exceed costs.

While substantial evidence from California and elsewhere supports the idea that pricing can instill a conservation culture, pricing options can only be offered in conjunction with smart meters. The devil is in the details: Pitfalls in implementation could easily turn a good concept into a bad investment. Very careful analysis and planning of implementation options will be required if the government’s bold vision is to produce the desired result.

Ahmad Faruqui and Stephen George, economists with the Oakland, Calif., office of Charles River Associates, were instrumental in the design and analysis of the California pricing pilot.