That's right about BP, Ann. One of the gauges was working and indicating that pressure was building. Instead, BP personnel chose to trust a broken gauge that indicated everything was fine. At least, that's what the book on the accident claimed.
Rob, as many have pointed out here, correctly working gauges and meters are only as good as the people reading them--or not reading them. The BP disaster was due at least in part to faulty oversight, i.e., lack of/incorrect monitoring.
Bob from Maine, I suppose one approach to this is to replace the gauge during scheduled downtime. That would mean some gauges would be inoperative during the wait for downtime, but that may be the most efficiently way to deal with this problem.
Rob; One issue about gauges is they are usually direct plumbed into whatever they are monitoring. Thus a broken gauge replacement requires shutting down that entire line and exposing the contents to contamination from outside or vice versa. Putting gauges on shut-offs that permit removal without leaks and making all gauges moveable such that the 'normal' is always in the same orientation. For most processes, gauges are 'trendicators' more than an accurate readout. The percentage of defective gauges is not surprising but what may be surprising is the number of gauges that no longer serve any purpose.