The paper's here, if you want to see how they came to the result. The PDF is paywalled, but you can view it in their online reader for free. https://www.scienc

The paper's here, if you want to see how they came to the result. The PDF is paywalled, but you can view it in their online reader for free. https://www.sciencedirect.com/science/article/pii/S096014812100344X https://www.sciencedirect.com/science/article/pii/S096014812... Properly understanding it is well beyond me, but it looks like it's about carefully arranging the turbines with respect to each other's wakes. While this is based on modeling and not empirical experimentation, it doesn't look like they just pulled the idea out of a hat. Getting the 15% improvement seems to require that the pattern is optimally oriented with respect to the wind direction, so I'm guessing real world benefit wouldn't be anywhere near that great.