[Dsg-rtpc] relief valve on non-flammable gas supply to gas panel

Mon May 6 16:49:27 EDT 2019

Hi sebastian,

I am willing to go along with whatever Bill wants as long as everyone understands that the system may not work exactly as intended.

Unexpected gas loss - An operator error in adjusting the pressure regulator may result in gas leakage out the relief valve and the cylinder rapidly going empty while in use. This is possible when the relief valve setpoint is lower than the maximum setpoint of the pressure regulator. Low pressure relief valves tend to leak when the gas pressure is close to but below the setpoint. We would need to have extra gas on hand to take these issues into account. Gas cost is $325 per cyl and has a 4-6 week lead time. So, we won't be able to replace empty cylinders if a problem develops during the run. The additional gas needs to be on-hand prior to the run.

The double component failure scenerio of the Pressure regulator failing open at the same time the MFC also fails open is what we are protecting against. The combination of relief valves and flow limiting orifices limit the gas flowrate to prevent dector overpressure. The orifices that are upstream of the pressure regulator limit max gas flow out of the cylinder to less that the capacity of the relief valve. The orifice  just upstream of the MFC, further limits flow.

We have some time before the Jan run, so we can play around with the system if we have the needed components. We would need to purchase 2 additional pressure regulators, $800. Additional ASME relief valves and orifices, ~$200.  Additional fittings, etc may also be required.

The MFC should be tested to verify proper operation at low inlet pressure, using the pre-mix gas, prior to purchase of the new components.

Cheers,
George

--
George Jacobs
Jefferson Lab (TJNAF)
STE 12
12000 Jefferson Ave.
Newport News, VA 23606
(office) 757-269-7021
(cell)   757-876-0480
(email) jacobsg at jlab.org

________________________________
From: Sebastian Kuhn
Sent: Monday, May 6, 2019 2:56 PM
To: George Jacobs; Carlos Ayerbe Gayoso
Subject: Re: [Dsg-rtpc] relief valve on non-flammable gas supply to gas panel

Hi George and Carlos,

Bill Crahen came to our RTPC design meeting today (just when Carlos couldn’t make it). It seemed to me that he and you guys are not quite on the same sheet yet - he suggested that we replace the pressure reducer we have with one that can limit the pressure to perhaps 11-14 psig so that there simply is no way the pressure becomes too high. While this means extra expense, maybe we can save some money by having all 4 HeCO2 bottles manifolded together so only one pressure reducer is needed (I would prefer this anyway since it will guarantee that the bottles are emptied uniformly and the gas composition is constant, which is not the case if each has its own pressure reducer). I will admit that I may not understand all the details, but I am hoping you can perhaps meet with him or give him a phone call and straighten this issue out - ultimately, we cannot win approval of our gas system against Bill’s advice.

Thanks - Sebastian

On May 1, 2019, at 9:06 AM, George Jacobs <jacobsg at jlab.org<mailto:jacobsg at jlab.org>> wrote:

I have added an orifice just up stream of the MFC, FO5, OKeefe G-17-SS. He flow at 75 psi = ~17.8 slm  and He flow at 15 psi = ~3.5 slm. Flow will be slightly higher for the 20% CO2 gas mix. Let me know if this works for you.

Please find attached files.

--
George Jacobs
Jefferson Lab (TJNAF)
STE 12
12000 Jefferson Ave.
Newport News, VA 23606
(office) 757-269-7021
(cell)   757-876-0480
(email) jacobsg at jlab.org<mailto:jacobsg at jlab.org>

________________________________
From: Bill Crahen <wcrahen at jlab.org<mailto:wcrahen at jlab.org>>
Sent: Tuesday, April 30, 2019 11:09 AM
To: George Jacobs
Cc: Bob Miller
Subject: Re: relief valve on non-flammable gas supply to gas panel

George, I could close MV3 and open MV4 and MV5 and directly pressurize the bubbler. These bubblers are glass, and they will not be connected to a system that can potentially expose them to 75 psig. Another note- a regulator is not a safety device. The relief valve (in this case) determines the system design pressure. If it seems more reasonable, we could install reliefs on the lines going to the bubblers (which presumably should never see more than a few inches of water pressure).
On 4/30/2019 9:56  AM, George Jacobs wrote:
Hi Bill,

We should plan to meet and discuss this system once the panel arrives.  Initial tests will be performed with out any detector volumes. Tests with detector volumes will be performed later, once the detectors are ready. Initial tests will be to verify MFC operation and transducer output.

I believe the MFC gas supply requirement specification is 10-40 psi. The pre-mix gas supply pressure regulator has a 0-50 psi output range. The MFC pressure rating is rated far higher, 1500 psi.  Initial regulator setpoint should be 15 psi for testing.

The RTPC gas supply pressure system ends at the 3 valves, MV3, MV4, and MV6. There is no way to shut off gas flow downstream of these valves before it vents to atmosphere. Pressure cannot increase as the gas will simply vent to atmosphere via the RTPC bubbler.

The DMS and RTPC volumes are not capable of holding pressure. They are not pressure system componenets, nor are the oil filled bubblers. The oil filled bubblers will vent gas at ~0.1-0.2 inches water column or ~0.007 psi.

The current  MFC maximum flow rate is 250 sccm. Even if the pressure regulator and MFC both fail open, the 0.017 orifice used on the high pressure side of the gas supply pressure regulators, limits gas flow to <10 slm.

The pressure transducers are capable of much higher pressure that the RTPC and DMS volumes can handle. Since the transducers are connected to the DMS and RTPC, and to the vent bubblers via the DMS and RTPC volumes, pressure cannot build beyond what is required to flow the < 10 slm flow to atmosphere in the worse case scenerio. The 10 slm flow is based on a multiple failure scenerio, ie the worse case, where both the gas supply pressure regulator and the MFC fail open.

The He4 gas supply flows to a He gas bag or balloon. The pressure in the bag is at most 1-2 inches water column, 0.07 psi, and limited by the bubbler oil level. Helium bag gas supplies are a very common setup in many experiments.

In my experience, low pressure relief valves tend to leak. There is also a potential problem using a relef valve with a setpoint lower than the max output of the gas supply pressure regulator. Improper pressure regulator setpoint can result in the rapid depletion of the gas supply. The gas has a 6 week lead time. So, we really don't want to go there.

Cheers,
George

--
George Jacobs
Jefferson Lab (TJNAF)
STE 12
12000 Jefferson Ave.
Newport News, VA 23606
(office) 757-269-7021
(cell)   757-876-0480
(email) jacobsg at jlab.org<mailto:jacobsg at jlab.org>

________________________________
From: Bill Crahen <wcrahen at jlab.org><mailto:wcrahen at jlab.org>
Sent: Monday, April 29, 2019 3:49 PM
To: George Jacobs
Cc: Bob Miller; Carlos Ayerbe Gayoso
Subject: relief valve on non-flammable gas supply to gas panel

George- the last version of your P&ID  I have from you shows a 75 psig
relief valve on both the premix and He cylinder regulators. The gas
panel has several components that would not be happy with that, namely
the pressure transmitters and the bubblers (which are glass). My
understanding is that you only need to supply a required minimum
operating pressure to the MFC (which is 10 psig).

If the relief valves had a 15 psig set point (McMaster Carr 4701K11,
ASME Sec VIII approved), at the flow rate your rotometer is  set for and
the MFC rate would only lead to ~ 0.2 psi pressure drop for 300' of 1/2"
stainless tubing, so you could set the regulators at ~12 psig, which
gives you a 25% margin on the relief valve. Alternately, you could
install the relief at the gas panel (which might be wiser, as this low
pressure unit would not be susceptible to ice, etc.).

Everyone let me know your thoughts- Bill Crahen

<RTPC-ValvePanels-5-1-2019.pdf><RTPC-GasSys-P&I-5-1-2019.pdf><RTPC-Gas-08-13-2018.xlsx>_______________________________________________
Dsg-rtpc mailing list
Dsg-rtpc at jlab.org<mailto:Dsg-rtpc at jlab.org>
https://mailman.jlab.org/mailman/listinfo/dsg-rtpc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/dsg-hallb_rtpc/attachments/20190506/71b43912/attachment-0002.html>