Flow through a failed control valve at RELIEVING CONDITIONS

[jcaiken]

I've just been looking at API 521 (2007) and discovered that Sections 5.10.3 and 5.10.4 (on page 28) specifically state that for estimating the relieving flow through a failed inlet and outlet control valves "All flows should be calculated at relieving conditions", i.e. the set-pressure plus over-pressure margin (e.g. 10%).

I'm used to taking the PAHH trip condition as the basis for upstream conditions (and based on other threads in this forum, so are others) and I'm concerned that I may have missed this significant 'change' to the standard.

Does anyone know of an acceptable mitigating basis to enable me to retain use of the trip pressure rather than relieving pressure?

 

[Longinthetooth]

You use the upstream vessel trip pressure for the "supply" and the downstream vessel at PSV set point +10% for the relief. If no trip on the upstream vessel, use the upstream RV set pressure for the supply. Every relief project I have worked on does this.

 

[don1980]

jcaiken - I think you're misinterpreting that sentence ("All flows should be calculated at relieving conditions"). It's referring to the relieving conditions of the protected vessel, and not the relieving conditions on the upstream side of the valve feeding that vessel.

For example, say you have a vessel with a MAWP of 150 psig, and you're evaluating the failure of a control valve feeding that vessel. When calculating the flow across that failed control valve, use the relieving pressure (165 psig) on the downstream side of the control valve. That's what this sentence is saying. Of course, that's to your favor because it reduces the flowrate across the failed valve.

The next question is, what value should be used for pressure on the upstream side of the failed control valve. That's an important decision, and the answer will vary from case to case. In most cases, it's appropriate to use the normal pressure - the upper end of the normal pressure range. Otherwise, you're calculation is based on multiple independent failures - double jeopardy.

 

[LittleInch]

I think Don1980 has a pretty fair analysis of the pressure requirement. I would personnal take your analysis of the PAHH trip pressure as being the most reasonable on to assume. Alarms on their own are nothing and without any executive action such as trip or shudwon or valve closure pre-determined and wired into the safety system can be ignored for this purpose. it may well be a bit conservative, but if things are going wrong, then pressures could be considerably above the normal or even high end of normal. You would need to look at each case individually to determine your confidence that whatever the "trip" function is would actually work and restrict the upstream pressure to your PAHH level.

The issue will be the independence of any failure mechanism and if there ia common mode of failure (e.g. loss of instrument air) that causes the upstream vessel to relieve AND the control valve to fail open, then the scenario could reasonably be required to be caterd for in the downstream vessel. No commmon mode failure, then double jeapordary as Don1980 says IMO.

In short, no easy answer.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way