Photo Of The Day

We had gusts over 100. various roads closed.

House over the road lost its oil tank and several will need new rooves.

Its still blowing.

@mdcc, true, but I'd say they were a factor.

@mdcc, true, but I'd say they were a factor.

Based on what evidence?

Not so windy yesterday, turbine was fine. Windy as fuck today, not fine.

Circumstantial I know, but i'll live with that.

But you'd design for 200mph, given that >150mph is a normal feature of Scottish weather, i.e you'd expect it to happen multiple times during the life of a wind turbine in that location.

150mph is not a 'normal feature of Scottish weather'.

Turbines are designed to shut down in winds over 25m/s (~55mph), which in a 25 year wind speed distribution of a pretty windy place* in Scotland, amounts to a very few percent of the year. Winds over 100mph are significantly rare. Wind turbines are designed to withstand the 100yr extreme 3s wind gust - which can cause resonance in the tower or blades. Something clearly went wrong with that turbine - not necessarily caused by the wind, but it won't have helped.

*the location of Scotland's most powerful wind farm (at the time of it's construction), which had a mean annual wind speed of 10.9m/s.

It's facing the opposite way to the others too, don't they turn into the wind automatically?

They do. Looks like the yaw motors have failed, if not completely then the power to them clearly has, and it has swung downwind.

Or the wind has changed direction since that turbine stopped, which could have been some time before the photo.

Seagull.

(Gravel looks shopped in gun pic)

150mph is not a 'normal feature of Scottish weather'.

Top gust in this storm was 165mph, previous high was only ~15 years ago at >170mph. In terms of the design life of wind farms, I'd call that a normal feature.

If I were trying to set fire to a turbine without physical access to the head, I'd send Stuxnet into the controller and get it to brake the shaft before feathering the blades. I'm going to assume that the standard controller has an interlock to prevent this, since it would be a good way of generating a lot of heat in the turbine head. If there is a bug in the software or the shut down sequence is manually controlled and somebody fucked up, I'm sure a fire would be a likely outcome.

A normal feature of the design life of wind farms, or of Scottish weather, which are you choosing?

Either way, the recent gusts are the highest recorded, and in any case not 'normal'. They are 'extreme'. And yes, wind farms have extreme design standards, but this is only one turbine that has caught fire amongst rather a lot installed in Scotland, so I would suspect a little more complex. And there is no proof yet (no doubting that there might be), that it is due to the extreme wind.

Design life of a wind farm is 25 years (typical UK planning permits only 20yrs in place), but they are designed to withstand the 100yr max 3s gust. Hell, we may have just had it!

Stuxnet - and I haven't come across it directly - appears to hit Siemens Industrial Software (thanks, wikiP). Ardrossan, where this turbine is, consists of Vestas turbines - they may have some Siemens generators, but they will* not* be running any Siemens software.

It would be pretty fruitless to send a computer virus to try and overrun a wind turbine, there are too many physical sensors that would trip it out in the event of an IT override. In this case, one of the physical sensors would have failed, which is a rare event.

Oh sorry i seemed to stumbled into the wind turbine accident investigation squad meeting.

:-) If I did a photo of the day today, it would be dodgy no-flash photo of scrummy nosh from The Fat Duck ... need something more topical ... please!

A normal feature of the design life of wind farms, or of Scottish weather, which are you choosing?

Normal for Scotland, and therefore normal for a wind farm sited on top of a hill in Scotland

Either way, the recent gusts are the highest recorded

Unless you have more data than I already alluded to, the recent storm hit a ~15-year high for recorded gusts, around 8mph below the record.

Stuxnet - and I haven't come across it directly - appears to hit Siemens Industrial Software (thanks, wikiP). Ardrossan, where this turbine is, consists of Vestas turbines - they may have some Siemens generators, but they will* not* be running any Siemens software.

Obviously one would hack whatever controller was in play, I just used Stuxnet as a known example.

What follows is speculation from somebody without your detailed knowledge of the subject, so feel free to mark my work:
The fact that it caught fire, rather than breaking up as other turbines have in high winds, suggests some possible failure modes and over-speed wouldn't be the first one on my list - you surely couldn't generate enough heat in even a failed shaft bearing. I would expect there to be very robust last-resort over-current protection, so an electrical fault, while possible, ought to be an unlikely candidate for fire starting. That's why I settled on the failure to coordinate blade feathering and turbine braking. It would be feasible to mechanically interlock these systems to prevent braking of a turbine set to a power-producing pitch, but I'm not sure that one would do so as a failure in the linkage could hold off the brake even when feathered, which seems like a higher risk than the one you're trying to mitigate. If the feather/brake interlock is implemented in software, that provides a potential attack vector/catastrophic bug.

Of course, the other possibility is a particularly fatty and therefore combustible goose was perched on the generator housing and got struck by lightning. Stranger things have happened in power generation/distribution accidents.

Of course, the other possibility is a particularly fatty and therefore combustible goose was perched on the generator housing and got struck by lightning. Stranger things have happened in power generation/distribution accidents.

I'm pretty sure that hippy wasn't in Scotland at the time of the incident, though.

!

Where was that taken?

That is one mahoosive gator. #wearenottopofthefoodchain

Brutus, an 18-foot-long saltwater crocodile, rises like a nightmare of mythic proportions from Australia's Adelaide River, lunging for the buffalo meat offered up by cruise operators. (Note that Brutus is missing a front leg -- lost years ago, according to local legend, in a shark attack.)

http://www.life.com/gallery/67641/2011-pictures-of-the-year#index/93

Would like to see what took his leg off

Would like to see what took his leg off

Wowzer

BBC News - Storm caused wind turbine fire - YouTube

I'm pretty sure that hippy wasn't in Scotland at the time of the incident, though.

Ha! repped.

Normal for Scotland, and therefore normal for a wind farm sited on top of a hill in Scotland
Unless you have more data than I already alluded to, the recent storm hit a ~15-year high for recorded gusts, around 8mph below the record.
Obviously one would hack whatever controller was in play, I just used Stuxnet as a known example.
What follows is speculation from somebody without your detailed knowledge of the subject, so feel free to mark my work:
The fact that it caught fire, rather than breaking up as other turbines have in high winds, suggests some possible failure modes and over-speed wouldn't be the first one on my list - you surely couldn't generate enough heat in even a failed shaft bearing. I would expect there to be very robust last-resort over-current protection, so an electrical fault, while possible, ought to be an unlikely candidate for fire starting. That's why I settled on the failure to coordinate blade feathering and turbine braking. It would be feasible to mechanically interlock these systems to prevent braking of a turbine set to a power-producing pitch, but I'm not sure that one would do so as a failure in the linkage could hold off the brake even when feathered, which seems like a higher risk than the one you're trying to mitigate. If the feather/brake interlock is implemented in software, that provides a potential attack vector/catastrophic bug.

Wind farms are designed for a 25 year life (often only installed for 20 years due to planning restrictions though). They are designed to withstand the 100yr max 3s gust, modelled for the particular hill they are on, which they may have just been hit by. The recent weather is still not considered 'normal' though. (I do have recorded wind data, having modelled around 150 wind farm sites in Scotland). Note - this is only one turbine, there are many many more that safely weathered the storm.

There are mechanical fail-safes all over the place in wind turbines (including the pitch and brake). They should over-ride any software control, a bug wouldn't be able to be introduced to the turbine control system other than extremely maliciously, a bit pointless as you'd need to have an extreme weather event at the same time to do any damage.

Your thoughts above are along the right lines - the fire (which was short lived) was the fibreglass nacelle housing, I would expect it was an electrical fault, or lightning. An electrical fault would be triggered by a failure of the pitch and/or brake causing the generator to overload if stil grid connected (which would need a failure on the breakers as well).

^she knows what she is on about :)

Would like to see what took his leg off

ftfy

http://www.dayswithmyfather.com/