How to slow down the venting of a 9000 psi volume? - 0.003" oriface not enough 1

[JCGJCGJCG]

Hello all - I'm a test engineer, and one of the tests my technicians perform is a proof test on a 42 cubic inch container, at 9000 psi. The catch is that one of the requirements of this test is the container cannot be filled or vented at a rate faster than 1000 psi/min (or 2000 psi/min for a 3 second moving average). This has become a huge problem when venting the container after the test and this rate is often exceeded, especially by inexperienced technicians (we vent it through a regulator to try to control the venting better. A 0.003" oriface was installed in the pressure line between the container and the vent, but it appears to only be effective at lower pressures - say 1000 or so.

I'm a bit at a loss on how else to slow this venting down - I haven't found a smaller orifice that would work with such high pressures. I believe the customer's venting requirements are based on a maximum allowable flow rate and the 1000 psi/min covers the worst case scenario of dropping from 9000 psi to 8000 psi in a minute.

The (simplified) current setup is that a tank of gas is connected to a booster pump, which is connected to a regulator. This regulator then goes into a gas line that goes into my container. There are two valves on the supply line - one is an isolation valve for the container, and another is a valve to vent the system.

One option I'm considering is to better automate the test to remove human error, but I want to also find a physical solution to prevent the vent rate from being exceeded in case of program failure - if the program aborts and brings things to a "safe state," the container will be vented.

Any help is greatly appreciated.

 

[georgeverghese]

Instead of stepping down from 9000psi to atm in one step, one option would be to install several orifices in series. You'll then get more reasonable values for the RO dia.

Also check that the vent line materials of construction are suitable for the lowest temp seen during depressurisation, else there may be a risk of brittle failure if the vent line ruptures due to high vibration.

 

[LittleInch]

I would have thought that what you need is a small piston which you control the speed of movement. Work out the volume expansion to get from 9000 to 8000, around 4to5cubic inches?, then slowly unscrew the piston. When you get to the piston end, shut it off, depressurise the piston, return it to the start point again and slowly draw more gas out. If you use a screw thread to control piston movement you should be able to control well the rate.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[racookpe1978]

Figure out the volume of gas that IS below the required pressure drop.
That is, assume the Delta P from 9000 psig to 8000 psig happens when the gas expands by 10% - but get the exact number.
Translate that 10% drop in pressure => 10% gain in volume => 4.2 cubic inch => how long of a 9000 psig rated tube 1/2 inch in ID. Assume, for discussion, that length is 4.2 inches.

Run your vent line from the container to a vent valve V1 through the orifice (no sense in throwing it away yet) to the a 2.0 length of 1.2 ID tube to a vent valve V2 to a second 1/2 ID tube to a vent valve V3 to atmosphere (or your recovery PV).

Process. End of test.

Close all valves, vent the container to the first tube through V1. Pressure goes down quickly, but cannot go lower than 8500 psig (or so). Close V1. Don't vent from the container yet, but open V2 to reduce the surge pressure by filling both Tube 1 and tube 2. Open V1 again at time = 2.0 minutes. Now, both tubes are at equilibrium pressure with the original container, and the "surge" drop is reduced to within spec because it went from 9000 psig to 8100 psig in a guaranteed minimum of two minutes. Plus, the half-filled tubes eliminate the "shock" by themselves acting as two more slowly-filled "dampers" which is what you really want anyway.

At t = 3.0 minutes, close V1, open both V2 and V3 and vent both tubes to atmosphere.
At t = 3.3 minutes, or when the tubes are empty, repeat everything above.
This cycles all three valves through another cycle, and repeats to reduce container pressure from 8100 psig to 7200 psig by repeating the steps above.
Repeat to go from 7200 psig to (0.9 of 7200 ) = 6480 psig.

Rinse, wash, and repeat until the orifice can maintain flow though the whole series with all three V1, V2, and V3 are open.

 

[JCGJCGJCG]

Thank you all - I think I'm going to try a mixture of George and Racook's suggestions and install additional orifices as well as step the pressure down via an additional tube of a set volume. Things will be slow, but that's what I need right now.

 

[TugboatEng]

Add an accumulator to the system. Maybe if this is a hydro test system have the accumulator charge the system through an orifice to limit the energy transfer during a failure.

 

[moltenmetal]

Consider a choke tube instead of just an orifice. A piece of capillary tubing of the type they use for HPLCs might be useful.

 

[zdas04]

I have done similar work with a series of choke nipples (massive billet about 6-inches long with an OD to match a 2-inch fitting and a drilled opening of any size). For the kind of control you are looking for, I'd use 4 of them in series with at least 4-ft of pipe between the nipples. I'd probably start with something like 1/4 inch hole and cut that hole size in half for each subsequent nipple. In my experience this lash up would keep you out of choked flow in any step (the requirement for choked flow is that the dP be frictionless, the choke nipple is long enough that friction is a significant factor). You might find that this set up is too slow for your purposes and need to adjust the hole size. I've sometimes had to adjust the last nipple to be the size of the next to last nipple instead of half the size. I think you can get these things made with NPT threads in a machine shop with any size hole you want for under $50/each. I always equip them with hammer unions to allow me to drop them in and out without much wasted time.

Stay away from simple orifice plates as their permanent pressure drop is only a fraction of the measured dP.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[EdStainless]

We make HPLC tube, 0.250" OD x 3mm ID.
This tube is strong (cold worked stainless), it wouldn't take a very long piece to create a significant pressure drop.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[ornerynorsk]

Can you cascade into multiple relief containers of known volumes, opening into only one at a time, to effectively throttle the process? The process of filling scuba tanks before the final run-up with the compressor comes to mind. Just a thought.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.