Households face power-pricing revolution
Households in the United States and the United Kingdom are about to experience a revolution in the way they pay for electricity.
Over the next decade, almost all homes will be fitted with “smart meters” recording the time as well as the quantity of electricity used. Most customers will face some form of dynamic pricing that relates the price they pay for each kilowatt hour (kWh) to the actual cost of generating it.
Smart meters and dynamic pricing are critical to using the generation and transmission system more efficiently while accommodating a growing share of renewables (wind, solar) on the grid without sacrificing reliability.
Power cannot be stored, and the amount demanded by customers (“load”) is highly variable, so system operators hold large amounts of generating capacity in reserve to cope with demand peaks or outages when generating units become unavailable.
Many generating units must be built and maintained even though they may only be used for a few hundred hours each year. The greater the variability in load the more idle capacity has to be maintained. In general, usage is higher during the day than at night, and higher in summer than winter, owing to increased airconditioning demand.
The problem will get worse over the next decade as the share of generation from renewables such as wind and solar, which cannot be scheduled in advance, increases. Even more back-up capacity will need to be held in reserve in case renewable power is not available at peak times.
CAISO LOAD CURVE
The attached chart shows total demand across the power grid run by California’s Independent System Operator (CAISO) over the last twelve months. It covers about three quarters of consumption across the state.
While “average” load is around 27,000 MW, load rises sharply during daytime hours in the summer when airconditioning demand is highest.
On almost all hours throughout the year (more than 98 percent) load is less than 40,000 MW. But on a very small number of hours (less than 2 percent) load increases substantially, sometimes as high as 45,000 MW. Enormous amounts of capacity must be kept in reserve to meet the peak demand experienced for less than 180 hours a year.
Even the distribution’s tail understates the amount of capacity kept in reserve, since the system is designed to operate with a safety margin, ensuring peak demand can be met even if one or more large power plants becomes unavailable due to scheduled maintenance or a fault.
The safety margin is set so the system will be unable to meet peak hourly demand no more than nine times a century. If peak demand is 45,000 MW, the system needs to carry a substantial cushion of spare capacity above this level.
The safety margin is normally set to at least 10 percent. So CAISO needs to maintain access to more than 50,000 MW of capacity (45,000 MW peak load plus 10 percent), even though the system “typically” needs to meet only about half that power demand.
If CAISO could reduce demand on just the 2 percent of peak hours each year, it could avoid up to 5,000 MW of load and the need to maintain more than 5,000 MW of idle capacity.
The load curve’s steep tail explains why demand response strategies designedto curb power use at peak periods have become one of the highest priorities for governments and system engineers on both sides of the Atlantic.
At the moment, demand response is largely restricted to industrial power customers, especially those with large heating and cooling loads, many of whom are on interruptible contracts allowing the system operator to order them to shed load for a period from 30 minutes to several hours at peak times, in exchange for rebates or a lower tariff.
In future, officials hope households can be brought within the demand management system by fitting them with smart meters that will vary prices significantly at peak times and/or allow the utility to reduce non-sensitive loads such as refrigerators and airconditioners when the system is stretched.
Under existing plans, most residential customers in California will be fitted with smart meters by 2013. Britain’s Department of Energy and Climate Change has announced plans to have all households fitted with smart meters by 2020.
Smart meters will record consumption in five-second or hourly intervals, and be capable of transmitting the data back to a central control and billing centre. Each meter will have a display unit showing both usage and current prices.
Meters will be capable of communicating with programmable thermostats and other household electronic devices. In theory, control devices could be instructed to turn up airconditioning systems by a degree or two or shut down non-essential loads when prices hit a critical level.
Once meters have been rolled out, households will be shifted from fixed-rate tariffs to variable ones. Current systems allow customers to “opt-in” to variable pricing but there is likely to be less choice in future.
The simplest time-of-use (TOU) tariffs would establish a relatively simple system varying prices by time of day, day of week, and perhaps season of the year. Prices would be fixed in advance. The intention is to encourage households to “load shift”, moving as much consumption as possible from peak daytime periods to the night-time when there is plenty of spare capacity available.
TOU tariffs have already been implemented for many industrial and residential customers. But smart meters would enable much more radical reform.
In the most ambitious system, customers would face real-time prices (RTP) linking the price they pay directly to hourly prices in the wholesale power market.
In an intermediate system, critical peak pricing (CPP), customers would pay very high prices for a small number of peak hours each year — perhaps no more than three hours a day for a maximum of ten days each year. These critical periods would be announced a day or so in advance. The rest of the time, customers would be on a standard TOU tariff.
The aim of both RTP and CPP systems is to provide customers with a strong incentive to cut all avoidable demand during the few hours when the system is most stretched.
While customers would face high prices at peak periods under all three systems, overall bills could be lower if generating capacity can be used more efficiently and the need to maintain so much idle capacity in reserve is reduced.
Much would depend on how savings are distributed between customers and shareholders. But the potential is enormous.
Peak load across the United States is currently 810 gigawatts (GW) and forecast to grow by an average 1.7 percent a year to 950 GW in 2019. But in a report for the U.S. Department of Energy, the Brattle Group estimated 2019 peak demand could be lowered by between 82 GW and 188 GW through the wider deployment of demand response linked to smart metering and dynamic prices.
If demand response technologies were deployed in almost all homes across the United States, coupled with a strong dynamic pricing system, almost all forecast peak demand growth in the next ten years could be avoided.
Such an ambitious demand response program would avoid the need for a huge amount of new generation and transmission infrastructure, while making it easier to accommodate variable sources such as wind and solar by matching them with increased variability in demand.
Full demand response is unlikely to be achieved (or be cost-effective). Policy will instead blend demand response with more investment in renewables and construction of additional gas-fired generation and transmission capacity to provide back up.