Sunday, 24 June 2012

Brewery Build #14: Fit-out, Leak Test & Burn-in

After spending some time, and copious quantities of pink plumbing tape screwing all the fittings together, fitting them to the kegs and getting everything setup, I've come to some realizations that would have made the build-process easier had I known them in advance.

Tapered fittings (such as American NPT) work best for pipe-to-pipe type connections, such as attaching valves to fittings or camlocks.  The tapered thread, combined with some thread-tape, locks the fitting in place and prevents leaks without additional seals or gaskets.

However, tapered NPT fittings are less useful when making bulkhead-type connections, such as connecting fittings through the keg walls.  Due to the tapered thread, the nut or fitting must tighten exactly as it compresses the gasket or washer against the wall of the vessel  However, especially when thread-tape is applied (which is essential to ensure there are no leaks along the thread) the fitting can easily lock-tight too quickly and not seal to the side of the vessel or not tighten quickly enough, potentially allowing for leaks along the not-sealed-thread.
In the photograph above, both a steel and silicon washer (not visible) are required before the nut can tighten the taper-threaded nipple to the side of the keg.

On the other hand parallel threaded fittings (such as British Standard Pipe Parallel thread (BSPP)) work much better as bulk-head connections.  Thread-tape prevents any leaks along the thread as the connection tightens securely against the side of the vessel.
In this case the parallel threaded nipple required only the locknut and a silicon o-ring to make a water tight seal on the side of the keg.  The disadvantage with parallel threaded fittings is that when creating pipe-to-pipe type connections an extra seal or gasket is needed that the fittings must fully 'tighten into' in order to create a water-tight seal.

With this in mind - and knowing that at 1/2 inch size both BSP and NTP threads mate without problem - bulkhead type connections through the wall of a vessel would be made best with an parallel BSP thread, while fitting-to-fitting connections like the ball valves, are best using tapered NTP thread.

Since the Chinese Ebay supplied lock-nuts have a groove cut in them (which most local fitting suppliers and LHBS usually charge considerably more for), the easiest and cheapest method for sealing through-keg bulkhead seals was simply to use the nut and a silicon o-ring (as per the picture of the thermowell above).  However as the fitting is tightened the o-rings can deform or break easily making them less than useful, and essentially useless in any situation that does not have a channeled-nut for them to sit inside.  As a result many of the fittings required both a silicon washer and stainless shim-washer so that the fitting would hold tight against the wall of the keg.

While it might reduce the life of the seal, and potentially be the source of contamination (but since the seals are silicon this should not be a problem), most bulkhead fittings were sealed on the inside.  This was to stop the water/wort from getting into the threaded connection and between the fitting and the keg-walls, which could happen more easily if the seals were located on the outside of the keg.

With all the fittings attached it was nice to see that the sight-glass worked as expected.  The silicon is transparent enough to easily see the water-line, however over time, I expect it will go cloudy so the little floaty-ball should be more useful.
With everything attached, fitted and sealed, it was time for a leak and burn-in test.  As with most things, the first few fittings that I attached leaked a bit and needed to be tightened or re-assembled, but by the time I was finished all of the seals were water-tight first time.  Once that was done it was time for a burn-in test, or more correctly a boil-in, by connecting the element directly to the power-socket and running it for about an hour.
While such a test, after all the leaks had been eliminated, might appear pedantic, in this case the silicon seal 'slipped' - likely due to the fitting being over-tightened - after about 45mins and ended up spilling boiling water all over the floor.  Since the HLT will be elevated, I'd much rather find the problem by testing, than when it was set in-place on the stand at eye-level.
The boil-in test also showed the consequence of my being lazy and not carefully cleaning the inside of the keg after cutting, drilling, filing and working with the keg, small patches of rust developed during the 1 hour long boil:
Most likely these were the result of small contamination on the stainless surface from the various drilling, cutting and filing processes.  Luckily using some of the Stainless Cleaner, a scrubby and a bit of rubbing removed them fairly quickly.
And soon the inside of the keg was looking as good as new again, next time I'll make sure I clean them more pedantically after any metal work.
In order to minimize the risks of the potential weak-point where the cord-flex attaches to the electric element, I wrapped some tape and spiral wrap around that point and then attached the cord securely to the plug housing.  Hopefully this should reduce any pulling or flexing at the point of weakness, and increase the safety and longevity of the fitting and element.
Even after the 1 hour test-boil the cord and connection were quite warm, so it's easy to understand how a bit of pressure or pulling could cause problems.

1 comment:

  1. For sure many entrepreneurs with small and big businesses are going to benefit from this. Keep it up!
    Plumber Birmingham, Al