Recently an accident occured in South Africa that at Duvha Power Station a Steam overspeed failure and causes catastrophic damages. I want to know the possible Root Causes. Please click the link and read the report with Pictures. I want to know your good views and root Cause analysis.

http://www.scribd.com/doc/61342541/Duvha-2011-Overspeed-Testing-Failure-V2

I shall be very thankful to you.

Regards,
Askani

 

I have also been searching for information on this and another catostraphic over speed event with no luck.
this article provided the most info
http://www.citizen.co.za/citizen/co...5&Operator-error-caused-Duvha-station-failure

It only states;
1) operator error
2) modification to controls

I have my opinions, but no facts.
I have queried other groups about if the performance of a true load rejection is an yearly requirement and the census has been no. So if this was supposed to have been a "normal" test of the OS and not a load rejection test, then the operator removed the generator from the grid while the unit was not at "no load". based open remarks that unit accelerated to 150% in 10 seconds, It must have been about 50% load when he opened the generator breaker

 

Below is the text that Eskom posted in their Nov 23rd financial report:

The investigation, by a team which included independent experts appointed by the insurers, pointed to the root cause of the incident as being a modification under taken by the power station in 2004 on the electro-hydraulic governor controller (known as the droop controller), while the direct cause of the incident is attributed to an operating error, in that the operator did not follow the set procedure while undertaking the test.

http://www.eskom.co.za/c/article/807/eskom-reports-a-strong-financial-performance/

 

The problem with root cause analysis is to know where to stop. From reading the reports, it seems that the immediate cause was either an incorrect procedure or failure to implement the procedure correctly.
We then need to ask:

- did the people carrying out the exercise know the reasons for the procedure, and the associated risks? If not, why not?

- Was there an element of complacency in that routine tests which never fail are often rushed through?

Both of these lines of questioning will lead back along a chain which will ultimately lead to management decisions - if this one hadn't gone bang, something else would have.

- why did the overspeed bolt fail to trip the machine at the correct speed? Did the bolt fail to operate, or did the bolt move as expected and fail to dump the oil?

Here you start to look at mechanical details, but again can go back a long way. Again management issues may crop up here as well. If you want to follow the contributions of management to major disasters, read some of Nancy Leveson's (sunnyday.mit.edu) papers. From all the blogs and other rantings that come up on Google (well worth a read for the sheer ignorance of engineering issues) there is also a political element that may have contributed.