Electrification was the single greatest engineering achievement of the 20th century, according to the U.S. National Academy of Engineering.
But buying electricity is not like buying a can of baked beans. Irrespective of how much power households and businesses use on any given day or month, they have come to expect that nearly unlimited supply will always be available at any time at the flick of a switch.
That makes pricing electricity particularly tricky. Nearly all the big questions facing the power industry in the next decade boil down to the question of how to value and pay for reliability during a period of unprecedented change in the way power is generated and distributed.
The costs of reliability are becoming much higher and more visible; it is as much a commodity as the power itself. The question is how to put a value and price on it? How much should customers pay to ensure power is always available?
How much extra should they be expected to pay to raise reliability by an order of magnitude, such as ensuring their power supply is 99.999% reliable, rather than just 99.99%? How many extra power plants should be built to limit the risk that customers are forcibly disconnected over the course of a year to one in 50, rather than one in five?
Electricity is a service rather than a product. Customers pay for continuous and ready availability. More than that, electricity is a utility. No great harm occurs if a supermarket temporarily runs out of baked beans or even bread. But even a brief interruption in the supply of power will severely inconvenience users and even endanger life.
"Most basic functions depend ... on a continuous supply of electricity. Without it subways and elevators stall, factories and offices grind to a halt, electric locks jam, intercoms and televisions stand mute, and we huddle without light, heat or ventilation," according to one report ("Brittle power: energy strategy for national security" 1982).
Unlike utilities such as water and gas, there is no economic way to store electricity in meaningful quantities. The amount of power generated must be nearly identical to the amount consumed at each moment of every day.
For other utilities, the water and gas stored in reservoirs, distribution facilities and pipelines provide a flexible buffer to help absorb momentary, daily and even seasonal variations in supply and demand. Power suppliers have no such luxury.
In spite of this, the power industry has achieved extraordinary levels of reliability. Most other industries cannot even come close. Excluding faults on local circuits, the reliability of supply on Britain's National Grid was 99.99954 percent in 2011/12 ("National Electricity Transmission System Performance Report" 2011/12).
That reliability is now coming under more pressure, and the costs of maintaining it are becoming more apparent.
Near-perfect reliability does not come cheap. Expensive transmission networks have to be built and maintained. Surplus power plants must be constructed and held ready to be brought online at short notice.
Some power plants must be kept below their maximum efficient level of generation ("part loaded") ready to ramp up within just a few minutes. Pump storage hydro plants and all sorts of gas and diesel turbines are also held in a high state of readiness to respond to any sudden imbalance between supply and demand.
In the past, the costs of ensuring reliability were hidden. In many places, integrated utilities were responsible for generation, transmission and distribution. The cost of reliability was bundled together with the cost of fuel and construction, transmission charges, meter reading and billing.
Now three decades of privatization and deregulation have blown apart the old cost-recovery model in North America and much of Western Europe. Utilities and transmission operators must now pay for reliability services separately.
The policy priority for more renewable generation is causing even bigger problems. Governments want a higher share of electricity to come from renewables such as wind and solar that are only intermittently available. Customers and power regulators want no loss of reliability. Power must still be there at the flick of a switch, whether or not the wind is blowing.
In the United States, Britain and elsewhere, grid operators buy not just units of actual power but also a variety of "ancillary services". The U.S. Federal Energy Regulatory Commission defines ancillary services as all those "necessary ... to maintain reliable operations of the interconnected system".
They include frequency response, reserve generation, voltage support, emergency assistance from other networks and "black start" capability to restart the network in case of a catastrophic loss of power across the entire grid.
In the year ended March 2013, Britain's National Grid spent over 800 million pounds ($1.2 billion) to buy ancillary services on behalf of the network and its users.
By far the largest component, 191 million, was paid to power plant operators for frequency response to help even out imbalances between supply and demand that lasted just a few seconds ("National Grid Monthly Balancing Services Summary" March 2013).
Another 121 million pounds was paid to generators to stand ready to provide extra power at 2 minutes notice. The grid also paid over 15 million pounds for black start capability in case the unthinkable happened (it didn't).
Britain has also proposed launching two more balancing services, the Demand Side Balancing Reserve and the Supplemental Balancing Reserve, to help keep the country warm and lit during the winters of 2014/15 and 2015/16, when spare capacity on the grid is otherwise expected to become very tight.
All these balancing services have to be paid for and will eventually be charged back to customers on their power bills.
OUTPUT OR CAPACITY?
In future as more wind farms are integrated onto the grid, the demand for balancing and other ancillary services is set to escalate sharply, and so will the cost.
Large numbers of fossil fuel plants will have to be kept under automatic governor control, part-loaded or in the short-term operating reserve (available within a maximum of four hours) to meet peak demand and to offset short-term variations in wind farm output.
To recover their construction and operating costs, these plants will either need to be able to charge very high prices for the small amounts of power they do generate or receive some form of capacity payment simply to remain available.
Policymakers in the United Kingdom as well as in California and other parts of the United States are all leaning towards capacity payments rather than relying on occasional sky-high power prices to provide the correct incentives to build spare capacity and maintain reliability.
But that raises the questions of how high the payments should be, and just how much spare capacity and reassurance customers can be expected to pay for. Should reliability costs be recovered on a per-kilowatt-hour basis, with customers paying in direct proportion to how much power they use. Or should some or all of them be recovered through a fixed standing charge?
Much of the interest in smart meters stems from the idea that they allow customers to reduce their demand to match the available supply, rather than pay for extra power plants that will be used only occasionally.
Either way, the costs of balancing and other ancillary services are set to become larger. The industry and regulators need to engage customers in a frank discussion about how much reliability they want and how much they will pay for it.© (c) Copyright Thomson Reuters 2013.