In a comment below this article, 'jjk' asks a very good question. He expresses it politely but I can be more blunt. If wind is providing a large percentage of total power, it will probably be when the demand for power is low at night. How do I know, he asks, that the effect I observe isn't simply a reflection of the fact that high percentage wind tends to occur at low demand moments, when the price tends to be low anyway?
I wish I'd thought of this before because I think 'jjk's suspicion may be partly correct. If, instead of looking at the percentage of demand provided by wind, I examine the correlation between the absolute amount of wind power in each half hour (MW not percentages) and the National Grid's buying price (the 'System Buy Price), the correlation is less clear-cut, though it still exists.
In other words when there isn't much wind, the average price that National Grid has to pay to buy electricity is higher (£54.50 per megawatt hour) than when the wind is strong (£49.90 per megawatt hour). Closer examination of the results also shows (not noted in the table above) that at the very highest levels of wind output the price tends to rise slightly.
When there is less than 500 MW of wind, the price averages £55.8, about £3.1 more than the average price in the table above. This contrasts with the £7 difference I estimate in the main body of the article
You can argue that this calculation is in fact too harsh for a reason that works in the opposite direction to 'jjk's hypothesis. High levels of wind output tend to occur in the winter, when the price of power is typically higher. So I think the table above probably sets a lower limit on the impact of wind on power prices.
On average, I think we can say, wind makes a difference to UK power prices for immediate delivery of somewhere between £3.1 and £7 per megawatt hour. Even at the lower level this washes away a large fraction of the consumer subsidy for wind.
Original article follows
You can argue that the subsidy for wind power in the UK costs the country almost nothing. The reason is that when the wind blows, wholesale electricity prices are lower than they would otherwise be. On the typical day when wind is producing about 7% of the UK’s electricity, the market price of power is about £7 per MWh less than when the air is completely still.
If, on average, wind power depresses the wholesale price of electricity by £7 for each megawatt hour consumed in the UK, the total impact over the year is about £2.3bn.  The total subsidy for renewable electricity paid by electricity consumers this year is capped at £3.3bn. This includes solar and other technologies such as landfill gas, not just wind. The subsidy cost for wind - about £2-£2.5bn - may well be less than the downward impact wind has on electricity prices. The net impact on consumers may therefore be close to zero. In effect, the whole burden of wind subsidy falls on the fossil fuel generators because they obtain lower prices than they otherwise would.
There's one obvious objection to this glib analysis: most electricity isn't traded just before it is needed. Much of the power that drives your home or office has been bought or sold months, or perhaps years, in advance. However, in the long run lower prices for immediate delivery seep through to the wider market. If you were an electricity retailer and noticed that power traded at perhaps £45/MWh just before it was needed, would you buy electricity months in advance at a higher level?
How do I get to the conclusion that wind depresses prices by an average of £7/MWh? Every half hour, National Grid has to balance the electricity market by buying or selling electricity. Total generation must match total demand, including losses in transmission, or otherwise the voltage on the UK network would move outside the strict limits that are set. The price that National Grid pays or the price it receives in the open market is recorded.
Also recorded each half hour is the amount of wind power that is generated by the major wind farms as well as the output of all other power stations. We can easily calculate the percentage of total generation that is provided by wind farms, offshore and onshore. Then I drew a graph that compares the price the National Grid paid to buy electricity with the percentage of the UK’s power provided by turbines.
This is what the chart looks like for the period between June 2012 and 19th January 2015 (the middle of last week).
This graph summarises 45,000 lines of data or records from almost a thousand days. The trend is clear: when the wind is hardly blowing the typical price of power that the National Grid faces is about £58/MWh and it falls as wind power increases. On those relatively few occasions that wind is providing more than 20% of electricity, the price is about half this level. On the average day over the last year, the main wind farms give us about 7% of total power needs. The typical buying price at 7% wind power is £51 per megawatt hour, £7 lower than when the wind isn’t blowing at all.
Has this trend varied year by year? A little, but the basic pattern is the same. When the wind is turning turbine blades, wholesale prices are relatively low. (There are only 19 days data for 2015 so we shouldn’t take much notice of this year’s figures).
As the number of wind turbines rises, we’ll see more and more days when this source of power rises to 20% or more of total UK generation. If current trends persist, this will take the price of power down to £30 or below. This is, of course, is exactly the phenomenon we see in Germany today, with prices often going close to zero or below on high wind days.
If the numbers in this note are true, they suggest that the subsidy paid to wind is balanced by a lower wholesale price in the electricity market. That’s the good news for consumers, and largely reflects the balance of supply and demand in the UK electricity pool. More wind means fewer high cost generators have to be incentivised to enter the market by greater than average power prices.
However the effect of this shift in the relationship of supply and demand had profound consequences for the profitability of fossil fuel generators. At the moment low power prices mean that many gas-fired power stations aren’t covering their full costs. And, as a natural result, few investors will want to build new fossil fuel plants. This is no bad thing, you might say, but it does mean that without massive intervention – in effect a renationalisation of energy generation or a guaranteed price for electricity – the rising number of wind turbines will inevitably destroy fossil fuel generation and eventually produce a highly unstable electricity market. This sounds an obvious point but it seems ignored by policymakers (and indeed by protagonists of renewable power).
Note about method
National Grid Buy and Sell prices are not precisely the same as 'market' prices. Each half hour, the Grid works out what demand is likely to be and what each generator has said it will produce. Then it estimates whether the whole UK system is likely to be in deficit or surplus of power. If the position is a deficit, it buys additional electricity to balance supply and demand. If there's a surplus, it does the opposite. When it is buying, it directly sees the prices but it doesn't exactly know what it would get if it were selling electricity. So it uses an estimate from the electricity market.
In a separate analysis, I have also looked at the System Sell price and the impact of different levels of wind generation. The curve is the same. When the wind isn't blowing, prices are about twice the level when wind is generating 20% of the UK's need. And, perhaps importantly, when wind generation rises above about 12% of UK generation, the price that National Grid obtains for the surplus electricity begins to fall sharply. We've seen, for example, several instances in the last few weeks of near-zero selling prices. In other words, in order to get someone to buy greater volumes of power National Grid had to accept very low prices indeed.
 Assumes total UK net generation plus imports less exports equals about 330 TWh.
 These are called the System Sell and System Buy Prices or SSP and SBP.