PRV for desiccant dryers? 1

[WheyProtein]

Can anyone provide some tips in regards to calculating the required PSV orifice area for desiccant towers?
I've been referring to API 520 but I am not sure what the relieving temperature should be as
it alters within the desiccant towers. The fluid is compressed air if that helps.

 

[RVAmeche]

API 521 is a better starting point if you're trying to figure out what process scenarios could result in overpressure. Basically, you look at any potential scenario (compressor overpressure, blocked outlet valve, external fire, etc) and calculate the relief requirements for all credible scenarios.

So if scenario is a blocked outlet valve with the compressor supplying 100F air, 100F is your relieving temperature.

Also note if you're sizing an air receiver (dessicant tower) for external fire, the relief valve will not actually protect the vessel. It's required by ASME but in your calculations you should make sure it's known that the vessel will fail during a fire scenario since there's no wetted surface area to dissipate vessel wall heat. Obviously the owner should also be aware if this is credible.

 

[WheyProtein]

Hi RVAmeche,

Thanks for the tip! If for example the temperature changes within the desiccant towers at around 100F during adsorption phase and 200F during regeneration phase, does that mean I need to calculate the required relief rates at both instances?

Also if I have 2 small compressor that has a capacity of 3.5m3/min but have a regulator before the an air receiver which allows a maximum of 5m3/min, then would my required relief capacity be 7m3/min?

Thank you!

 

[RVAmeche]

It somewhat varies depending on your philosophy, but while analyzing a system for overpressure scenarios you typically don't consider double jeopardy scenarios (defined in API 521) where two unrelated things occur. An example would be a power outage and fire happening at the same time. On the other hand, you may lose air pressure and thus lose control of your pneumatic control valves; this isn't double jeopardy.

Look into the dryer operation more. If its a standard dessicant tower dryer skid with two towers; one is typically absorbing and one is offline for regen and not seeing air input directly from the compressor at that point; it may be using a onboard blower to regenerate the material.

Your compressor can output a maximum of 3.5 m3/min (assumed to be at the desired set point of the dryer relief valve; if not find that point on the curve). I also assume both the compressors feed into the one regulator. If the regulator can fail open resulting in 7 m3/min then that's a viable scenario. If the system is configured so a failed open regulator would only result in 5 m3/min, then that's your worst case.

You need to know your process very well to analyze the system for failure scenarios. It's also not something to be taken lightly or a box to check; pressure relief is a critical safety component of a plant. I assume you have a senior engineer who will be reviewing the calculations.

 

[WheyProtein]

You are a life-saver! Thank you for your insight. The dryer does indeed use a separate blower during regeneration phase. I will be sure to get my manager to review my work, I am still new to this field of engineering. Cheers!.

 

[TiCl4]

Most air compressors should already have a PSV on the outlet of the compressor (or receiver tank, at the very least!) that would be rated for 100% compressor speed and blocked outlet case. If the MAWP of the desiccant skid is at/above the setpoint of the PSV, I'm not sure you would have a credible overpressure scenario.

For those more knowledgeable than I, this raises a question:

If the MAWP is greater than 10% of the PSV on the compressor outlet, I don't see a credible overpressure case (other than fire, which is a case-by-case basis). However, what if the MAWP falls within 10% of the PSV setpoint or is the same as the PSV setpoint? Technically the desiccant bed can be sent air above its MAWP, but it, ipso facto, it is being protected by that same PSV that limits overpressure to 10% of the MAWP.

Can the PSV protect both the compressor and the receiving tank in this case (i.e. is this allowed by ASME?) All I could find in the code is below, which would indicate to me that the MAWP of the desiccant bed would need to be >10% of the PSV setpoint to qualify


(UG-125)
(g) The pressure relief devices required in (a) above
need not be installed directly on a pressure vessel when
either of the following conditions apply:
(1) the source of pressure is external to the vessel and
is under such positive control that the pressure in the vessel cannot exceed the maximum allowable working pressure at the operating temperature except as permitted in
(c) above (see UG-98), or under the conditions set forth
in Nonmandatory Appendix M.
(2) there are no intervening stop valves between the
vessel and the pressure relief device or devices except as
permitted under UG-135(d)

 

[RVAmeche]

Typically there's isolation valves between the receiver and compressor for maintenance reasons so one PSV located on the compressor discharge or receiver probably won't be able to cover both pieces of equipment.

I've had inspectors say we need to put a PSV on a receiver even though the compressor couldn't generate enough pressure to hit the MAWP. He said put a relief device on it and we said okay; so sometimes its up to the inspector for that area.

 

[TiCl4]

RVA - what is the overpressure case for the receiver or desiccant bed if that isolation valve is closed (other than fire, of course)? My point is that by closing the isolation valve, you remove the source of overpressure.

 

[RVAmeche]

I agree, but the AHJ inspector said "put a relief valve on it" so we did it.