Filter vessel - Thermal relief valve or normal relief valve 3

[EngCutie]

Hi, I have attached a drawing to help me explain.

Filter vessel - MAWP is 20Barg, I have a normal relief valve sized for when the pressure goes above the MAWP pressure, but my client wants a thermal relief valve. I’m a little confused, does that mean the pressure inside the vessel would never actually exceed the MAWP pressure without thermal effects? I read that thermal relief valves normally have smaller orifice size. Is it possible to have a relief valve that will relieve due to thermal effects as well as when the pressure exceeds the MAWP?

Upstream of the inlet is a pump, but it’s outside of our scope, so I’m not sure exactly what else is there.

Please help, thanks.


Cutie


Thank you, EngTips

 

[Chance17]

What is the filter dimensions. Diameter ?? Length ??

 

[pleckner]

"Thermal" relief valves are not thermal at all but pressure relief valves. The industry calls them "thermal" because of the fact that the scenario is a liquid filled vessel that is heated (thermal effect). When the liquid heats it expands and puts enormous pressure on the vessel. The relief valve relieves this pressure. Again, it is really a pressure relief valve.

Since the relieving rate for this scenario is typically very small, there are relief valves on the marker with very small orifices, smaller than the typical API D orifice, but they are still pressure relief valves. So no, they are not special and they are still code valves.

You can't have the client dictate to you, the process engineer, what the controlling relieving scenario is, this is your responsibility. You must check all credible relieving scenarios and come up with the one that produces the larges relief valve.

That pump is NOT outside of your scope if it is part of the system that is being protected by the relief valve. You must consider excessive head due to a possible blocked-in scenario. If the pump can't overpressure the filter, then that is not credible, but you still must check it. What about fire?

 

[EngCutie]

Because of the flow rate and everything, the normal relief valve is sized to have ‘G’ orifice, and a thermal relief valve as you mentioned ‘smaller than the typical API D orifice’. I’m not the process engineer, I’m not sure if there will be an overpressure (not due to thermal effects) scenario, well that’s one of my questions. The pump is not on the P&ID, according to one of our process engineers, our client should have relief valves at downstream of the pump, he thinks there is no way the vessel will experience overpressure if not due to thermal effects.

I have just read the filter vessel operating instructions, it says ‘ filter must be provided with pressure relief valves to insure that pressure that develops from thermal expansion of the liquid product will not damage the vessel when the system is NOT operating’

So now I’m even more confused, it looks like I need a thermal for the vessel when it’s not operating, then what about when it’s in operation?

Client only mentioned thermal in the spec, nothing else ...
Thanks


Cutie


Thank you, EngTips

 

[jrwchem]

Cutie,

Since this is a filter, the main reason why you would need a thermal relief valve is if the minimum operating temperature is below the ambient temperature. In this case if the filter is blocked in, then heat from the atmosphere can increase the temperature of the fluid inside the filter which increases its volume, and therefore the expansion needs to be relieved. Many times "thermal relief" valves are used to protect piping, and are small "D" orifices. If you already have a "G" orifice i am pretty sure that is big enough to also handle a thermal relief case. The "G" orifice is likely there for protection against fire, blocked discharge, or some other contingency.


JW

 

[EngCutie]

Thanks JW,

The operating temperature is ambient temperature for this project (-10/45DegreeC).

I remember seeing the relief valve sizing software (I don’t have one), the sizing basis has to be filled in- fire or black discharge or thermal, etc, I guess if thermal is selected, it will just give me a smaller orifice?

I already sized the relief valve as its ‘normal’ (with G orifice), and proposed it to the client, but my client came back ask for a thermal, I think if I try to size a thermal, I might get a smaller orifice, then it makes me wonder what happens if the system has overpressure because of other causes other than thermal effects.

It sounds to me that all I need to do is to let my client know the relief valve I proposed can also be used as thermal relief valve. But will a smaller orifice be ok for this kind of application? I wouldn’t mind use a smaller orifice, because it’s cheaper I think…


Cutie


Thank you, EngTips

 

[pleckner]

What causes the large pressure increase in the confined space is the fact that the vessel is 100% liquid full and liquid, not being very compressible, has no where to expand. Once the relief valve pops, it spits out a small amount of liquid, dropping the liquid level in the vessel thereby creating a small but definitive vapor space. At this point the small vapor space will act as a bladder and there is no longer a thermal relief scenario. The relief valve only has to be large enough to allow this little "spit" of liquid.

Any relief valve on the market will be "big" enough to handle the relieving flow and that is why we hardly ever calculate the size. I just install a 3/4" x 1" relief valve and be done with it. As a matter of fact, calculations for a "thermal" relief scenario is not even required by OSHA unless it has to do with large and long pipe lines or "large" totally liquid-full vessels sitting out in a potentially hot ambient environment.

As far as the operating instructions go, you can't just rely on that. The vendor gives you just one of what may be several scenarios to look at from their experience. They have no idea if fire or blocked-in or whatever may be credible. You or some other process engineer must fully document that all cedible scenarios have been evaluated and the pressure relief valve sized for each. The purchased valve will be the largest determined.

And unless you see a P&ID that shows a PSV at the discharge of this "not in your scope pump", you can't assume one is there.

If "thermal" relief is truly the only credible scenario you have then don't worry about it. Document this and specify the 3/4" x 1" relief valve and have a great day.

 

[jrwchem]

Cutie,

I sense it may be worth breaking down the whole basic process of sizing PSVs.

Step 1: You need to consider all credible cases of over pressurization. These cases can include fire, blocked discharge, thermal expansion etc...

Step 2: For each credible case you have to calculate how many lb/hr need to be relieved to maintain pressure below guidelines provided in ASME (e.g. less than 10% above the MAWP generally)

Step 3: Size the required PSV orifice for each case. Note each case has a different back pressure, lb/hr, and conditions. Which ever case requires the largest orifice is said to govern. As long as you purchase a PSV with this orifice then all other cases will also be covered.

Regards,
Jon

 

[EngCutie]

Thank you, pleckner, enjoy your day too.

“increase in the confined space is the fact that the vessel is 100% liquid full and liquid, not being very compressible, has no where to expand” – so this is the blocked discharge scenario? (When outlet valve fails I imagine)

The fluid is oily water; I suppose the worse case is actually blacked discharge. (If it’s gas application, I’d have to consider fire exposure?)

Thank you, Jon. These steps are very helpful. I will certainly follow them in the future when selecting relief valves.

All right, one more question:
Is my understanding correct: fire creates the greatest relieving requirements (biggest orifice size), the second is blacked discharge, then there is thermal (smallest orifice). If I go to a valve vendor without telling them it’s thermal or fire, they would just size it as blacked discharge, so ‘normal’ relief valve is actually blacked discharge. The only difference between them is the orifice size.

THANKS


Cutie


Thank you, EngTips

 

[BenThayer]

if blocked discharge is a credible scenario, it almost sounds like the filter's MAWP was not specificed correctly.

typically a filter should be designed to take the pressure of supply pump as the downstream users take or not and throttle or control the flow, etc. In other words, if the downstream user shuts off the flow, the PSV should not open and spill the contents of the pump. now the tank the pump is taking suction from will be emptied or pumped to wherever the PSV discharge goes.

it sounds like you are being tasked with the sizing and or specification of the valve. if that is true and the process engineer has access to other P&ID's that shows PSV's in place, you should perhaps be confirming that those PSV's are the correct size.

depending on where you are and the codes that govern, some codes do not require filters to have PSV's. the size of the filter or if it is considered to be a piping component could impact this.

 

[jrwchem]

Cutie,

"my understanding correct: fire creates the greatest relieving requirements (biggest orifice size), the second is blacked discharge, then there is thermal (smallest orifice). If I go to a valve vendor without telling them it's thermal or fire, they would just size it as blacked discharge, so 'normal' relief valve is actually blacked discharge. The only difference between them is the orifice size."

The approach you described is unsafe. You need to calculate the relief load required for each case, and then determine how big (i.e. the area in in^2) the valve has to be for each case. Then the largest area is used to specify the valve for purchase.

Every case is different. If the stream composition changes, surface area changes, etc... a different relieving load would be determined. The vendor can not (and should NOT!) provide you with a valve based simply on the statement that it is for a blocked discharge, or fire.

You need to perform an engineering evaluation of each case, and determine how much it needs to relieve. You can not generalize. Well ok, you can a bit... fire loads are generally higher than thermal loads. But you still need to calculate the loads, and quantify how many lb/hr the PSV is required to relieve. the vendor will need to know how many lb/hr are required to pass through the valve, as well as things like temperature at relief, set pressure, back pressure, etc.... Only then can a valve be selected and purchased.

I believe this may be the first time you are being exposed to selecting PSVs. I suggest you have a look at API 521 for an introduction on how relief loads are determined.

I hope this helps.

regards,
Jon

 

[EngCutie]

Thank you, Ben

Hi Jon,

This is my first time. I’m going on a training course next year. But at the moment, I’m really confused, “calculate the relief load” & “determine how big (i.e. the area in in^2)” - how do I do that?

I always thought the vendors had to do the sizing… I have given the vendors: min&max flow rate/temperature/ set pressure/ back pressure. In this case, it’s just a vessel full of oily water. I didn’t tell the valve vendors what the sizing basis was, they all assumed it’s blocked discharge (they did the sizing - G orifice)… I think blocked discharge could be the worse scenario in this application.

My understanding is: this ‘blocked discharge’ relief valve covers the thermal case – If this is correct, I think I can try to convince my client to accept it. At the moment, my client wouldn’t accept it, because it’s not sized as ‘thermal’ (on the spec, it says, all relief valves has to be thermal case)… I can try to get a ‘thermal’ sized, but my worry is: ‘thermal’ has a smaller orifice, there might be worse conditions than thermal expansion (where a bigger orifice is needed)… and my problem is: there is no one knows if such conditions exist or not.

Thanks…


Cutie


Thank you, EngTips

 

[EngCutie]

Oh, I forgot to mention, we don’t have the API521 standard in our company... but I have a copy of the Crosby pressure relief valve handbook, there are some formulas, I think they can be used to do the calculation.

Cutie


Thank you, EngTips

 

[jrwchem]

Good morning,

We need to be clear that sizing a valve, and determining the relief load are two different things. The first is simple (generally for 1 phase), and the other requires engineering skill and evaluation.

The difficult one is determining the required relief load. This is how much mass(lb/hr) the PSV has to pass under the given scenario. You really need to read thru API 521 to get a feel for their recommendations.

The easy one is sizing of the PSV. This means determining the orifice area. This is simple once all of the relief conditions are established (in the hard part described above). This is what the Crosby Manual outlines how to do.

Usually the PSV vendor will do the easy step - size the PSV. Unless you have a different arrangement with the vendor, and they also have the scope to perform the engineering calculations to determine the relief load as well. If this is the case, then you would have had to send them quite a bit more information. Pump curve, design pressures, Filter dimensional drawings, plot plan (for fire circles), etc....


If you are tasked with the hard step (determine the relief load), you really dont have a choice, and you really should read API 521. One text book that might provide some further insight is

"Pressure Safety Design Practices for Refinery and Chemical Operations"

I am curious, do you work for an engineering design firm, or for a chemical/petroleum company?

Regards,
Jon

 

[EngCutie]

Hey, I work for an engineering company (O&G), I’m new here, they used to have subcontractors to do the sizing and purchasing of relief valve, now it’s part of my job, I think that’s why we haven’t got the relevant API standards.

I think all I need is to size a relief valve.

I know the maximum flow mate of the system, but I don’t know how to determine the flow rate when the worst case scenario happens, is it possible for the flow rate to actually go above the system maximum flow rate?

Thanks, I will try to get the book.

Still very confused :S


Cutie


Thank you, EngTips

 

[jrwchem]

Hi,

Yes it is possible for the PSV required flow rate to exceed the normal operating mass flow rate thru the filter.

Since this is a simple filter, the only case that could likely exceed the blocked discharge (i.e. normal flow if you don’t take any credit for backing up on the pump curve) is the fire case. For the fire case you determine the wetted surface area (the whole thing in this case unless the filter is way up in the air with no potential for a pool fire directly underneath -i.e grilled floor), and then use formulas found in API 521 to apply a heat flux. This heat flux can cause the liquid to expand, and eventually boil. Then depending on the latent heat of vaporization the heat input will generate a certain mass flow rate that needs to be relieved. The relieving temperature is the boiling temperature at the relief pressure (Pset +21% for fire case).

You really need API 520, 521 to do this sort of work. I am surprised an engineering company doesn’t have access to these recommended practices. I would ask (demand?!) that they purchase them for you.

Good luck
Jon

 

[EngCutie]

Thank you Jon, I will ask on Monday.

Cutie


Thank you, EngTips

 

[pleckner]

EngCutie:

Please don't take this wrong but process safety calculations are too important to leave it up to the inexperienced. These must be done by someone who is a lot more experienced than you are. I'm not criticizing, just trying to emphasize the importance and danger involved in not getting it correct.

You asked me in a previous post,"...increase in the confined space is the fact that the vessel is 100% liquid full and liquid, not being very compressible, has no where to expand" – so this is the blocked discharge scenario? (When outlet valve fails I imagine)"

This is both the thermal expansion and blocked-in scenarios. You won't have a blocked-in case for liquid unless the vessel is 100% full.

And the answer to your question as to which order of scenario will produce the largest orifice size was too general and at the best, only correct for a typical vessel. For distillation columns, loss of cooling in the condenser will give the largest orifice. Also, there are a lot more scenarios to consider in systems such as control valve failure, pressure regulator failure, loss of power, loss of a utility, etc (see API 521). Yes, these might not pertain to you now but they can pop up latter and you have to know how to deal with them if you are going to be responsible for doing these calculations.

Another thing about PSV vendors, they will size a PSV based on the sizing information you give them but they won't develop the information themselves. You do that or hire an outside expert to do that.

 

[EngCutie]

Hi pleckner,

I think you are totally right. Thank you for your very helpful reply.

That’s how others normally did it:
- Give process details and all relevant relief valve specifications (from clients) to valve vendors
- Vendors size the valves and send back sizing report with quotation

For this very project, I’m not sure whether my client did the safety calculations or not, but I guess they have done, and then decided a thermal relief valve is enough.
I have given all the process details to the valve vendors, but I forgot to mention it was a ‘thermal’, so vendors came back with ‘blocked discharge’, one of client engineers said ‘operation temp is -10/45DegreeC, need a thermal’. If I follow the ‘rules’ all I need to do is ask vendors for a revised quotation with ‘thermal’ to basically comply with client’s specification.

I could have just done that, but I would like to know the reasons behind it because of the “importance and danger” and ‘You can't have the client dictate to you’ as you mentioned in your replies.

My colleague just assumed that the client had other safety systems upstream of the vessel and only needed a ‘thermal’ for the vessel.

The vessel is 100% full when it’s operating. I simplified the drawing, there are control valves, needle valves, etc…

-------------------

At the moment, we don’t have API 521, because of what people have been doing (and we didn’t have to purchase too many relief valves in the past), so bosses didn’t think it’s necessary to have this standard. I have been doing some reading, and with everyone’s help, I understand it a bit more now.

Thanks


Cutie


Thank you, EngTips

 

[tickle]

EngCutie

Specifying "thermal" to the valve vendors will not get you an answer. You need to calculate the required flowrate. This is related to heat input into the vessel (assume it is solar radiation - though it could be due to fire). You will need to know the thermal coefficient of cubical expansion and the isothermal coefficient of compression for the liquid. - simply put; the heat causes the liquid to expand this increases the pressure so you have to relief some liquid at a rate proportional to the heat input (noting that solar radiation will reach a maximum temperature)

It is not a difficult calculation.

Back to your OP - yes in O+G thermal expansion can be the only method of exceeding the MAWP in a vessel (that and fire).