Hydrotesting entrained air calc. 1

That's one of those very difficult questions. I could see calculating/estimating the amount of air if you have pockets in the piping with no high point vents. But if you are just talking about filling the line with liquid and calculating how much air is entrained in the liquid during the filling process, good luck. It's one of those numbers personally I wouldn't have much faith in, you could use it to prove just about any sort of hydrotest results IMO.


Don't tell me, an non-engineer wrote that procedure

TD2K,

Those are my feelings also and yes the proceedure was written by a non-engineer. My hope was not to use a high entrained air estimate to justify a hydrotest but to have it at the begining of the hydrotest, just after pressuration to know if the feild engineer had to drain the line back down and re attempt a better fill.

Hello,

Procedure could be time consuming. I once wrote a computer program to do just this (it took about 80 hours to write). Think ideal gas law (for changes in temp & pressure), compressibility of the hydro test fluid, expansion of the pipe segment from pressure and temperature changes (typically only radial for buried pipe and axial & radial for above grade), and fluid expansion/contraction from temperature change. Then use superposition theorem to shove all of these variable together. Each equation is fairly simple. Look in Roark’s Formulas for estimating the volume change of the pipe segment from pressure. If you leave temperature out of the equations, the calculations are not too bad. Just iterate to the solution.

IanMiller (Mechanical):

TD2K has hit the nail on the head. Don't place any criteria or faith on instructions prepared by lay people. I further concur with TD2K in acknowledging that we, as engineers are on our own on this practical and tough subject of assuring a 100% liquid fill in a complex object such as an underground pipeline.

You're going to have to put a lot of faith on your field pipeline crew and the elevation details they can furnish. I would not take the dissolved air (or gas) in the water, since this will stay dissolved - especially as the pressure is increased. I would be on the look out for high traps where the air will segregate and accumulate from the water and plan on positively venting those. You can only do as good a job as the basic data you are given. I don't envy your position.

Lots of luck, and safe testing.

Art Montemayor

Hello,

After reading Montemayor’s response, I realized that I did not state what I thought was obvious. Stupid me. I assumed that you would estimate the amount of entrained air after you had some data from the hydro test. Look at the pressure rise verses the amount of fluid added during the hydro test. Then iterate solutions for the factors stated in my last note.

If you are trying to do this before the hydro test from elevation and survey data, it is not practicable (too many uncertainties.) The reason one would be interested in the amount of entrained air is to evaluate the sensitivity of the test. If there is no significant amount of air, the test would be sensitive to a leak. If there was a significant amount of air, there could be small leaks with little pressure change during the test.

On the first page of my program you fill in data about the pipe that can be obtained before the fill (and should be). That will be used to estimate the volume of the line and the amount of water needed to bring it up to pressure. Then you would add data about the test such as begining and ending temperature, and the amount of water that was added or subtracted durring the test. Then in the background it will accomplish all the necessary calculations and output a statement like "This is an acceptable hydrotest". That much I have completed, and even have a second program which can be used for antifreeze in the works. These have been the hidden agenda of the questions I have been asking over the last few weeks.

Now it has been suggested that the entrained air piece should be included in my assesment. If we included the data to form a PV curve durring the pressuration, the deviation from the ideal curve should be related to the amount of entrained air. Then if the % entrained air was too high we would require the line fill be redone and or the high points be vented. This will probably eliminate many of the hydrotest which pass with small leaks because the large amounts of entrained air mask the pressure drop. This is something that does not currently occur on our hydrotest because no one knows how to complete the entrained air calculation.

I would under no circumstances allow a high entrained air content to be used as a reason to pass a test with poor results.

So Thats where I'm going. I have one week. Thanks to everyone for their help.

Ian Miller

Additional comments to the above assuming you are testing a pipeline:

1. Use a pig for filling to minimize the amount of air. May require temp pig lauchers and receivers. You will also need these for drying the line after testing.
2. Use a meter to measure the amount of water fill. Knowing the pipe volume and making an adjustment for the expansion of the pipe diameter due to pressure then you should have a better guess.
3. You should have a stabilization period after filling and bringing the line up to pressure.
4. The Haliburton manual has a bunch of equations to acount for temperature, etc. Do not know where to get a copy. They somehow just get copied and passed around.
5. Watch out for elevation changes so that the yield strength will not be exceeded at the low elevation point. Testing a hydrocarbon pipeline with water can cause problems because of the difference in liquid density.

Ian,
I'm currently on a committee writing a safety manual chapter on static testing. Our approach to this topic was to require a hold period after filling. The pressure during this period is to be held at a low positive value and you can keep adding liquid until a high-point vent stops issuing gas and you stop having to add liquid to maintain a constant pressure. At that point you raise the pressure to test pressure and may not add any more liquid until the test is complete (but you can drain some if you have temperature problems). We will allow some small pressure drop (less than 5% of test pressure) over the test period to account for temperature decreases.

Basically we avoided a very difficult and approximate calculation by adding the soak period.

David

As 1969grad said the best way to fill is with a hard poly pig and a slow fill. Depending on pipe size I like to try and run about 5 to 6 barrels a minute max. Any high points will need a temp flange or test head with a bleeder valve and make sure of a good flow of water through any bleeders. Then I bring the line to full pressure to check for leaks at any flanges or threaded fittings. The next step is to let the temperature of the fluid stabilize with a few hundred lbs on the line. Production doesn't like this because they are losing money. Depending on where your water is coming from and at what temp this can sometimes take several days but if the water is cold going in usually overnight is enough. As a layman with well over a 1000 tests behind me I can tell you that a good one that is fully underground will not gain or lose a pound in an 8 hour test. I can send you the calculations for press v temp change with figures for pipe size, wall thickness, length and above ground factors. Your underground temp probe needs to be away from any above ground piping. Tom