A friend who is an airline pilot tells me that the addition of winglets to the wingtips tips civil airliners reduces fuel consumption by 5%. Now the fact is that all modern airliners do sport winglets, and the airline operators are not going to add them just to make their planes look good. They are designed to reduce the tip vortex, which is a cause of induced drag.

I can see no reason that the same considerations should not apply to wind turbines. Has the addition of winglets to turbine blades been tried already? Is there some fundamental reason that it doesn't work with turbine blades? It is hard to see what this might be.

It could even be that reducing the tip vortex would reduce bat kill from smaller turbines, since bats might be mistaking the vortex for clouds of insects.

So I took this question to the Wind energy group on LinkedIn, and this is what I picked up:

There are winglets on Siemens 2.3 MW turbines.   They have been introduced in Enercon turbines in different shapes since 1994. If you see the good CP values ( up to 50 %) of this blade design and the low noise values of this turbine types the application of winglets make sense. One  figure given is that winglets increase yield up to 6 %.

Efficiency is very important in wind energy because 1 % yield plus makes approx. 3-4 % revenue plus in wind economics.

Winglets make sense and do improve performance on turbine blades, in a similar manner as on an aeroplane wing, so a number of turbine manufacturers do include them now, though the design is somewhat different to those on aeroplane blades. 

Historically winglets were not included because the flow around a turbine blade tip is highly complex, i.e. due to the radial and turbulent components being relatively large. Another point would be that aircraft blade lengths are constrained by airport infrastructure, so you need to get maximum performance from a fixed blade length. Windturbines on the other hand can achieve improved performance by simply making the blades longer, which might well be the more cost-effective.

Noise reduction is a side effect of good controlled vortex from the winglets. There is a general rule among blade experts, if you hear one blade to be noiser than the others (lightning damage or equal deformation close to the blade tip) you lose approx 1-2 % yield. 

I estimate the effect of winglets to be 1.5-3 dBA in noise reduction . 

Approximately 80 % of the total noise of a turbine is erodynamic born.

So thanks to all on the list who shared information. Very shourtly after setting the question, I noticed winglets on a turbine visible from the Paddington-bristol rail line, ? at Didcot.
 

So wind turbines generally have a 5% inprovement in efficiency to come.

Incidentally, there was an anti-wind troll on the list, who seems possibly to be thought disordered. I am impressed by the frequency anti-wind and climate change "sceptics" show this feature.

The key significant accusation they bring about wind is the low load factor - the poor match between production and consumption. The answer to this is to increase power spread through HVDC lines, and power storage capacity through pumped water, compressed air, batteries (distributed as electric vehicles) and through electrolysis.