More reliable windy weather reports
Julia - Not knowing when the wind will blow and how hard is one of the main drawbacks confronting energy companies that rely on wind power to supply their customers. If they bank on seeing a certain level of supply, but a law means they're short of power, they have to make up for the shortfall by buying supplies off their competitors, often at shockingly high prices, undermining the viability of their business. But if we improve our short term weather forecasting for turbines so predictions about wind speeds and therefore outputs can be improved, the situation is much less risky for the operators who are more likely to invest. That is what Martin Shields from Colorado State University is trying to do as he explained to Chris.
Martin - Our problem is, how do we provide a better wind forecast to utilities so they can better plan for producing tomorrow's electricity demand.
Chris - That sounds like a bit of a trivial problem, though - it might not be windy tomorrow, we'll just get the electricity from somewhere else? Is that not how it works?
Martin - Well, sure, you can go out on the market and buy it from someone else, but it's a lot more expensive when you expected the wind to blow and it didn't. You have to go out and pay a lot of money for electricity on what's called the spot market.
Chris - Are our forecasting methods good enough to give that kind of fine grain detail, though? What sort of resolution of energy generation, and also wind speeds, are we talking about?
Martin - Well, we are looking at winds that are very localised. We need to know how the wind is going to blow near the wind turbines where the electricity is produced. That's a tough act for forecasters to do. It's really difficult to know how fast the wind is going to blow at any particular time in a very small locale.
Chris - So how are you going to solve that for them?
Martin - We work with scientists at the National Oceanic and Atmospheric Administration, part of the US government, and every day they're hard at work, fine tuning their models. What we do, in our project, is look at how the old model that's currently running does in predicting wind 12 hours ahead versus the developmental model, which is behind the curtains. 12 hours later, we see how good those predictions are. We look at what the cost of a mistake is under each of those regimes. What we found is that, as the model is improved by the scientists, there are substantial savings and that can subsequently be passed on to consumers through lower electricity rates.
Chris - Does this give generators more confidence, essentially, to invest in this and go down this path because rather than take the sure fire way of generating electricity, you burn some coal, or some gas, or some oil, you know how much that's going to cost you, you know that you're going to get electricity at the end of it, and therefore very low risk. Versus, if you've got to take a gamble on the weather, high risk. Whereas if what you are doing can be brought to bear and prove that actually they can save a few bob, it does mean they're more likely to invest in this sort of market.
Martin - Exactly. The uncertainty of wind is an impediment to its adoption. Sometimes it blows and sometimes it doesn't. But, also, the uncertainty about the uncertainty is an impediment. What we're trying to do is remove the uncertainty about the uncertainty, and that provides value to the utility and helps increase the likelihood that they'll use more of it.
Chris - Where are you basing the analysis because is this going to be applicable everywhere? And what about with an eye on the future? People are very worried about climate change and some of the predictions about the rate at which that is going to happen are quite scary because they suggest quite near term changes and they're very variable. There's a lot of those uncertainties that you're referring to geographically as well. Are we looking at an even more uncertain future and therefore how much certainty can you bring to the uncertainty at the moment?
Martin - That's a tough question, Chris. We're doing this for the continental United States and, even within the US, our wind forecasters do better jobs in some places than others. As we introduce climate change - we know it's happening, what we're trying to do is make this transition to whatever the future holds as costless as possible, or as least costly as possible.
Chris - Are utility companies receptive to this kind of thing? Can you see them actually having trust and confidence in the sorts of numbers that you are beginning to generate?
Martin - Yeah. We've talked to a few utilities who are clamouring for this information. It really is an expensive problem when there's a mistake made and anything they can do to keep rates down to consumers, they're excited about. We're just hoping that we can help in that effort. It might be a few dollars per year or a few pounds per year for the average customer but, when you add that up across the economy, it can run into the hundreds of millions of dollars in annual savings and that's money that can be spent on other things. People can have something a little bit more exciting in their lives than electricity.