jari001
Chemical
- Aug 9, 2013
- 478
Hi everyone,
I have a pilot metal system where the burst discs (with a burst indicating strip) of the vessels are joined to a piping network leading to a crash tank via sanitary connections. I am able to perform outboard leak testing using helium on this system and have found that the sanitary flange connection has an average leak rate of 10e-4 mbar*L/s. The system has vessels that range from 200L to 50L and interconnecting piping (total length on the order of 200ft). The lab scale system for the same chemistry is an order of magnitude smaller vessel volume wise and very little tubing runs; KF connections are used on the lab scale system instead of sanitary connections. The glass system had an inboard leak test performed at the KF fittings and their average leak rate was 10e-6 atm*cc/s. Is there an accepted method to compare these leak rates given the geometry difference of the gas flow path (even if it was one style of fitting despite the leak testing style)?
My current thought is to compare the leak rates normalized to the respective dP (absolute value of the difference between the system pressure and the pressure of the surroundings) of the two different methods given the simplification that the actual leak behaves as flow through an orifice is both instances.
Thanks for your time.
I have a pilot metal system where the burst discs (with a burst indicating strip) of the vessels are joined to a piping network leading to a crash tank via sanitary connections. I am able to perform outboard leak testing using helium on this system and have found that the sanitary flange connection has an average leak rate of 10e-4 mbar*L/s. The system has vessels that range from 200L to 50L and interconnecting piping (total length on the order of 200ft). The lab scale system for the same chemistry is an order of magnitude smaller vessel volume wise and very little tubing runs; KF connections are used on the lab scale system instead of sanitary connections. The glass system had an inboard leak test performed at the KF fittings and their average leak rate was 10e-6 atm*cc/s. Is there an accepted method to compare these leak rates given the geometry difference of the gas flow path (even if it was one style of fitting despite the leak testing style)?
My current thought is to compare the leak rates normalized to the respective dP (absolute value of the difference between the system pressure and the pressure of the surroundings) of the two different methods given the simplification that the actual leak behaves as flow through an orifice is both instances.
Thanks for your time.
