Robin Khan emailed this photo to me, it looks very much like a Marbled Duck (teal) but there does seem to be a few anomalies:
1. The scapulars and mantle feathers should really be well marked with diffuse but obvious pale spot at the tip of each feather, this bird appears to be have neat(ish) pale fringes.
2. Speaking to John Waldon apparently the bird may be as big as a pintail , which would be too big for a Marbled Duck
3. Perhaps the flanks are too scalloped rather than spotted?
4. the tail may be too pointed, difficult to tell from this picture - but Marbled duck does have a long tail.
The crest on the back of the head is fine for Marbled Duck as is the bill length. The it would be good to have some feedback from field views especially any flight views of the wing pattern - upper and under.