Consideration of Human Error in Risk Assessment 10

[Wolfram]

German authorities require for the approval of risk assessments about offshore windparks the consideration of human error. The authorities require the consideration of "normal" operator error (e.g. ship officer fails to read the radar information correctly) as well as negligence or gross negligence (operator sleeps during watch or is drunken). Who can inform me how this topic is handled in risk assessments for chemical or petrochemical plants.

 

[Wolfram]

owg

if you reed the first publications about human error reliability or human error analysis (e.g. NUREG handbook about Human Error Analysis or the publications from Hunns and Daniels) you may find the indication that all human error rates were collected and analysed by the participation of hundreds of test subjects and thousands of operator actions. In no publication you will find a remark that the test subjects were drunk or high. Instead you will find a lot of remarks about the composition of test groups and test subjects (e.g. age, sex, education and physical condition). Again nothing was mentioned about drunken test subjects.

I think you make a big mistake in your risk analysis if you include acts of negligence and gross negligence to the given or published human error rates.

 

[JOM]

Wolfram.

I think we've now got a better appreciation of your situation. It's not so much a question of protecting the turbines, but protecting shipping and the environment.

Is there some contradiction in the argument against the windfarm because of the possibility of oil pollution? The windfarm produces clean, green power? There should be some positive arguments in favour of the farm.

Recently I heard of a scheme from a DVD manufacturer or perhaps a video store chain. They propose producing DVDs that self destruct when the hirer starts using the item.

This means that after a number of days the DVD becomes unplayable. So the hirer throws it away. At present they have to return their rented DVDs to the store.

This scheme would generate a new waste stream of plastic discs and containers. The company argues that because customers do not have to return the DVDs to the store, there is a reduction in vehicle journeys. Reduced fuel consuption. Reduced greenhouse gases. so overall, their would be a nett benefit to the environment.

Clever argument.

Will the windfarm generate power that would otherwise be produced by oil fired stations using imported oil? Would that result in lower oil imports and fewer tanker movements?

That would mean reduced numbers of tanker incidents and oil spillage. So the windfarm might generate a nett gain for the environment.



Cheers,
John.

 

[HamishMcTavish]

Hi Wolfram, found the enclosed which might give you either contacts or references of use.

http://www.cs.auc.dk/~pba/Preprints/HumError

could be of use as a method to compile empirical data for probability of accidents caused by drugs/alcohol.

The contents of the thesis shown at

http://www.geocities.com/tsgreg2002/abstract.htm

lists chapter 4 as looking at the role of alcohol and drugs in shipping accidents. Couldn't see anything more than the contents but you could try mailing the author.

Regards, HM.

No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[HamishMcTavish]

Wolfram,

JOM said "Is it fair for the authority to ask you to assess the occurrence of human error in an industry you are not part of - shipping?

With the very large volume of ship movements in your area, surely someone has figures for ships going where they are not supposed to - reefs, beaches, sand banks, oil rigs. Does the cause really matter?"

Perhaps looking for the contribution from negligence is a red herring since 'other' human error will play a part too. You could take a look at historical data / accident reports and say x% of the accidents were judged to have a root cause of negligence (never that simple) and then put another or-gate into your fault tree. The problem is whether you could review enough data to build up a picture worth using.

I think that to substantiate that you could contact the operators of the types of ships capable of causing catasrophic or major environmental incidents and ask them how they encourage people NOT to drink etc, what kind of shift system they operate and what the manning levels are for ships of those types. That way you can get a feel for the safety culture of the shipping.

I think that the requirements to ASSESS the risk would have to be done but by the OPERATOR of the ships. Either yourselves or the German government need to request/contract them to do this for you. Surely there must be a requirement on the operator to have produced risk assessments for their work?

I read the paper on the link you provided and was wondering if there were any better criteria to use for harm to the environment than small, serious, big and severe. Has anyone used numerical criteria to assist you in targeting the types of operators to ask the above question? If it were categorised in say number of barrels of oil released or number of birds killed then you could start to think about whether that size of ship COULD be damaged by the impact.

Then you paint a picture of defence in depth: The wind farm is as far away from the shipping lane as we can make it, it will be marked on charts, there are engineered collission avoidance systems being developed, there are two operators on duty to support each other, the safety culture of the shipping operators typically supports the skippers to be alert etc and even if there is a collision the type of ships capable of causing environmental damage is not harmed by the (glancing?) impact with one of our wind turbines.

I think after that you have to start making ALARP arguements since the public at large will receive a benefit from the power the windfarm will provide. Surely that is in the national interest.

Regards, HM.



No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[Wolfram]

Hamish

all the time fascinating how you get the references.
The publication from Mr. Koester (Danish Maritime Institute) is an excellent publication about categorization of human error in marine industry. Furthermore it describes quite plain the problems of understanding and analysing human errors. In fact it is one of the best publications about this topic within the last five years.

The second reference is nothing news. More or less it is the same as already published by Prof. Soares (University of Lisboa) about human errors in marine industry.


JOM

your statements about net gain for the environment are correct but it does not solve the problem.


The problem :
1.How is human error considered in qualitative and quantitative risk assessments?
2. How are acts of negligence or gross negligence (e.g. the drunken skipper or sleeping operator) considered in qualitative and quantitative risk assessment?

 

[Wolfram]

Hamish

if you are dealing with risk assessments for offshore installations it is worldwide common practice to consider the shipping traffic around the offshore installation (please look at

http://www.standard.no/standard/NORSOK_standards/2169/Z-013.pdf

which is one of the best standards for marine risk assessments).
You cannot shift the problem to shipping companies or national authorities. And if you read the excellent article of Mr. Koester you may see which difficulties occur if you try to analyse accidents reports about marine accidents. It is frustrating and currently not the method to derive human error data which gives you a flavour about negligence and gross negligence.

So who can show me an official statement how acts of negligence or gross negligence are handled in qualitative and quantitative risk risk assessment.

 

[HamishMcTavish]

Hi Wolfram,

I found the following link on the UK Maritime and Coastguard Agency site

http://www.mcga.gov.uk/c4mca/mcga-regs/windfarm

which is the steps taken to address navigational safety in teh consent regime for establishment of wind farms off the UK coast.

I am still reviewing the NORSOK standard.

I agree with you that you have to take account of the impact that YOUR structure will have on surrounding area and traffic, however you seem to be implying that it will be your responsibility to carry out human factors assessments of how likely it is that drunken skippers will crash into your structure i.e. activities that you have no control over. You might as well include the risks that the drunken skipper will crash anyway, but I do not see the point.

The assessment of tasks that you cannot control does not make sense; the whole point of risk assessment is that if you evaluate the risk and it is too high when compared against the relevant criteria then you take steps to reduce the risk. If you have no control over the risk then this last point is not possible.

You can assess the actions of the control room operators of your windfarm failing to interact properly with the surrounding shipping (and maybe even how likely it is that they will be drunk) but this would fall under the advice given previously by myself and others.

Regards, HM.



No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[Wolfram]

Hamish

thank you for the given reference of UK MCGA but the requirements/recommendations of this authority are very similar to the one already published by the Dutch and German authorities.

I have never said that it is my responsibility or task to perform a HF assessment for all ships passing the offshore windfarm. But I have been asked by national authorities and environmental presuure groups to consider acts of negligence and gross negligence for the QRA.

If you are not able to perform an individual or ship specific HF assessment to solve this problem you are allowed to use generic human error rates (published by NUREG, Hunns and Daniels, Swain and Gutman and many others)like the ones listed by MarkraM (see above).

In the frault tree model and Monte-Carlo-Simulation we included several times the basic event human error (e.g. misinterpretation of radar signals, wrong course of the ship, switching wrong UHF channel etc.). For all the basic events we allocated generic human error rates and together with the technical basic events and failure rates we got a nice result about collision frequency and collision risk.

Now authorities and environmental pressure groups argue the risk assessment is wrong because acts of negligence or gross negligence are neither qualitatively nor quantitatively considerd.
Unfortunately the published generic human error rates are not applicable for acts of negligence and gross negligence. Furthermore the qualitative part of the risk assessment the so called HAZAN was done without consideration of negligent & gross negligent acts like switching off the dead man alarm, switching off the collision avoidance alarm, looking TV during watch etc.. All these acts are reallity in marine industry and it is difficult to defend a QRA without it.
So the solution could be to include qualitatively all these acts of negligence and gross negligence into the risk assessment (by the way who has experience in doing this and is it reasonable to do it?)and to define technical and organisational measures of protection.
But how to do it quantitatively? I don't know. Now we put in the fault tree model the basic event "acts of negligence & gross negligence" directly below the top event "Collision" but without allocation of probabilities (only as qualitative event).

How do you handle this problem (negligence & gross negligence) in nuclear industry?

Regards
Wolfram

 

[JOM]

Hello Wolfram.

I believe the 2005 World Wind Energy Conference is to be in Melbourne, Australia - my home town.

Your problem sounds like a good paper to present to the next conference.



Cheers,
John.

 

[owg]

Wolfram
The following reference from NUREG discusses the role of drugs (legal and illegal) on the performance of nuclear station operators.

http://www.nrc.gov/reading-rm/doc-collections/nuregs/contract/cr6751/cr6751.pdf

I did not see any statistics there but the long list of references may lead you to some useful numbers.

HAZOP at

http://www.curryhydrocarbons.ca

 

[Wolfram]

curryhydrocarbons

I am not sure whether your tip with a more or less vague statement about the list of reference is a help or a burden.

 

[HamishMcTavish]

Wolfram,

I think from the previous posts that you are unlikely to get a definitive answer to your question in one place, more than likely you will have to piece it together from various references. To rebuff someone who is offering a tip or pointer to where you may find the strategy seems a little rude.

I think that if you consider a deterministic arguement based on the numbers of actual accidents in the area that you plan to site the windfarm then that will give you the bounding case. If I were you I would give a ROBUST reply to your regulators / crtitics telling them that they are asking for too much detail from someone who is not the duty holder (i.e capable of reducing the risk) and ask them what criteria they would like you to use. In the nuclear industry the usual way of treating external events is to generate statistics from historical records.

From my previous posts it should be clear that some negligence could be from 'normal' human error and some from violations BUT that the charge of negligence would be made by the courts, after the considering all the evidence to establish whether the operator or the organisation (or both) were negligent.

Regards, HM.

No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[owg]

HamisMcTavish - Thanks, I needed that.

HAZOP at

http://www.curryhydrocarbons.ca

 

[Wolfram]

Hamish

Hamish you are right my remark about owg's pointer was uncalled and I am sorry about that.

I agree with you that it is time now to come to an end.
Looking back to all the tips I have received during the last days I say thank you very much to all the one's who tried to help me. Nearly all tips were helpful and it is my task for the next days to put the bits and peaces together.

I will try the ROBUST reply and also use your arguments. We will see how this matter will end. Everybody who is interested about the progress of this issue may contact me under wbr@gl-group.com at January next year.

Cheerio
Wolfram

 

[azertyuiop]

wolfram ,
I started to read the long thread just out of curiosity , but spent a lot of time to read it all , because your problem isn't easily solved.

I can't give you any concrete help on this matter , not being into this specific sector , nor doing Hazop's.

However , I think you also could look into another direction to collect further intelligence : my country (belgium) decided some time ago to implant 65 x 2.5MW windturbines in an offshore park just on the outlet of the English Channel , turbines supplied by a renowned danish windpower manufacturer , that also posted intel on this on his website (a competitor of your company ?). Given the location of this 65 turbines implantation amid one of the biggest marine traffic locations in the world ( ships going to Antwerp , Rotterdam , Hamburg and Bremen ) , they surely would have to also complete some analysis similar to your problem. The turbine park will be implanted om an offshore sand bank , that surely already is put on marine maps , but I suppose this information will help you find another channel of information to try to solve your problem.

Interested to see how this will pan out . . .

 

[Wolfram]

azertyuiop

we already delivered a QRA for a Belgium offshore windpark project. The Belgium authorities are very pragmatic and professional. There is no problem with the consideration of human errors in risk assessment.

Regards
Wolfram