A solar park sells in the hours the sun shines, and those are the hours everyone else's panels are selling in too. The capture rate measures how much that costs.
The capture price is a generator's revenue divided by the MWh it produced: the market price weighted by its own output rather than by the clock. The capture rate is that price divided by the baseload average. An illustrative 100 MW solar park earning 9.76M on 153,500 MWh captures 63.58 per MWh against a baseload average of 85.66, a capture rate of 74.2 per cent.
Worked in full in Energy Trading by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The year's 8,760 hours are grouped into five rows, each with an average price and the park's average output in those hours. Grouping loses detail but keeps the mechanism visible; a real calculation runs the same formula on every hour. Prices are per MWh and illustrative, not a forecast for any market.
| Row | Hours | Price | Average output, MW | Generation, MWh |
|---|---|---|---|---|
| Midday, high sun | 1,200 | 48.00 | 70 | 84,000 |
| Daylight shoulders | 1,500 | 78.00 | 35 | 52,500 |
| Low sun, morning and evening | 1,700 | 96.00 | 10 | 17,000 |
| Evening peak, dark | 1,200 | 120.00 | 0 | 0 |
| Night | 3,160 | 85.00 | 0 | 0 |
| Year | 8,760 | 85.66 | 153,500 |
The park generates 153,500 MWh, a capacity factor of 17.5 per cent. The baseload price in the total row is the time-weighted average of all 8,760 hours.
Baseload price = Σ(hours × price) / Σ hours = 85.66
Revenue = Σ(hours × output × price) = 9,759,000
Capture price = revenue / Σ(hours × output) = 9,759,000 / 153,500 = 63.58
Capture rate = capture price / baseload price = 63.58 / 85.66 = 74.2%
Excel, with hourly columns: =SUMPRODUCT(Output,Price)/SUM(Output) for the capture price, divided by =AVERAGE(Price).
Row by row, the midday hours produce 54.7 per cent of the year's output at the year's lowest price and earn 4.03M. The shoulders produce 34.2 per cent and earn 4.09M. The low-sun hours produce 11.1 per cent at 96.00 and earn 1.63M. The two rows with the highest prices, the evening peak at 120.00 and the night at 85.00, produce nothing. The park sells most of its output where the curve is cheapest and none where it is dearest, which is why its price is 22.09 per MWh below the average.
Value the same 153,500 MWh at the baseload price and the revenue is 13.15M instead of 9.76M. The difference, 3.39M a year, is the shape discount: a covariance between the park's output and the price, not a view on the price level. It is 34.7 per cent of the true revenue, and it sits in every model that multiplies a generation forecast by a baseload forward curve.
The capture rate is not a property of the panels. It is a property of the market the panels sell into. Every new solar park in the same zone adds output in the same hours and pushes the midday price down further, so the rate moves with the fleet, not with the site.
Cannibalisation shows up in one row. Hold everything else and move the midday price: the baseload average barely moves, because those are only 1,200 of 8,760 hours, but the capture price moves hard, because those hours carry more than half of the park's output.
| Midday price | Baseload average | Capture price | Capture rate | Shape discount | Revenue |
|---|---|---|---|---|---|
| 60.00 | 87.31 | 70.14 | 80.3% | 17.16 | 10.77M |
| 48.00 | 85.66 | 63.58 | 74.2% | 22.09 | 9.76M |
| 40.00 | 84.57 | 59.20 | 70.0% | 25.37 | 9.09M |
| 30.00 | 83.20 | 53.73 | 64.6% | 29.47 | 8.25M |
| 20.00 | 81.83 | 48.25 | 59.0% | 33.57 | 7.41M |
| 0.00 | 79.09 | 37.31 | 47.2% | 41.78 | 5.73M |
A midday price falling from 48.00 to zero takes 6.57 off the baseload average and 26.27 off the capture price. That asymmetry is the whole risk in a merchant solar case: a forward curve can be right about the baseload level and still overstate solar revenue by a widening margin every year.
Suppose a buyer offers a pay-as-produced PPA at 70.00 per MWh. Quoted against baseload it looks poor: 81.7 per cent of 85.66. Quoted against the capture price it is 6.42 per MWh better than selling merchant, about 0.99M a year on this output. The buyer takes a 15.66 discount to baseload for absorbing the shape, but the shape is worth 22.09 at merchant prices, so it is paying 6.42 per MWh above the market value of this profile: a premium for price certainty, not a bargain on the power. Always compare a PPA price with the capture price of the profile it covers, never with the baseload forward.
The frequent error is the wrong average. Three averages look plausible and only one is right. The simple average of the three producing rows is 74.00. The time-weighted average over the 4,400 daylight hours is 76.77. The generation-weighted average is 63.58. The first two ignore that output is concentrated in the cheapest hours, and both overstate revenue by more than 15 per cent. The same weighting error, applied to a retail customer's consumption rather than a generator's output, is the subject of how much of a fixed-price retail margin the load shape takes; for a dispatchable plant, which chooses its hours instead of having the weather choose them, see how to calculate a clean spark spread.
The price duration curve, the averages over different hour sets and the shape multipliers are live in the free workbooks for the book's case, where the same weighting is applied to a plant and to a retail load.
There is no universal benchmark, because it depends on how much solar the market already has. The rate falls as midday prices are pushed down. In this illustrative case it is 74.2 per cent at a 48.00 midday price, 64.6 per cent at 30.00 and 47.2 per cent if midday hours clear at zero. Track it against the share of solar in the market, not against a fixed number.
No. The capture price weights each hour by the MWh produced in it, and output is highest when prices are lowest. Averaging the three producing rows gives 74.00; weighting them by hours gives 76.77; weighting them by generation gives the true 63.58. Only the last one multiplies by output to give revenue.
A pay-as-produced PPA fixes the price on whatever the park generates, so the buyer takes the shape. At 70.00, a PPA struck at 81.7 per cent of baseload, the generator earns 6.42 per MWh more than its merchant capture price, about 0.99M a year on 153,500 MWh. Its 15.66 discount to baseload is less than the 22.09 the shape costs at merchant prices, so the buyer pays a premium for price certainty.
This article is one calculation from Energy Trading. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: infrastructure, data centres and energy → · All 453 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.