Peak-load pricing – How can a regulator develop incentives to reduce peak demands, which would require capacity expansion?

[Response by Sophie Trémolet and Diane Binder, November 2009]

There is a general consensus about the fact that tariff structures matter for efficient pricing and investments1. Several options are proposed in the literature for optimum pricing and capacity, including marginal cost pricing, Ramsey pricing2cost allocation regulation3 and peak-load pricing. All options are based on the idea that pricing structures need to take into account underlying costs and that price mechanisms may signal the value that consumers attribute to further capacity expansion. An “average price” pricing system would have the effect of encouraging higher consumption during peak periods and lower consumption during off-peak periods — which producers and consumers do not want.

Peak-load pricing (PLP) is very similar to block rate tariffs, insofar as both methodologies differentiate classes of customers by their consumption pattern (volume or time of the day) and offer them different marginal rates. PLP can be useful when the marginal costs vary depending on when the service is used. For example in the telecommunications sector4, the operator builds its network with the capacity to serve peak demand. As a result, fixed costs are caused by peak demand. To facilitate marginal cost pricing, the operator would maximize profits by charging higher prices during peak hours so that customers would pay prices in line with the marginal costs incurred by their consumption. The effect of PLP is to induce some consumption to shift away from the times of peak demand, and toward times of lower demand. Such demand response takes place when a customer changes its consumption pattern in response to stimuli. Consumers are rewarded — in the sense that they pay less — for using the service when there is ample un-utilized capacity, rather than when demand takes up or even exceeds all the capacity. Utilities are rewarded because this system makes more efficient use of existing capacity and avoid capacity expansion when not needed in absolute terms.

Many countries have been adopting PLP and load control schemes since the California power crisis in 2000 and 2001. Peak demand grew in California much more than forecasted by the California Energy Commission, leading to exponential peak prices; however, because of retail price control, the increase in wholesale price was not matched by any increase in retail price for customers and had therefore no impact on reducing demand. Special demand management programs were implemented, including demand relief by which large users agreed to reduce their demand during energy emergencies, energy efficiency and conservation programs (Sweeney, 2002).

However, introducing PLP has some costs that need to be taken into consideration and must be weighed against the welfare gains of more efficient pricing. PLP requires sophisticated measurement of customer usage and advanced metering. Many utilities may lack information that allows differential pricing across periods of consumption and would therefore need to upgrade metering equipment so as to introduce PLP.

There are several alternative approaches to introduce incentives for reducing peak demands, which have been adopted largely in the electricity sector:

  • Two-part pricing system based on consumption: with this system, a fixed charge reflects capacity and distribution charges whilst a variable charge is based on on-peak and off-peak costs. This approach eliminates distortions caused by average cost-pricing, enabling customers to face the true costs of additional electricity purchase (Berg, 2006). Customers thus pay a share of overheads, and on-peak and off-peak prices are paid at incremental cost.
  • Two-part pricing system based on continuity: this methodology emphasizes the difference between two categories of peak demand: a demand that exists steadily over the year (for industry production for instance) and a specific demand that exists only during peak period (such as air-conditioning in the summer). Therefore the demand can be divided not on the basis of peak and off peak, but rather on continuity versus sporadic demand. An idea is that customers pay off-peak price during periods when the demand is continue; during peak months, customers pay the price for a quantity up to their average monthly consumption during off-peak period. The peak period price will be paid for any additional quantity bought. In this system, the additional demand does not affect the off-peak price and hence no subsidies are required (Peles, 1981).
  • Direct load management: As an alternative to PLP, direct load management, or rationing, is offered as a technique for dealing with peak usage (Berg, 1978; Peles, 1981). Each customer obtains a right to buy electricity up to a certain limit. A customer can raise the limit by buying in advance a right to buy additional quantity. The distribution company charges the same price for all KwH bought, but it has the right not to supplycustomer above the upper limit. In this way, customers pay in advance for a given capacity, while the electric company can control the peak demand by not supplying customers above a certain limit.
  • Load management via storage capacity installed at the customer’s location. During off-peak periods, storage is filled by the utility and then is drawn down by the consumer in the next peak-period. Utilities have an incentive to cover storage costs since this system allows savings in generation capacity costs.



  1. Refer to FAQ question: “Price Path – How does a regulator or a utility design a tariff structure that will gradually align prices with efficient costs over time?
  2. See Deviations from Marginal Cost Pricing: Ramsey Pricing in BoKIR.
  3. Refer to FAQ question: “Cost pass-throughs – Should bulk tariffs be transferred directly to final consumers through adjustment clauses or comparable pass-through mechanisms?
  4. Refer to BoKIR Peak-load Pricing.