Burst disc filling flare KO Drum

[Justice100]

Hello,

I am looking at a burst disc on a S&T heater using water as heating medium located about 200m away from a nitrogen blanketed expansion vessel via 10" header. The expansion vessel is located above the header. If the burst disc were to spuriously burst then we will direct heating medium to the flare KO drum. The heating medium pumps will be tripped on burst detection (with redundancy) and we have a shutdown valve on the inlet to the heater and a check valve on the downstream side to prevent this. I am just wondering what would happen if the valves fail but the pump trips. The expansion vessel is at 10 barg so will push out a lot of water... or will it just bubble through the water? I suppose it will be a bit of both but I thought it would mainly push it out given the routing.

 

[georgeverghese]

Since you have a check valve on the downstream side S/T exchanger - heating medium side, reverse flow from the expansion drum cannot occur. It will be forward flow through the pump and the SDV (which has failed to close?).

Strange you have a unreliable BD - it should be replaced with a spring operated PSV. Or is there some reason for the BD selection ?

 

[Justice100]

Hi George,

The pumps will trip and an SDV will close to prevent flow in the forward direction and the check valve in the reverse direction... it is just incase these protection measures fail then if the expansion vessel empties the heating medium system into the flare KO Drum then we could overfill it. Check valves fail but I am not sure it is realistic for the gas in the expansion vessel will blow all the liquid between the vessel and the disc out like blowing through a straw.

Burst disc is required for tube rupture scenario where PSV was not considered to react quickly enough.

 

[georgeverghese]

This concern you have seems to be valid, but the root concern is the BD reliability. You need at least 20% margin on pressure between max operating pressure and BD setpoint.
The API suggestion for BD is only for the case when overpressure source pressure >1000psig or so. Even then I would much prefer a spring loaded PSV.

By the way, why do you have this tube rupture BD on this HX - can you not cater for this tube rupture with a larger PSV on the expansion drum ? Check valve fail close in this case is not valid. Can you lock open any manual valves on heating medium exit from this HX ? Where is the heating medium control control valve - better to have this on HX inlet.