I did a quick review of ASME Section VIII, Div 1 page 93 and Appendix M. There is nothing in Code words that specifically prohibits this. However, the Code recognizes that good engineering practices must be followed by the owner of the vessel when installing pressure relief devices. I would recommend against any in-line filter that could affect fluid flow and pressure.

If you want to build a case with certainty, I would contact he local Jurisdiction. I would bet that they will NOT allow for any potential obstruction upstream of the safety relief valve.

you would have the include the possibility of filter failure in your design conditions and such a failure not affecting the relief capacity of the releving pressure of the installation.

Typically such reliefs are augmented by a back pressure control valve (manual or automatic) that relieves at a lower pressure than the safety device and can be maintained without compromising mechanical safety. It would not be unusual to have a motorized scraped filter ahead of that valve.

Stangman,

Maybe you didn't mean it the way I've interpreted but a filter upstream of a relief valve will not change the relief valve set point.  Certainly, I think all agree, that the filter could affect the relieving capacity of your relief system.

That said, you may want to review UG-135 Installation, I think there is language there that would support not installing a filter since it could obstruct the flow to the relief device.  Besides the possibility that the filter element could fail and obstruct the flow, after repeated times of flow through the filter the fouling noticed on the relief valve occurs on the filter.  How do you propose to service the filter?  You will have to install a block valve to isolate it which will require additional administrative controls to address the block valve in the relief path.

In situations where the process fluid is fouling, it is probably more typical to install a rupture disc upstream of the relief valve.  In some cases, a clean purge stream could be introduced at the relief valve inlet flange.  

However, these alone are only part of a solution.  I would say your real problem is the "occassional" relief "as expected" to which you refer.  Sounds like it happens often enough to want to do something about it.  I would also be working to resolve the reason "why" the relief is occurring.

I support the use of a rupture disc ahead of the relief valve.  As an upgrade, you might want to consider a rupture disc that is often used with viscous polymers, a self-sweeping design where normal flow sweeps the surface of the disc clean with no dead-ended flow paths.  This requires modifying process piping to do this.

The more you learn, the less you are certain of.

Does it matter that the valve in question is a thermal relief?

Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.

A rupture disc won't work since it would only be a one time use in a very inaccessible area.

I may be wrong, but a filter would require changing the relief valve setting since the upstream process piping would experience a pressure greater that the piping between the filter and the PRV i.e. a filter adds a pressure drop before the PRV.

Stangman,

A filter (not completely plugged) would create a pressure drop only after you reached the setpoint and there is flow through the relief system.  The filter would then be no different than any other component (elbow, valve, rupture disc, etc) in the inlet pipe to the PRV.  Too high a pressure drop in the inlet piping during flow and the PRV could experience undesirable closing and opening.  Prior to reaching the setpoint, there should be no flow and no pressure drop so there is no reason to compensate the set pressure.

You say installation of a rupture disc would be very inaccessible and maybe a rupture disc is not what you need, but it would seem installing a filter would also be very inaccessible.  Then it becomes more of an issue that you will not be able to easily maintain the filter.  Even without flow through the filter, just having the filter exposed to a dirty liquid will over time will cause fouling in the filter.  Just seems that there is no way to ensure the filter will not obstruct the flow.