Complete DCS failure - loss of logic configurations

[anoop]

Recently in our coal-thermal power plant, two 500MW units of DCS had a simultaneous complete loss of logics of all 214 processors (active and redundant included) leading to extremely dangerous plant conditions. The incident occurred when a faulty common domain ethernet switch was restarted. The incident repeated again after 3 days while reconnecting uplink communication of another common domain ethernet switch, leading to similar situation. With both units under safe shutdown, it was tested again and the complete failure occurred for a third time. Root cause analysis is still in progress by DCS supplier, no solution has been received yet. I request all to please share & discuss similar incidents. Edited January 4, 2012 by anoop

[panic mode]

no details or part numbers...? ok, looks like maxDNA is made by Bharat Heavy Electricals Limited from India (or BHEL, coincidence?)

[Ken Roach]

The same guy posted the same information to Control.com and a few other places. I think what he's really fishing for is somebody to say "Hey, my maxDNA system did exactly the same thing !". I'm thinking that's not going to happen.

[anoop]

@Ken Roach: Yes I did, but my intention is to get to the root of this incident through technical discussions. I don't see the lighter side of your comment, since the danger in such a situation cannot be imagined unless you know what can go wrong inside a power plant that goes totally out of control. Please try to comprehend worst-case scenarios for boiler, turbine and generator. Edited January 4, 2012 by anoop