What fluid state to use when sizing a Relief Valve for a fire case if the shell-side fluid is steam? 2

[Rubje]

Context:

-We are trying to add a Relief Valve to protect the shell-side of a shell and tube heat exchanger

-The over pressurization scenario for the exchanger is the external fire case.

-The shell-side fluid is steam which gets condensed. The exchanger is equipped with a steam trap to evacuate the condensate

My question is: if you were to make the calculation, would you assume that the fluid state is liquid or vapor?

API 521 offers two calculations for the fire case, if the vessel is filled with liquid or gas. I'm in conflict with some colleagues as we disagree with the state of fluid.

I say that since it's a S&T and it has a steam trap, the vessel will mainly be filled with vapor. On the other hand, my colleague thinks that since there is a bit of condensate in the exchanger, it's enough to consider it a liquid case and use the liquid state calculation.

Thoughts?

 

[Latexman]

When not running the shell may be full of condensate. When running the shell is mostly vapor. A fire can occur anytime. Size for both and pick the largest relief.

Good Luck,
Latexman

 

[Rubje]

Thank you for the response.

If we consider the exchanger to be filled with vapor, the temperature will surpass the exchanger's design temp (before we reach the design pressure). When this happens, there will be a mechanical failure and the exchanger will essentially self relieve.

In this case, a PSV won't protect the H/E in the fire case, correct?

i just have a hard time imagining that there's a significant amount of liquid since there's a steam trap and it's not a kettle type exchanger. Assuming the trap isn't working is a double jeopardy scenario.

 

[Latexman]

A PSV in vapor service will not protect from over temperature during fire. Other methods can be used, or assume the risk. Also, more than likely, a PSV is still required by Code, so you still need a sizing basis.

It is common/normal practice for maintenance to acid wash and water rinse the HX to improve heat transfer. If a fire occurs then, there is liquid.

Good Luck,
Latexman

 

[Sabeau]

Relief temperatures exceeding design temps are quite common in gas expansion scenarios and care must be taken to ensure the PSV will actually relieve before this occurs.
Latexman gives good advice in sizing for the highest credible relief load, however, you need to establish whether this is just a liquid vaporisation or if the gas expansion scenario is credible.

In some cases you may be able to reduce the set pressure of the valve to decrease the relieving temperature, this is one thing you may wish to look into.

 

[shvet]

@Rubje

When not running the shell may be full of condensate. ... A fire can occur anytime.

Think about a probability of simultaneous fire and overfilling if HE is overfilled only during a couple hours a year. It is up to you to decide what conditions are "normal" and whether to include in scenario conditions other than "normal".

API 521-2020
4 Causes of Overpressure and Their Relieving Rates
4.1 General
... The design is based on the following:
a) the normal operating pressure at operating temperatures;
b) the effect of any combination of process upsets that are likely to occur during normal operations;
...
Table 4—Effects of Fire on the Wetted Surfaces of a Vessel
...
Normal operating level up to the height of 7.6 m (25 ft)
...
4.4.13.2.4 Fire-relief Loads
4.4.13.2.4.1 General
... Ultimately, the user shall decide whether a scenario is credible or not.

 

[Latexman]

Think about a probability of simultaneous fire and overfilling.

Very good advice! This is not shvet's first rodeo, nor mine. SGOTI knows nothing about your industry/business/plant, because you did not share, so we default to our experiences. Mine is semi-batch, emulsion polymerization latex plants. 2-4 batches per day is normal. And, we do have some HX's, not all, that are "down" 2-4 times per day for 1-2 hours at a time with various levels of condensate in them.

In a refinery with planned outages every 5 years or so, it is obviously different.

Also, this problem may, or may not, be a decision worth very much effort. Size the relief both ways and see. Is one a D orifice and the other an E orifice. Both can fit in a 1.5" x 2" PSV body. In my plants, that's about what I would expect for this situation, but I'm going by memory. So, I would size both, and, if it didn't make much difference, pick the largest relief, and move on. In your situation, it could be very different, so think it through. It's what our companies want us to do.

Good Luck,
Latexman

 

[Sabeau]

"Mine is semi-batch, emulsion polymerization latex plants"

Finally we learn where the name is from!

 

[Latexman]

Hi Sabeau,

It was no secret. Take a look at thread1088-468919. I think there are older threads on here about folk's usernames, but that was the first hit I got when I used Search.

Good Luck,
Latexman

 

[Rubje]

@latexman: "Other methods can be used"

What are some of these other methods?

Sabeau: "you need to establish whether this is just a liquid vaporisation or if the gas expansion scenario is credible."

I believe the gas expansion scenario you are referring due is the block discharge case, correct? That's usually the thermal expansion cases in many PSV scenarios.

 

[Latexman]

Adding fire resistant insulation will delay the temperature exceeding the MAWT.

A fire water spray system (not sprinklers) will protect from exceeding the MAWT.

Though I am not that familiar with it, I understand a depressurization system may protect the HX too.

Good Luck,
Latexman

 

[Sabeau]

Hi Rubje,

By gas expansions I am talking explicitly about a vessel or container filled with gas which is expanded due to thermal energy input from a fire.
In many cases (but not all) the design temperatures of the vessel and associated equipment is often exceeded because of the large amount of energy required for the pressure in the vessel to reach a level sufficient to cause the PRV to relieve. This is obviously not as common in liquid vaporisations because of the phase change that occurs and the relatively lower saturation pressures of liquids under pressure.

Hope this helps, please let me know if I have misunderstood your question.

 

[Rubje]

Thank you Sabeau and Latexman!

 

[RVAmeche]

If you're strictly talking about fire protection for a vapor filled vessel, as others have mentioned there are references out there that a PSV won't adequately protect the vessel. It's still required by code of course, but the limited heat transfer means the vessel walls will likely fail first.

 

[Rubje]

@RVAmeche
With the abundance of information that the psv won't protect in fire case why doesn't the code make an exception for PSV requirements in this circumstances? Perhaps the committee just likes to be extra conservative?

 

[RVAmeche]

Not sure, I'd guess the thought is mainly "it can't hurt". We've had some instances where we had documentation that the system was safe by design, ie pumps/etc couldn't cause overpressure and no other sources of overpressure existed. The inspector said I don't care, add a relief valve. So we added a small relief valve and noted that it was added at their request.