Required and rated flows for back pressure / header determination 5

[Justice100]

Hi all,

I am trying to understand when to use required and rated relief flows for header sizing for conventional pressure relief valves.

API 521 states in table 8 that for a spring loaded PRV then design for rated capacity in the tail pipe and required in the main header. My understanding is that this is based on coincident relief flows where all the PRVs will not be relieving their rated flow simultaneously.

In the section for Single device disposal systems it states "Pop-action PRVs normally have the backpressure calculation based on the rated capacity of the valve. The design of the disposal system should be checked for adequacy under such conditions."

If I have a disposal system with many relief valves then for a scenario where a single relief valve lifts should the entire system be designed for the rated flow? For a plant where the flare size is determined by one very large relief valve this would seem important. Similarly, if you have a coincident case where two relief valves lift, one a large PRV and the other small then would it make sense to calculate the back pressure based on the rated flow of the larger valve?

A final API 521 extract
"Typically, the required relieving rate is used for the flare header, flare tip, and knockout drum design with spring-loaded PRVs. However, there can be instances where a higher flow rate than required can be encountered that affects operation of the downstream equipment. For example, most spring-loaded PRVs discharge 50 % or more of their rated capacity at set pressure. Consequently, the initial flow rate can be greater than the required relieving rate. In this case, the rated capacity can be used as an upper-limit flow rate when designing downstream components such as scrubbers, thermal oxidizers, and liquid seal drums."

 

[The Obturator]

I can't comment fully to your dilemma, but I picked up in your first line that you have conventional design PRV's. Conventional spring PRV's alone will ne limited to typically 10 % built op back pressure maximum (or more in lime with their overpressure eg., 21% for fire case Ref: API-520). With everything else you describe, you might have an issue with other back pressures generated from other PRV's relieving into the same header. I would not know how your system PRV's were selected if the contractor/specifier/engineer/vendor was not aware of any other back pressures. Like the sizing of PRV's, worst relieving case is taken and the maximum back pressure should have been considered, meaning probable selection of balanced PRV's. I'm sure others will fill in on your posted issue.

Per ISO-4126, only the term Safety Valve is used regardless of application or design.

 

[georgeverghese]

Q1. API 521 states in table 8 that for a spring loaded PRV then design for rated capacity in the tail pipe and required in the main header. My understanding is that this is based on coincident relief flows where all the PRVs will not be relieving their rated flow simultaneously.
A1. Yes, only for coincident relief loads

Q2. In the section for Single device disposal systems it states "Pop-action PRVs normally have the backpressure calculation based on the rated capacity of the valve. The design of the disposal system should be checked for adequacy under such conditions."
A2. Yes, but I would interpret this to be the built up backpressure in the tail pipe only due to the rated flow. The main header need only be checked for required flow. The imposed backpressure at the end of the tailpipe would be the highest backpressure coincident with the relief event at this PSV.

Q3. If I have a disposal system with many relief valves then for a scenario where a single relief valve lifts should the entire system be designed for the rated flow?
A3. No, see A2

Q4. For a plant where the flare size is determined by one very large relief valve this would seem important. Similarly, if you have a coincident case where two relief valves lift, one a large PRV and the other small then would it make sense to calculate the back pressure based on the rated flow of the larger valve?
A4. Only if the two PSVs' have staggered setpoints, with the larger one at the lower setpoint. And provided that the rated flow from this larger PSV exceeds the required flow. If the rated flow from this large PSV is still less than required, then use the rated flow from both large and small PSV.

Q5."Typically, the required relieving rate is used for the flare header, flare tip, and knockout drum design with spring-loaded PRVs. However, there can be instances where a higher flow rate than required can be encountered that affects operation of the downstream equipment. For example, most spring-loaded PRVs discharge 50 % or more of their rated capacity at set pressure. Consequently, the initial flow rate can be greater than the required relieving rate. In this case, the rated capacity can be used as an upper-limit flow rate when designing downstream components such as scrubbers, thermal oxidizers, and liquid seal drums."
A5.Would follow this guidance for low and low low pressure vent and flare headers only where the design pressure of connected equipment is very low; say those with MAWP < 30psig (ie. where built up backpressures of 3psig or more is not permissible) . In these headers, transient peak pressures of a few psig may develop due to short bursts from these PSVs'. Thermal oxidisers are low design pressure, and liquid seals in vent seal drums may be of the order of no more than 2m water column or so. Use the recommended 50% of rated capacity for conventional PSVs' if this exceeds required relief rate. Not valid for vents and flares collecting from vessels with design pressure >30psig I think. If the controlling relief load in this LP or LLP header is addressed by a pop acting conventional PSV, then use 100% rated capacity.

If your plant has a ESD blowdown system, then the max imposed backpressure on the end of the tailpipe for the PSV in question (see A2)would include the peak backpressure developed at this tailpipe location due to facility cold blowdown (not firecase blowdown) together with any coincident contribution from continuous vent or flare loads.

 

[don1980]

API 521 is providing general guidance which is applicable for most (i.e. common) cases. One should recognize that there will always be exceptional cases where deviation from general guidance is justified. That’s why human critical thinking hasn’t (so far) been replaced by AI software. To recognize when an exception should be applied, understand the reasoning behind the general guidance. You can usually figure this out by sitting back and thinking it through for a minute.

For example, consider a particular fire zone in a plant. Let’s say there are 100 vessels of various sizes and with different wetted areas (each vessel with its own PSV). We know that each vessel isn’t going to heat up and relieve at the same exact time, and we also know that each PSV will cycle open-closed based on its over-capacity and the heat intensity experienced by the protected vessel. Since we have 100 PSVs (100 unique situations) we can confidently say that the flow rate in the downstream collection header will be much closer to the sum of the required flow rates rather than the sum of the rated flow rates.

Now consider a different case in which there are only 10 vessels instead of 100. Obviously, our probability assessment is now different. With only 10 PSVs in the system, there’s a greater risk that the collective flow rate (header flow rate) may sometimes be higher than the sum of the required flow rate of each PSV. So, if this is a design in which there’s little tolerance for being undersized, then a deviation from the API 521 guidance is justified. Be more conservative and use a higher flow rate.

If, for example, we have a case in which there are only 2 vessels (2 PSVs) in the system, then it’s even more obvious that the API 521 general guidance (use sum of required flow rates) isn’t sufficiently conservative. When applying general guidance, always think about your specific application and assess whether or not that guidance is suitable for the application. If it’s not, then make the appropriate adjustments for your specific design.

 

[Justice100]

Thanks for your responses Don and George. It would seem some differing intrepretations from very knowledgable people.

My interpretation of API 521 is similar to Don's. However, I think perhaps the required relief rate might be applicable to the header on the basis of it being a dynamic system i.e. that a conventional relief valve will initially pop open at set pressure but will not relieve the rated flow for a sustained period and will relieve somewhere between the required and rated before either reclosing and cycling or staying open at a lower than relief valve design accumulation pressure depending on how much the required relief flow is in comparsion to the rated flow and if the conventional valve has any (minor) modulating characteristics. Since the header will be large compared to the tail pipe by the time the flare system has been pressurised then the required flow might be more appropriate. In a similar vein for a system designed for a single relief valve designing the flare for the required flow would be ok on the basis that the rated flow would be very short lived. I feel API 521 could put some more guidance/explanation on this.

 

[don1980]

Justice100, I think your understanding of this issue is pretty good. The key to assessing special cases (i.e. cases in which the header is fed by just one PSV or just a few PSVs) is to understand the fundamental mechanics of how pop-valves work - how they act. When a pop-valve pops open, it only opens to approximately 50% lift. What happen next will depend on the ratio (required flow)/(rated flow). If that ratio (for this given scenario) is greater than 50%, then the system pressure will increase, forcing the lift from 50% to a higher value. Full-lift isn't achieved until/unless the system pressure rises by 10% above set pressure. But, of course, the system pressure won't rise if that PSV (at 50% lift) is already supplying the necessary flow (the required flow). If the flow through that PSV (at 50% lift) is approximately equal to the required flow, then the PSV will tend to remain open, without any significant changes to that lift position (50%). If the required flow is less than 50% of the rated capacity, then the PSV (at 50% lift) is flowing more than the system demands, so the system pressure starts to fall, and this in-turn will cause the PSV to re-close. Once closed, the system pressure rises again, pops the PSV open, and the cycle repeats.

The API 521 statement which you labelled "A final API 521 extract" is pointing out a special case in which the total header flow (from one or just a few PSVs) could be greater than the sum of the required flows. What makes this situation possible is the mechanical action of pop valves - the fact that they initially pop to approximately 50% open. To illustrate this let's say, for example, you have two PSVs feeding a header, and the required flow from each PSV is only 20% of their respective rated capacities (20% of each valve's rated flow). By applying the generalized API 521 guidance, you'd size the header for the sum of those two required flow values. But, based on our knowledge of how pop-valves work, we know that that sum is too low for this specific case. If those two valves open at approximately the same time, we know that they'll pop to ~50% lift, so the actual flow rate in the header will likely be much higher than the sum of the two required flow rates (each being 20% of the PSV's capacity). This is what that API extract is trying to explain. It's saying you may want to size that header by summing 50% of each PSV's rated capacity, rather than 20%.