Hydrotesting; Delta T, Delta P! Need formula that accounts for Thermal Expansion in Pipe and H20.

See this FAQ.
Also get a copy of BS PD 8010 pt 1. This has a substantial section on how to account for temperature change during hydrotest and hence is able to be quoted as an international standard approved by a standards body and not something found from the internet, though I think the FAQ is very good, it will never carry the "weight" of a code.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

The FAQ shows you the magnitude of the pressure change, but doesn't fix your problem. The numbers in the FAQ work out to around 100 psi/F. This means that if you fill the system with 60F water in the morning and pressurize it to 170 psig, by dark it would have either busted the pipe (30F temp rise would raise the pressure 3170 psig) or you would have drained some water out. Then over night temperature drops to 60F and your system is no longer full and parts are under a vacuum.

One way that I've gotten around this on 24 hour tests is to fill the system the day before, pressurize it early in the morning, drain the excess pressure into a container, and use that same container for makeup to maintain test pressure the whole time. If you run out of water you let the pressure drift down. A successful test is one that ends at test pressure without adding any water from anywhere except the tank (you may or may not end up with water in the tank). This only works if the first temperature change is an increase. A clean chart and all liquid accounted for satisfies every client I've ever worked for.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

I'd missed the details last night. How are you currently accounting for the temp drop?

An 11F drop but only 10psi drip doesn't look right unless you are constantly tipping this up, but presumably measuring carefully the amount you're putting back in.

A goods stabilisation period (8 to12 hours) before you start the test is usually a good idea like zdas04 says, fill it one day and start the test the next day.

Pressure testing a low pressure 2" copper line for 24hrs sounds a bit OTT. How long is it?, is it all buried?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

Current testing procedure:
(pipe is hung just below deck (ceiling) in an exposed/open building (not roughed in))

1. Fill line (from lowest point)
2. Bleed off the air at the high points
3. Pressurize the system using a water/air tank (air as the means of supplying the pressurizing force)
4. Take pressure and temperature readings (roughly no less than 30min since system was filled / ie system sat for 30min)
4.a. Temperature is taken with a laser gun aimed at the pipe, (in the shade as it is under a concrete deck)
4.b. Pressure is taken with a dial gauge tapped to a valve the is used on essentially all the tests
4.a.1. Temperature gun measures in increments of 1/2 a degree, aka 0.5s
4.b.1. Pressure measures in increments of 1 Psi (gauge is new for this job, and has been calibrated at the factory)
5. Wait 24 hrs
6. Take final temperature and pressure readings

*7. Per GC/Owner if the pressure drops - test fails! (GC/owner doesn't record/look at/read/measure temperature, only our QA/QC, because we understand thermo and fluid concepts)
*8. Wait another 24 hrs
*9. And wait again... for cryin out load this is TEXAS Weather temperature matters!!!

Yes the math of my original posts data may not exactly jive as inaccuracies in the equipments precision, and trapped air may/is still in the system (clung to the pipe walls).

@Zdas04 we are not allowed to bleed and fill water.

I have to say that the first hydrostatic test I was involved in was 1972. I've designed and executed a few (hundred) tests in the intervening 43 years. The procedure that you described is the first one that I've ever seen (and I've reviewed a couple of other people's tests) that did not have an analog pressure recorder (mostly charts, but some of the new recording electric gauges that capture 10 Hz or better data work). Writing down pressures every half hour on a 24 hour test? INSANE. Using temperature guns? NUTS. Remember the 100 psi/°F? You can pick up or lose a degree in a few minutes. If I was confronted with an owner that imposed the procedure you described, I would refund my fee on any uncompleted work and wish them a good life. The only way that I can see your system passing this test is if everyone agrees to lie about the real data and make up fake data. Not a really ethical outcome.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

zdas04 said:

Writing down pressures every half hour on a 24 hour test?

Click to expand...

They waited 30min before taking the initial reading, not recording every 30min. Sorry for the confusion there.

I suspect you have some considerable amount of air in there which is preventing the water from acting like we would normally expect as a drop of only a few psi for such a drop in temperature means that something is not right and none of the equations etc will produce the correct figures. for A/G pipe, a visual inspection normally suffices and is better at detecting small (pin hole leaks).

Also I'm finding it kind of difficult to work out how you get the water to equalise with ambient air temp so fast. 92F is quite warm water.

For this size of pipe etc can't you use air?

Otherwise you might just need to wait until the temperature drops to the minimum and then start the test from then?

This does sound rather ridiculous though from your client.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

Thank you all,

I have been researching this none stop and with the help here, I have come to the below.

Life is well.. just that "LIFE".. and trying to account for the temperature and pressure in the case is near impossible.

Unsolvable problems:

1. Testing the system over a 24 hr period with the outside temperature constantly changing.
2. Not being able to reach a point where we can confirm without any uncertainty that the system has reached a stable state.
3. Accounting for the different line sizes, tees, 90s, that would effect the "vessel". (temp. expansion/compression effects on the pipe)
4. The tight guidelines/procedures enforced by the GC/Owner.
5. Having a tolerance of "must be at least higher than the initial pressure to pass".
6. Throwing temperature out of the testing procedure (GC/Owner) does not include in their analysis.
7. The accumulation of air in the system that can not be purged via the air vents that are located at the high points.
8. The testing equipment accuracies.
9. (might be a stretch) but using city water in lieu of pure H2O.

Only reasonable solution: Test in a way that mitigates temperature, ie 15min-1hr tests.

Consensus... Since the GC/Owner's testing subcontractor is the one who holds final approval and responsibility on approving the test and whether it is passing... let them use fictitious/weak standing/unbackable values.

Thank you again for the help.

We ran long hydro tests over changing temperature for the USN nuke plants - but didn't worry about "water volume" issues but rather "ran the pump" (a small PD pump) to keep pressure at the required point despite temperature changes. Drained if P got too high. Pumped if P got too low.

The "leaks" - if any at all! - were assumed through valve seats, because all joints and the new welds were monitored. No water through a joint at max pressure? No leak at that joint.