This is a topic that is near and dear to me, as it is one I learned very early on in my career as a means to safely perform PRD work on an RV or RD that is connected to a header/manifold. Although this design MUST be adequately done AND must be managed very tightly, it can add a level of safety for workers doing work associated with the relief system or header/manifold.
Let’s first squash the rumor that having an intervening valve(s) before or after a pressure relief device (PRD) such as a relief valve (RV) or a rupture disc (RD) is just a rumor and has no basis in fact or engineering design! In fact, the two most prevalent pressure vessel codes, ASME and NB, both allow this design, albeit with some stringent requirements.
Special Note to NH3 Refrigeration readers: IIAR-2, 2014 also allows this in 15.4.1.
So why would we even consider this?
Imagine the scope of work when changing out a PRD when the process is still live at certain parts. As I wrote about in 2014 (Does changing a Relief Valve on a “Header/Manifold” require a line break permit at your facility?), this task can prove to be an extremely high risk, even at ground level. But a line-break permit may not be adequate in protecting the workers, so in this article, I want to explain the “safety need” for these intervening valves.
In the “old days” pressure vessels were equipped with a single PRD, and to do work on this single PRD or to change it out, the vessel had to be emptied, isolated, and verified gas/vapor free before the PRD could be removed. This achieved a high level of safety, but at a considerable cost, as the prep work and verification took the vessel out of service for an extended period. So the idea of putting an intervening block valve(s) between the vessel and its PRD was discussed, and now we see both ASME and NB allowing this design. Then came all the environmental regulations and processes were now needing to discharge their PRD’s to a destruction device and thus the need to have an intervening block valve(s) between the PRD and the destruction device became a necessary evil for working on the PRD and the destruction device.
This design allows merely for the isolation of the PRD from either the vessel or the header/manifold (when attached to a header/manifold). This intervening valve (see my 2016 article “Butterfly valves and relief valve line isolation”) allows for the PRD to be isolated from the “live process,” both a vessel in service and a header/manifold, to be worked on. And when the PRD discharges to a header/manifold, isolating it from the header/manifold becomes safety critical. If any other PRD were to lift/burst either prematurely or during a high-pressure event, the header/manifold would act as a conduit to convey the hazard directly to the worker(s) at the header/manifold opening. So having an isolation valve AFTER the PRD and/or BEFORE the header/manifold can play a critical safety role in working on the PRD.
However, and this is a HUGE “HOWEVER,” these intervening valves before or after a PRD can create a potentially catastrophic situation if they are not tightly managed/controlled. SAFTENG members MUST refer to my other postings regarding the CODE REQUIREMENTS for this type of design as EVERY ASPECT is safety critical. See the following:
National Board’s position on “intervening stop valves” in relief systems (2018)
Car Seal Program and Relief Valves (2011)
And now we should begin exploring the need for a “car-seal” program that would ensure compliance with BOTH NB and ASME code requirements. These two codes differ in their requirements for the “administrative program” needed for this design – PLEASE see my article earlier this year as the NB has set the bar higher with “constant attendance” when these valves have deviated to their CLOSED position.
The intervening valves MUST be able to be SECURED in the OPEN, as well as LOCKED in the CLOSED position for LOTO purposes. The SECURED in the OPEN is where our car-seal program comes into play and the LOCKED in the CLOSED position is where we have to comply with OSHA’s Control of Hazardous Energy (LOTO) requirements. We are permitted to “lock” OPEN the valves, as long as we do not use our LOTO locks for this purpose; as LOTO locks are ONLY to be used for LOTO during servicing and maintenance activities. A facility could use a special colored lock that is not part of their LOTO program for their car-seal program, but locks are not required. These valves MUST BE SECURED in their SAFE POSITION (i.e., OPEN in this scenario), this is where “car-seal” comes into play (see the 2011 article listed above).
When work on the PRD is necessary, and this scope of work exposes the worker to a potential release of the hazard, the PRD should be isolated using our LOTO program/process and these intervening valves. But there MUST BE controls in place that would prevent the CLOSING of these intervening valves, and this is done via our “car-seal” written program/procedures AND our LOTO process. Now comes the most CRITICAL part of this management system… RETURNING THE INTERVENING VALVES to their SAFE POSITION – BEFORE we return the vessel to service. This is all covered in my SAFTENG Car Seal Program (See Document Library), which I will be updating this week with the new NB “attendance requirements.”
But bottom line, having an intervening valve before or after a PRD is an acceptable design (which should be part of our documented relief design documentation) and can add a level of safety unable to be achieved by a line break permit alone. This design becomes even more critical when the PRD is located in hard-to-reach areas (e.g., working on top of a vessel or from a lift basket of some sort).