What makes the seal? And how to test?

[CHarr]

I have some new radiators (they're Kudox) and some Honeywell TRVs, which come with a tail to screw into the radiator, and then the valve fastens to the tail with a nut and olive.

I understand that the olive, not the thread, makes the waterproof join, and many people say don't bother to wrap the olive with PTFE - it's a waste of time and certainly don't wrap the thread with it, as the thread just provides the compression, not the seal.

So how does the tail form a water tight seal in the radiator? There doesn't seem to be anything except the thread, and I can tighten to a certain point and then no more. I have wrapped these threads with PTFE, but I'm curious about how the seal is formed.

 

[RickMoore]

The PTFE tape makes the seal on threaded joints

 

[CHarr]

Does anyone have any suggestions for a method of testing for leaks in a new CH system before a boiler has been installed? (ie it's not a closed system and isn't connected to a water supply).

I was going to connect up the as yet bare ends of the feed and return pipework to a couple of old lengths of garden hose and fill from a tap, directing the other end to a drain. I'm hoping that if I partly restrict the drain end, I can leave the tap trickling water into the pipes (to achieve a positive but modest pressure) while I bleed the air from each rad in turn, checking for leaks top and bottom.

If all is well, would it be better to leave the (untreated) water in the system pending installation of the boiler, or should I empty everything out again to minimise corrosion?

 

[RickMoore]

Pressure test should be 1.5x the maximum working pressure (assuming copper) so test to 4.5 bar. Make sure the pressure holds, and if physically check joints for leaks/weeps. You could use a pressure reducing valve off the main to fill and give you the correct pressure.
If your planning on leaving the system filled flush and add inhibitor.

 

[CHarr]

Pressure test should be 1.5x the maximum working pressure (assuming copper) so test to 4.5 bar. Make sure the pressure holds, and if physically check joints for leaks/weeps. You could use a pressure reducing valve off the main to fill and give you the correct pressure.
If your planning on leaving the system filled flush and add inhibitor.

I've used polypipe - would you test to a different pressure? Is something like this suitable: http://www.screwfix.com/p/prv-with-gauge-15-x-22mm/41943?

 

[RickMoore]

I think from memory polypipe suggest 6 bar for min 1 hr, and you will need to top up during the test as the plastic expands to do it properly.

To be honest if you test to 5 bar and get no leaks you can be pretty confident in your installation.

That prv would be fine so long as your main has sufficient pressure.

 

[Madrab]

You could also dry test it, I prefer doing this as it avoids any nasty water damage if a there's a leak.
Seal off the pipe work and pressurise using a pump to no more than a couple of PSI using a dry pressure gauge, only use low pressures as there's a lot more punch to compressed air and is only designed as a let by test (checking for leaks) rather than a pressure test
On 2nd thoughts, scratch that, I wouldn't want you to have any problems as a DIYer and as it's plastic pipe you're using