Tuesday, 13 December 2011

The last word on safety

...well, for a little while - maybe!

The recent incineration of a turbine at Ardrossan has caused some odd pieces of information to be released - from a couple of sources. We'll start with total accidents over the past 5 years. According to the Daily Telegraph who quote Renewables UK, there have been 1,500 accidents or incidents over the last 3 years - that's approaching 1 a day. Now, a number of these will be related to construction, and quite possibly a large number - but since the 1,500 number isn't broken down to any useful degree, we don't known the distribution of operational and construction incidents. Curiously, the same article clearly identifies the reason why accidents involving wind turbines are not identifiable: its because a wind turbine is classed as a machine - not a structure and therefore there is no requirement to report mechanical failures. Renewables UK seem to have a handle on the problem (through its Lessons Learned Database), its just a pity they don't seem to want to place that information in the public domain. Instead, Renewables UK hide behind a probability that 'the risk to the public is 1 in 100,000,000'. Now, where did that number come from? What is the basis for it?

The really nasty bit about the Renewables UK quotations in the Daily Telegraph article I have referenced above concerns setback distances. Renewables UK state that planning and safety rules meant turbines were always at a certain minimum distance from roads and homes, reducing further the risk to the public. Sorry, did I miss something? What is this minimum distance? Is this distance legally enforceable, or just a recommendation or an industry rule of thumb? The only restrictions on placing wind turbines close to residential properties that I am aware of are based upon predicted noise output - which is where we end up with the ridiculous situation where these massive structures (sorry machines) end up being placed as close as 650 metres to isolated residential properties. And with quieter turbines being developed all the time, I imagine that minimum distance will reduce further. I'll return to risk assessments in a minute.

The last piece of odd information comes from Infinis, the owners of the Ardrossan facility. According to this article, staff at Ardrossan abandon the wind farm when wind speeds exceed 55mph. Ask yourself why? 55 mph isn't all that windy after all - what are they concerned about?

Now returning to risk assessments, and in an effort to try and find out whether the wind farm developers take the possibility of blade throw, ice throw and or catastrophic tower collapse seriously, I reviewed the Health and Safety section for a number of the Environmental Statements for wind farms that I have linked to. Only the proposed Hill of Ochiltree document makes significant note of such a possibility, and here is what they have to say on the matter (para 13.4.6, chapter 13 refers) :

A recent risk assessment1 for a wind turbine application in West Northampton considered the risks to public health of an operational turbine. Structural failure and ice throw were considered in detail in the report. Following clarification from the relevant planning department it was confirmed that the 1 in 10,000 probability of component failure equated to a 1 in 1,000,000 probability of turbine collapse. An annual limit of risk to public health of 1 in 100,000 is proposed based on recommendations in Health and Safety Executive published guidance2 and British Standard3. The turbine applied for in West Northampton is similar to those proposed at hill of Ochiltree i.e. modern 2+MW 3 bladed turbines. Therefore the conclusions in the risk assessment can reasonably be applied at Hill of Ochiltree. While turbine failure is a potential risk factor, the actual public health risk is extremely low.

Perhaps we have found the source of that risk statistic quoted by Renewables UK - which appears to have been rather mis-quoted. Now, I'll leave those risk numbers alone - there is something far more interesting if we dig a little deeper. I tried to find the risk assessment carried out that this paragraph refers to. I couldn't find it, since the link provided as a reference is for an old planning application. However, I believe it refers to a planning application for a 125 metre turbine at an Asda store submitted to West Northamptonshire District Council in 2007/8 (planning reference 07/0356/FULWNN). I would really like to see that report - so if you have a copy, please get in touch. Whilst the report itself could not be found, I did come across the report to the Northampton Area Planning Committee that made reference to a risk assessment carried out on a 125 metre wind turbine planning application. The really relevant part of that planning report is quoted below:

The Risk Assessment carried out for the proposed 125m turbine at the ASDA Brackmills Depot (WNDC Planning Application: 07/0356/FULWNN) showed that ice could be thrown up to 680m from the turbine – substantially greater than the separation between the turbines and the A14 and also further than the nearest residential dwellings. We are not aware of any proposal to address ice-throw as suggested by the Highways Agency. We also bring to your attention that vibration monitoring technology does not detect even ice build up and so can not be relied upon to prevent ice throw. We strongly recommend that the turbines should not be allowed to operate during any period when icing could occur and that they should remain disabled until such a time as any ice on the blades has thawed.

Hmm. So, somebody has sat down and calculated that ice could be thrown 680 metres from a 125 metre turbine. Presumably then, shattered blade parts could also be thrown this sort of distance. And how close are similar (or larger) sized turbines being placed to residential properties here in Scotland? Closer than 680 metres. If you are interested in the 'mechanics' behind ice throw, take a trip to this page.

Next, I went on the hunt for information regarding blade throw. The only place I could find significant reference to this was in an environmental statement for an American wind farm called Desert Claim, in Washington State USA. Page 3-162 (actual document page number 6) makes a weak attempt at tabulating maximum blade throw distances. Now, it has been quite a while since I studied graduate level physics, but the approach taken by these developers is a gross over-simplification; no account is taken of a rotor over-speed condition (the calculations are based upon the maximum normal operating rotational velocity of the rotor) and no account has apparently been taken of the wind and how it could effect the trajectory of a detached blade (or part of). How would an 80 mile an hour wind effect these estimates? This in actual fact is not a complex situation to model. It involves lots of variation in key variables, but it would be relatively straight forward to build a software tool that modelled the relationship between wind direction, blade mass and rotational velocity and provided detailed information on the sort of distance a detached blade could travel in extreme weather. This to my mind should be provided by the turbine manufacturers.

The point of all this? Well, clearly some work has been done on the risk of ice throw - but it is not widely quoted or referenced. I could find no reference to any serious attempts to model the risk zone of blade throw. Both of these potential hazards are brushed off as insignificant by many developers. And catastrophic failures? Developers and industry bodies seem quite happy to hide behind (mis-quoted) probabilities on the basis that apparently demonstrates how safe their equipment is. The reality is, the probability becomes more significant as the number of installed turbines increases. I for one will not voluntarily go within a country mile of one of these things - and this is why:





What I find odd is that our society is becoming more and more risk averse - when it is convenient. The video above shows what can happen - after all it did happen, and in all probability, it will happen again - only next time, perhaps somebody will get hurt - or worse. What happens then I wonder? Wouldn't it be better to make EVERY REASONABLE effort to mitigate the potential impacts of such occurrences through the use of sensible setback distances rather than dubious probabilities?