Author Topic: Is it correct to expect a system that will be circulating water to be airtight?  (Read 746 times)

a_scanner_brightly

  • 5 O'Clock Shadow
  • *
  • Posts: 49
  • Location: Eugene, OR

I'm attempting to convert an open loop radiant floor heating into a closed loop one.  I inherited such a system when I bought the house and found it was not very sanitary (or up to code in our state) and ended up disconnecting it from the primary water heater and blowing all of the water out of it (which was nice and brown, mmmm).
 
Someone here recommended that before I try resurrecting it, that I test it for leaks by pushing air into it and measuring the pressure to see if it holds.

I've done this.  Results: it seems to leak, very slowly.   It's possible it had a viable leak and nobody noticed because it was an open open loop system pulling new water in from the main via the primary water heater.

More about the results: I filled it to 35 PSI yesterday and now that I check on it, 24 hours later, it's slid to about 30 PSI.  Previous tests went about the same, and I'd re-tighten all of the fittings and try again.  The rate of pressure loss seems to level out over time.  Perhaps there's a low enough pressure where it will be airtight?  Could probably down-regulate the system to 10 PSI?

But, let me back up and start from first principles.  Is it correct to expect a system that will be circulating water to be airtight?  I can't imagine a system that's not airtight would do just fine being watertight, but maybe this is too high a standard.

Any thoughts?

lthenderson

  • Handlebar Stache
  • *****
  • Posts: 1165
I'll start off by saying I have no experience in water heating systems. However, I would expect that if it can't hold air 24 hours which is the industry standard, it will leak water. While it might not leak much water, my concern would be where that leak is and if the small leak might become a larger leak suddenly when I wasn't home. I think I have heard or read of some sort of gas that can be injected into systems to help identify where the leak is coming from. You might be able to do something like that to identify it.

Jon Bon

  • Pencil Stache
  • ****
  • Posts: 567
  • Location: Midwest
Get a spray bottle fill it with water and and bunch of dish shop. Spray your joints until you find the offender.  Then you can at least move forward with whatever plan you decide.

FYI temperature also effects pressure, but probably only 1-2 psi or so. So it might read differently in the heat of the afternoon and the cool of the morning despite it not leaking at all.

Also how are you measuring the pressure? I know the big box stores sell a screw on pressure gauge for about $8 bucks which is a godsend in doing this type of work.

AccidentalMiser

  • Pencil Stache
  • ****
  • Posts: 632
Yes, if it’s not airtight, it won’t be liquid tight.  The size of water molecules and nitrogen and oxygen molecules are essentially the same.

When you say the “rate of pressure loss seems to level out over time” do you mean that the pressure loss stabilizes to, for example, five psi per day?  Or do you mean it’ll go down to, for example, 20 psi and stop lowering?

Do what @Jon Bon said and “snoop” your joints with soap.  Dawn works really well for this. 

Keep up posted!

a_scanner_brightly

  • 5 O'Clock Shadow
  • *
  • Posts: 49
  • Location: Eugene, OR
Quote
Get a spray bottle fill it with water and and bunch of dish shop. Spray your joints until you find the offender.

Fantastic idea.  Trying this!

Quote
When you say the “rate of pressure loss seems to level out over time” do you mean that the pressure loss stabilizes to, for example, five psi per day?  Or do you mean it’ll go down to, for example, 20 psi and stop lowering?

Every time I've done this it lost air quickly at first and then slowed and stopped.  This latest round of testing I spliced in a dedicated inline sharkbite pressure gauge ($22) to rule out putting on/taking off the pressure gauge as a source of leaks and I also re-gorilla tightened everything one more time and filled it up and left it alone. 

On the first day, right after pumping, it read 35 PSI.

Yesterday it was 30 PSI, a loss of 5 PSI, which prompted my post.

Today, it's still 30 PSI.  A loss of 0 PSI.

Quote
FYI temperature also effects pressure, but probably only 1-2 psi or so. So it might read differently in the heat of the afternoon and the cool of the morning despite it not leaking at all.

I thought temperature might affect air pressure too but couldn't really correlate temperature change to the loss.  Maybe the air warmed up while I was pumping it in (the bike pump got pretty hot) and then as it cooled to ambient it lost pressure?  The system only holds about 5 gallons of water.

Maybe I'll try the dish soap spray test right after pumping it up to, say, 50 PSI and seeing if I get any bubbles.
« Last Edit: September 13, 2018, 11:54:58 AM by a_scanner_brightly »

JLee

  • Walrus Stache
  • *******
  • Posts: 5019
Quote
Get a spray bottle fill it with water and and bunch of dish shop. Spray your joints until you find the offender.

Fantastic idea.  Trying this!

Quote
When you say the “rate of pressure loss seems to level out over time” do you mean that the pressure loss stabilizes to, for example, five psi per day?  Or do you mean it’ll go down to, for example, 20 psi and stop lowering?

Every time I've done this it lost air quickly at first and then slowed and stopped.  This latest round of testing I spliced in a dedicated inline sharkbite pressure gauge ($22) to rule out putting on/taking off the pressure gauge as a source of leaks and I also re-gorilla tightened everything one more time and filled it up and left it alone. 

On the first day, right after pumping, it read 35 PSI.

Yesterday it was 30 PSI, a loss of 5 PSI, which prompted my post.

Today, it's still 30 PSI.  A loss of 0 PSI.

Quote
FYI temperature also effects pressure, but probably only 1-2 psi or so. So it might read differently in the heat of the afternoon and the cool of the morning despite it not leaking at all.

I thought temperature might affect air pressure too but couldn't really correlate temperature change to the loss.  Maybe the air warmed up while I was pumping it in (the bike pump got pretty hot) and then as it cooled to ambient it lost pressure?  The system only holds about 5 gallons of water.

Maybe I'll try the dish soap spray test right after pumping it up to, say, 50 PSI and seeing if I get any bubbles.

How much pressure is this system rated to take?

a_scanner_brightly

  • 5 O'Clock Shadow
  • *
  • Posts: 49
  • Location: Eugene, OR
UGGHH I'm an idiot.

I've been taking a picture of it every day and writing the PSI down, but I haven't been comparing the pictures.   The pictures from each day show that the pressure's been the same this whole time (35).

https://imgur.com/a/P7IhC8a

I've been misreading the pressure each day because there's a different number of hash marks on the 0-50 bucket vs the 50-100, 100-150, and 150-200 buckets.

Why do they design pressure gauges like this?  Is it because the sensor can't reliably sense differences between 0-10 PSI?

lthenderson

  • Handlebar Stache
  • *****
  • Posts: 1165
I was racking my brain at why something would lose 5 lbs of pressure and then stabilize and not lose anymore. I wasn't coming up with anything so it is good to know you figured it out. Onward ho!

Jon Bon

  • Pencil Stache
  • ****
  • Posts: 567
  • Location: Midwest
Nice! We have all been there.

Well if it ever leaks you now know how to fix it.....


Curmudgeon

  • 5 O'Clock Shadow
  • *
  • Posts: 27
Why do they design pressure gauges like this?  Is it because the sensor can't reliably sense differences between 0-10 PSI?

Yep.  Sometimes automobile speedometers are done the same way - more marks between 20 and 30 mph than between 0 and 10.

a_scanner_brightly

  • 5 O'Clock Shadow
  • *
  • Posts: 49
  • Location: Eugene, OR
Thanks for the encouragement! :)

Moving forward with a big parts order to get this project done.  Will report back here how it goes.

Prairie Stash

  • Handlebar Stache
  • *****
  • Posts: 1609

Why do they design pressure gauges like this?  Is it because the sensor can't reliably sense differences between 0-10 PSI?
Yes. that is correct.

most sensors, guages, scales etc. have a "sweet" spot where they're reliable. The sweet spot is the range where its accurate. Outside that range, the measurements are terrible, why bother having hash marks there?

I work with a lot of instruments. Most people I train fail to get that an instrument that can read close to zero accurately is going to suck at reading something large; for a rule of thumb the sweet spot is 10-90% of the range on the instrument. It works with weights, pressures, temps etc., don't go to the extreme high or low of your tools, its better to switch tools, instrument, devices etc.

The mechanical design is just too hard, imagine a sensor thats extemely sensitive for low ranges but also super strong for high ranges, thats a tall order to build. Its far easier to build stuff extremely sensitive OR extemely strong, theres no such thing as a tool that does everything the best.

Also note, zero is an artificial limit. I have guages that have negative presures, zero is just a place holder and not an absolute limit. Instruments that force a zero reading are generally lower quality, when they read under zero you can tell if the calibration is off.