Statkraft, a solar farm operator, guarantees to its customers a certain Performance Ratio (PR) per site. This guarantees that the amount of energy delivered will not fall below a percentage of the system’s theoretical maximum output under ideal conditions. The company must pay heavy penalties (sometimes totaling in the £millions) for every percentage point if it falls below the agreed performance level.
The PR guarantee for each site is based on simulations using Typical Meteorological Year (TMY) data. One challenging factor of this is clipping - which occurs when the power generated by Photovoltaic panels exceeds the inverter’s maximum capacity, causing the system to cap or "clip" the energy output. While this optimises the system for average conditions, excessive clipping during unusually sunny periods can cause PR metrics to misrepresent actual performance.
Statkraft wanted to develop a reliable and accurate way to remove the effect of clipping. Initial approaches involved complex calculations of energy lost from clipping and using that to adjust the PR calculation. “This risked masking data that might provide important information, such as inverter derating - the intentional reduction of an inverter's capacity during higher operating temperatures to enhance its efficiency and longevity,” says Elena Koumpli, who heads up the Performance Analysis Team at Statkraft.
Aiming to find a clipping formula that was both reliable and simple for customers, Statkraft turned to NPL via the Measurement for Business (M4B) programme.
NPL developed a model of a typical solar farm based on Statkraft’s portfolio and used synthetic weather data to simulate the system under various realistic weather conditions and climates. The team then applied statistical analysis to the resulting data to establish the level of sunlight that caused the system to cross the clipping threshold. This was used to create a new equation for defining Performance Ratio which was simple enough to be put in contracts and easily calculated and evaluated.
The standard PR equation is: Energy output / expected energy output (under perfect conditions). Under the new system, on days when the levels of sunshine crosses a certain predefined threshold, a new calculation kicks in, capping the expected energy output to align more closely with the operational limits of the inverters.
“Essentially, we’re capping clipping based on a threshold for sunlight, rather than a threshold on the system output,” says James Blakesley of NPL. “This gave us a PR metric that more accurately compensates for clipping than using output data, has no unintended consequences, and works consistently for any system, anywhere in the world”.
By adjusting the PR calculation, Statkraft has a fairer way to assess performance, which will help it avoid being hit with unfair penalties when the actual weather dataset comprises many sunny days, especially during PR testing.
“Missing the PR guarantee by just 0.5% could lead to a penalty of £100,000 for a project valued at £50 million,” says Elena. “For a portfolio of over 1GW that could mean penalties of several million. That is pushed even higher where actual weather datasets have much higher solar radiation values than TMYs, and we do see differences as high as 8%, depending on location and year.”
The calculation is also helping with future project design. A recent solar deployment had its test phase during a particularly sunny week. The clipping calculation was essential in providing accurate data, without which it would have failed to meet its PR KPI, despite performing well.
“The fewer risks of falling below the PR due to non-predictable factors the better for our business and reputation” says Elena Koumpli. “And the fewer penalties we have to pay, the more money we have to invest in new projects, speeding up the rate at which we can deploy the renewable energy that is critical for transitioning the world away from fossil fuels”.
NPL and Statkraft are collaborating to incorporate this clipping correction into future industry standards for solar PR, providing all solar operators and customers with fairer, more reliable performance metrics. More reliable KPIs provide greater confidence in the operation of solar farms, reducing risks and encouraging further investment.
NPL helped us define a more accurate Performance Ratio metric which will help us - and the rest of the Solar industry - avoid expensive penalties where calculation inaccuracies could incorrectly imply we have missed targets. Having NPL behind the new metric ensures that we are doing right by our customers, and gives our customers confidence that the resulting equation is a better way to get a reliable picture of solar PV performance
Elena Koumpli, Head of Performance Analysis - Statkraft