Heat Exchangers/Bursting Discs/Burst Sensor

[samuray]

Hi, everyone.
Based on your knowledge and experience I would like to hear your advice/comments on the following:
Bursting discs are to be used as the pressure relief device on a Proccess plant shell and tube heat exchangers. These bursting discs are equipped with burst sensors (wire) which trigger an alarm on the central control room once the disc bursts. The problem is that this type of sensing device have been a source of innumerous false alarms therefore, the question is:
What would be the consequences/implications if this sensors were removed living the bursting disc without a burst sensor which would imply that we would be relying on the operator expertise to detect a burst event?
thanks

 

[TD2K]

The consequences depend on your system, it's impossible to answer you with no details. Where does the fluid go? What is the fluid? Does Operations have another way to determine a disc has failed? How long will that take? Is that delay acceptable? And so on.

Once the burst disc fails, flow will continue until the system is emptied or the system is isolated.

 

[samuray]

TD2K,

The fluid goes to the flare, and we have water(low pressure)flowing in the shell and gas(high pressure)in the tube, and yes operators can(depending on their expertise)determine that a burst has occured by looking at the flare changes and other plant changes, my question is if based on your experience do you think that it would be safe to do so(remove the sensors).

 

[TD2K]

I assume that the piping to the flare and flare system was designed to handle the initial flow of gas and liquid resulting from a rupture disc failure. I'll point out that may not be true but since it's going to the flare now, I'll make that assumption.

If you remove the sensor, the first thing that comes to mind is how long it will take the Operator to realize a disc has failed? How much liquid can flow to your flare during this period? What is the impact of this longer 'relief' to the flare on your process? Can the flare handle this amount of liquid (note, the flare and its knockout drum may have been sized assuming the operator would shut off the flow of fluid within a certain amount of time given the notification from the sensor). Are there any other potential flare load issues that could result from this relief continuing for a longer period of time?

I would suggest some sort of Hazard analysis be done on the sensor removal to look at the various possible risks (this would be mandatory in the US for a PSM covered process), what safeguards you have in the system and then decide whether removing the sensor is acceptable. There's no way anyone here can give you a 'yes/no' answer.

Finally, have you determined why the sensors are giving you false alarms and can that be corrected?

 

[samuray]

TD2K,

thank you, very usefull and constructive response, i'll go through all of that (i have gone through some already).

take care