Opinion, Berkeley Blogs

Measuring the economic costs of electricity outages

There’s nothing like a power outage to make you appreciate just how much you depend on reliable electricity. According to the numbers from the Energy Information Administration, as of Saturday, Nov. 3, nearly 2.5 million people in Sandy’s path were still without power.

Economists and policy makers have tried to measure the value of having electricity – plus the converse, the lost value from an outage – in several different contexts. None is entirely satisfying.

electricity outage map
Electricity outages following Hurricane Sandy

Regulators and utility planners bat around a term called the “value of lost load,” referring to lost electrical load. In principle, it’s the amount the average consumer would pay to avoid an outage.

Electrical systems in the U.S. are designed to sustain most unplanned downtime at electric generating plants and most downed transmission and distribution lines. For instance, when a nuclear plant trips unexpectedly, there are enough other plants that can be brought online to produce electricity to keep the lights on.

When utility planners and regulators decide how much to spend on building redundancy to weather unplanned events, they try to trade the costs of having excess capacity against the benefit of avoiding outages. Estimates of the value of lost load factor prominently in the benefit calculations.

Unfortunately, regulators are using value of lost load estimates that are based on dated and flimsy research. This piece explains many of the shortcomings in the studies that attempt to measure the value of lost load. One major flaw is that most values do not account for the duration of an outage. I was lucky enough to avoid the outages associated with Sandy, but I lived through three days without power (or running water, since our water pump was electric) when I was younger and the third day was definitely worse than the first.

Empirically, estimates of the value of lost load range widely, from $2 per kilowatt-hour to $20. Most regulatory decisions use something on the low end of that scale. Given that a typical household in New Jersey uses 23 kilowatt hours per day, and probably less than 20 kilowatt hours per day in October and November since no one is using air conditioning, these estimates suggest the average household without power suffered $40-80 of losses per day. That seems low.

Of course, even if the regulators used a very high estimate of the value of lost load, the cost-benefit comparison probably wouldn’t suggest that we should invest to protect our electrical system fully against a storm-of-the-century such as Sandy. It might, however, suggest that more lines should be put underground.

Another way to think about the value of electricity is to look at how people’s lives change when they first acquire electricity in their homes or communities. International lending agencies like the World Bank have funded rural electrification programs around the world for several decades. They are keenly interested in measuring how these programs contribute to economic development so they can compare investments in the electrical grid to other projects they support.

While an estimate of the value of electricity that is based on households in rural Kenya or Ecuador that have not previously had grid power may have little relevance to customers in New Jersey, the issues researchers confront in developing these estimates are revealing.

One difficulty in evaluating precisely how people’s lives change when they gain access to electricity is disentangling the impact of electrification from other changes happening at the same time. For instance, does electrification create more job opportunities in areas that already have paved roads? Or, might it look like that in the data as governments target road and grid development to areas that are developing rapidly?

Similarly, electricity outages in the aftermath of Sandy may be particularly costly as they delay storm repairs. The value of electricity depends on the context, which makes it hard to put a precise number on it.

Future research needs to deal with these difficulties because they drive important decisions about how to allocate society’s resources. How much should a place like the World Bank spend on rural electrification as opposed to building health clinics? How much should US utilities spend to prevent electricity outages, particularly during storms? Several recent studies have made some progress estimating benefits of recent electrification efforts in Brazil and South Africa, but there’s certainly room for more.

Cross-posted from Energy Economics Exchange (tag line: Research that Informs Business and Social Policy), a blog of the Energy Institute at Haas.