Emergency Handling of DCS Control System Failures (III)
Thread Content
What will happen if a single CPU in the control station fails, and how should it be handled? We urge all marine enthusiasts to actively participate by leveraging their own DCS systems!
The system automatically switches to the Standby controller; the fault should be resolved, or else a new CPU needs to be replaced
Different systems, different processing conditions. For redundant systems. . Of course, for a non-redundant system, there’s only one option: shut it down. Hehe, as for handling the situation, of course it’s necessary to cut power immediately and replace the card (these cards probably don’t support hot swapping, to prevent accidents; thus, the card needs to be replaced while the system is powered off). What’s similar in this case is the alarm – it’s a high-priority system alarm, and the problem won’t be resolved until it’s dealt with. Then, it’s necessary to switch to a redundant CPU, either simultaneously or as the first step to take. The system is good; there’s no noticeable difference in the manufacturing process. Replace it with a new CPU, reinstall the data, and then perform the switch (between primary and secondary modes) to see if it works well; if not, replace it again – there’s money to spare. If everything is normal, then O~. The difference is that some systems have problems or hidden flaws, and these become apparent at that time. If another CPU cannot take over, then the system will fail to function. It could also be a matter of the manufacturing process – many signals go offline for several seconds, and as a result, many controls at this station are switched to manual mode. This is an issue related to redundant switching that needs to be addressed. Experiments need to be conducted during shutdowns, and things need to be handled gradually until the system can respond quickly
Phenomenon: If a single CPU fails, the real-time alarm window on the operation station will display the corresponding alarm message and emit an alarm sound. Action: Report the situation to the control room; the instruments should be dealt with immediately. This fault does not affect normal operation, but it is necessary to increase inspection frequency.
Another point is that whenever changes are made to the control scheme, it’s important to create comprehensive backups, and to copy those backup files to another workstation as well; this way, in the event that the hard drive on the engineer’s workstation fails, there will be no need to panic.