In any case, the factor of at least 6 in energy consumption per passenger-km between planes and trains boggles the mind. Those whom the Gods wish to destroy They first make mad. -- Euripides
To recapitulate: because you have to consider territories affected in a half kilometre radius anyway, and because the site of the best wind is not necessarily along the railway, and because you have long transmission distances along the railway line anyway, railway-feeding wind farms might as well be built as normal wind farms somewhere near the line. Railways might opt to build wind farms, and use as much electricity as their wind farms produce, thus not having to spend on buying net electricity elsewhere, but that net zero would cover having to get electricity from the grid and sending surplus to the grid depending on intermittance. It's best if what they buy is regenerative energy too, and it's best anyway if non-rail users have regenerative sources too, so why not cut it short and create a wind-et-al-fed grid which feeds rail too. *Lunatic*, n. One whose delusions are out of fashion.