PDA

View Full Version : Pressure relief valve



DaBit
06-08-2003, 01:52 PM
Hi,

I am looking for a pressure relief valve which should open at approximately 20 bar (~300psi), and dump the (high side) overpressure to the low pressure side of the refrigeration system.

Compressor unit will be about 1HP running R404a and ethane in an autocascade.

What valves are recommended for this purpose?

Many of the valves I thought suitable for this, such as the Danfoss KVR (which is in fact a condensing pressure regulator) are not designed for safety purposes.

cooltrain
06-08-2003, 04:31 PM
How about using a solenoid valve to be controlled by a pressure switch sensing at the relief valve outlet? ..... When the relief valve opens the pressure switch energizes the solenoid and opens a line to the dump zone. Of course you need pipe linkage from the relief.
That is my simple guess and could be cheaper too than a special valve.

In larger system, we use the built-in recovery unit to contain the gas into a cylinder using solenoid and p-switch principle.

DaBit
07-08-2003, 07:09 AM
Originally posted by cooltrain
[B]How about using a solenoid valve to be controlled by a pressure switch sensing at the relief valve outlet? ..... When the relief valve opens the pressure switch energizes the solenoid and opens a line to the dump zone.

Why not pipe the relief valve directly to the low pressure zone?

I suppose that relieving overpressure to the low side is fairly common practice; relieving it to air should be done on emergency only.

Thus, I suppose that standard valves exist for this purpose. I only have to know which ones.

The problem I have with solenoid & p-switch is the dependancy on electricity. If anything fails (p-switch, solenoid, wiring, power supply), the system doesn't work anymore. Though the HP cutout will prevent catastrophic failure, this is not a desirable situation.

A mechanical valve is far more reliable than electric circuitry.

FreezerGeezer
07-08-2003, 10:07 AM
A PRV will get weaker over time, and at some stage will start to lift before you want it to. Many of our clients treat a PRV as a single use valve.
Perhaps there's a no return valve on the market that can be set to only allow flow over a certain pressure one way, and stop it the other way - this may be better than a PRV.
However, I would be inclined to modify cooltrain's system.
Fit a tee before the solenoid valve, with a PRV on it, venting to atmosphere. That way, if there is any problem with the solenoid valve, the PRV will act as an ultimate safety, and vent the over pressure refrigerant to atmosphere. Obviously, the PRV would want to be set for a higher pressure than the solenoid system control.

DaBit
07-08-2003, 11:14 AM
Originally posted by FreezerGeezer
[B]
Perhaps there's a no return valve on the market that can be set to only allow flow over a certain pressure one way, and stop it the other way - this may be better than a PRV.

Something like an EPR?
The pressure setting of ~20 bar might be a problem with those valves, though.


However, I would be inclined to modify cooltrain's system.
Fit a tee before the solenoid valve, with a PRV on it, venting to atmosphere.

Then I would prefer using cooltrains system, with an additional HP cutout switch.

Still, I would prefer a mechanical valve; I suspect the HP switch/solenoid to start oscillating when the set pressure is reached.

Gary
07-08-2003, 05:17 PM
Compressors have an internal relief which dumps high side pressure to low side.

FreezerGeezer
07-08-2003, 10:18 PM
That's true, Gary. However, I think that DaBit is trying to create a pRV safety system that does not vent to atmosphere if it can possibly avoid it?

DaBit, if you set the H.P. control up correctly, you won't get the short-cycling effect.

DaBit
08-08-2003, 01:05 PM
Originally posted by Gary
Compressors have an internal relief which dumps high side pressure to low side.

Is this relief suitable for few-times-safety only, or could it be used for every start?

I foresee another problem with using this builtin valve: when it becomes active, both the R404a and R170 in the discharge gas mixture is dumped to the low pressure side. This is not what I actually want; the R404a must be led to the expansion orifice so it can do it's duty, while the R170 must be dumped to the low pressure side since it cannot condense yet.


That's true, Gary. However, I think that DaBit is trying to create a pRV safety system that does not vent to atmosphere if it can possibly avoid it?

I don't want to vent at all to atmosphere. That would destroy the balance of the autocascade refrigerant mixture.
The problem is: during startup the condenser for the ethane is not yet down to temperature, so I cannot condense the ethane at that moment. Therefore, to prevent the high side rpessure from skyrocketing due to ethane not condesning, I must do something with it, and just recirculating it until it can condense is the easiest way.

Gary
08-08-2003, 03:14 PM
The internal relief is a safety device only. The pressure should never get that high.

When the compressor starts, the bulk of the refrigerant is shifted to the high side. How high the pressure goes is determined by the amount of ethane vapor and the internal volume of its container. It is not going to skyrocket without limit. It will rise to its maximum pressure and stay there until it starts condensing. If that maximum pressure is too high then the internal volume of the high side needs to be increased.

DaBit
08-08-2003, 04:50 PM
Originally posted by Gary
If that maximum pressure is too high then the internal volume of the high side needs to be increased. [/B]

I already though about that.

I basically have two options:

- Keep the volumes as small as possible to limit the charge needed. Limiting high side volume is particulary effective since pressure is 10-15 times higher than at the low side. Thus, every high side cubic inch counts as 15 low side cubic inches. Furthermore: what I noticed during construction of my captube R134a liquid chillers is that fixed orifice systems seem to be more responsive to load changes when high side volume is kept small.

This also means that I need a mechanism to keep high side pressure limited.


- Increase high side volume. For me, this has a few disadvantages: more refrigerant is needed, and I have no idea how much high side volume I need since I have only a ballpark idea how much refrigerant I need to get the system working.

Hmm, luckily I never thought this was going to be easy :D

cooltrain
09-08-2003, 06:41 AM
Hi DaBit
Allow me for another try. As you said that you're not happy with the electric circuitry like I've suggested, then this one might work.

A simple check valve between the relief valve and the dump zone will only allow a flow when the relief outlet pressure is greater than the dump zone.

In this manner you may prevent a glitch.

Hope this will help.

DaBit
11-08-2003, 09:13 AM
Originally posted by cooltrain
Hi DaBit
Allow me for another try. As you said that you're not happy with the electric circuitry like I've suggested, then this one might work.

I might pick it as the 'best available solution'...

I thought this was just picking the right valve for the job, but that seems to be harder than I imagined.


A simple check valve between the relief valve and the dump zone will only allow a flow when the relief outlet pressure is greater than the dump zone.

But then I still need a valve which opens at a certain pressure...

Gary
11-08-2003, 11:25 AM
I'm thinking a CPR at the compressor inlet might do the job, decreasing the pumping capacity of the compressor to limit the high side pressure, while increasing the proportion of refrigerant held in the low side until temps drop.

DaBit
11-08-2003, 11:32 AM
I cannot imagine what you mean. A CPR is a valve meant to maintain a sufficiently high pressure in the receiver, right?

How exactly would one of these valves do the requested job at the compressor inlet?

Gary
11-08-2003, 11:43 AM
A CPR is a crankcase pressure regulating valve, whose purpose is to limit the load on the compressor. It maintains a maximum set pressure at the compressor inlet.

DaBit
11-08-2003, 12:24 PM
Originally posted by Gary
A CPR is a crankcase pressure regulating valve, whose purpose is to limit the load on the compressor. It maintains a maximum set pressure at the compressor inlet.

This would indeed work, if the compressor shell was not the largest storage of refrigerant vapour in the system. That 2500-3000cc of free volume is probably sufficient to hold all the vapour I need.

And I cannot control that volume with a CPR :(

Derek
14-08-2003, 02:12 PM
Just for information.

We had Henry PRV's on some R22 and other more exotic fluids. Site regulations mean we check operation annually.

In general they worked OK but 1 in 8 stayed closed to 150% of cracking pressure (corrosion). 5 in 8 were non linear opeing at much lower conditions each time (mechanical fatigue) and 2 in 8 stayed open (seal degredation and corrosion).

Generally I have to support PRV's as 'one shot' items.

On a more recent system we moved to all stainless and used Swagelok valves these were repeatable (should be for the price) but we had elastomeric seal issues resulting in random opening and remaning open after operation. Not a Swagelok issue really as our HFC's (R404A and R245fa) and lubricants have a random effect on everyting from PTFE to EPDM and the valve and refrigerant manufacturing industry has yet to catch up with material compatibility. Typically we are told 'its HFC compatible' bit like PC's the cheaper the component and the more complex the use the less compatible it probably is.

Since the UK PER (EU PED) we have just implemented the fitting of bursting disks to lines (high burst rating) for any potential hydraulic lock risk locations. Again these require annual visual inspection and typically I propose replacement (corrosion risk mainly)

If venting from high to low is required then an all (compatible) metal wetted parts pressure regulator or PRV is the grail. Electrical systems and anything with software will fail its only natural.

The only real way out for designers is a risk analysis for all the systems potentail failure modes where you mitigate the risk in someway. Solicitors like this kind of thing.

DaBit
14-08-2003, 02:27 PM
Originally posted by Derek
[B]Just for information.
Generally I have to support PRV's as 'one shot' items.

That seems to be how everybody thinks about them.


(R404A and R245fa) and lubricants have a random effect on everyting from PTFE to EPDM

Now you really have my attention. What effects have you seen from R404a contact with PTFE?

I am intending to use PTFE hoses with stainless braid to transport the suction vapour.


If venting from high to low is required then an all (compatible) metal wetted parts pressure regulator or PRV is the grail.

Any suggestions on which one to use?


Electrical systems and anything with software will fail its only natural.

I know; I am designing electronics and software for a living. In one of the earlier replies in this topic I wrote: The problem I have with solenoid & p-switch is the dependancy on electricity. If anything fails (p-switch, solenoid, wiring, power supply), the system doesn't work anymore.


The only real way out for designers is a risk analysis for all the systems potentail failure modes where you mitigate the risk in someway.

Very important, and hard stuff.
But in this case I want to 'misuse' the PRV to solve a design problem, not for safety only (though it is a pretty side effect).

It seems that increasing high side volume and keeping static charge limited is about the only way to do things failsafe.

Derek
16-08-2003, 03:11 PM
R404A, R22, PFC's and POE's

We have experienced a exposure based increase in thin wall PTFE seals of 5 to 10%. Tendency is to jam valves open on quick release couplings and PRV's to either open slowly (weeks) or jam tight. Radial seals are the most prone. I have assumed any compression seals actually benefit from this action.

Only sucessful material so far has been DuPont Delrin. When you look at the chemistry any fluoro fluid is going to react in someway with 'fluoro' materials. Should see the effect of fluorosilicate o rings with R123.

I'm using Swagelok (those guys had better put me on commission soon) 4R3 (format PRV's as both high to low and as a pressure bypass for R404A and with, after a few 'losses' modified neoprene seals. Check out www.swagelok.com. They supply a range of spring and seal kits so in development we didn't need to by a shed load of valves.

Sound like your doing something similar to us. Most of my life is spent 'righting the wrongs' of systems engineers who think refrigerant is like software and can be programmed to know what to do!