, ,

The gas line lets oxygen in too – tubing, ball locks and connecting

Posted by

–

In How much oxygen gets in? we tested what lets oxygen into a closed fermenter. But a fermenter isn’t closed all the time – it’s connected to CO₂ for pressure and transfers. So we ran the same tests on the gas line: the tubing, the valves, the ball locks, and what happens at the moment you connect.

In short:

  • A gas line with an EVA tube let in 30 µg of oxygen per hour, and 82 µg/h with the gas connected. With a stainless pipe instead, 12.8 µg/h.
  • Connecting a ball lock pushed in about 600 µg in one go. With a ball lock that bleeds as it connects, about 20 µg.
  • A gas line that had stood closed for a long time had collected about 1,700 µg, which went straight in when the valve was opened.
  • With the valve open and the gas on, the supply added about 20 µg/h.

Same chamber, same method

The tests ran in the same chamber and at the same pressure, 2.8 bar absolute, as the gasket tests, and the ingress is calculated the same way. The reference is the chamber with only a ball lock post and nothing connected: 8.2 µg/h.

Tubing

In the first tests the gas line had about 2 m of Kegland EVABarrier tube, 5 mm ID × 8 mm OD, between two push-in fittings and a valve. That alone let in 30 µg/h, with the gas disconnected. With the gas connected through the regulator, 82 µg/h – more on the regulator below. When we replaced the EVA tube with a 316L stainless pipe, the same line let in 12.8 µg/h, still a little more than the reference because of the push-ins and the valve. EVABarrier is sold as a barrier tube, but two metres of it let in about 17 µg/h more than the stainless pipe.

Valves and ball locks

Festo makes stainless 8 mm push-in fittings specified for oxygen, with 1/8″ thread, and we use them in many places. The “Festo” ball lock in the tests is a stainless ball lock with its usual connection removed and replaced with a Festo push-in.

With the valve closed, the connectors themselves made little difference. A Camozzi or DuoTight valve with a short PTFE tube and a ball lock with a Festo or DuoTight push-in gave 8.4 to 9.0 µg/h – the same as the bare post. Even 10 cm of EVA between a DuoTight valve and a DuoTight ball lock didn’t show. The combination of a Camozzi valve, 10 cm of EVA and the Festo-fitted ball lock, on the other hand, leaked so much gas that it couldn’t be measured.

The lowest of all was the ball lock we use today, with a welded 8 mm stainless stub going straight into a DuoTight valve and no hose connection on the ball lock at all: 4.5 µg/h, which is less than the bare post with nothing connected. We don’t have a good explanation for that yet.

The moment you connect

The biggest numbers came from single events, not from steady ingress. Connecting a purged and pressurised gas line through the Festo-fitted ball lock pushed in about 600 µg of oxygen in one go – most likely the air trapped in the connector as it’s pushed onto the post. With a ball lock that bleeds gas as it’s connected, the same thing added about 20 µg.

That’s the ball lock we use today. Instead of a hose connection it has a welded stainless stub that goes straight into a tap and a T-piece, and we’ve removed the O-ring in the valve that opens when the post is pushed in. With the tap open, CO₂ bleeds out of the connector and flushes it before it seals on the post. The branch with the second tap is for flushing the hose when the ball lock has been connected for a while without pressure.

The worst was a gas line that had stood closed for a long time. The gas in the hose and the regulator had picked up oxygen, and when we opened the valve, about 1,700 µg went straight into the chamber.

We don’t remember exactly how long the line had been closed, but there were two days between this test and the one before, so it may well have been days.

For scale: in 60 litres of beer, 600 µg is 10 ppb and 1,700 µg is 28 ppb – from a single connection. 600 µg is what the reference chamber let in through its gaskets in three days; 1,700 µg is more than eight days’ worth.

Gas on

With the valve open and the gas on, the chamber took in 23 µg/h, compared with 3.8 µg/h with the valve closed in the same set-up. The gas supply itself added about 20 µg/h.

We suspect the regulator – even though we had already replaced its hose fittings with compression fittings and sealed all threads with fresh Loctite before the tests.

What we learned

  • Keep EVA tube out of the gas line where you can – stainless pipe cut the ingress from 30 to 12.8 µg/h.
  • Bleed the ball lock when you connect: about 600 µg without, about 20 µg with.
  • Flush a gas line that has stood closed before you open it to the fermenter.
  • The gas supply isn’t oxygen-free either.

All the numbers are in the table at the end of How much oxygen gets in?, and the instrument is described on The oxygen lab.