Water content measurement in gas well stream

[Mafuta]

I would wish to know the best way to measure water content in the gas well stream upstream of the inlet separator (free + vapour water)for accounting in the daily production report. I have a problem with separations, so I end up having incorrect water volumes.
What is the normal practise in getting daily water volumes out of gas process facility?

 

[zdas04]

Everyone has incorrect water volumes. The industry and the regulators have taken a "don't ask, don't tell" approach and the worse a number is the more everyone tries to pretend it is good.

For water vapor, the industry has always ignored it. If separator pressure is above 50 psig or so, this isn't horrible since the 100% RH point is around 300 lbm/MMCF at 50 psig (less than one bbl/MMCF). I have a client who is using vacuum as a deliquification method and all of the water that gets to surface is vapor (around 5,000 lbm/MMCF or 14 bbl/MMCF), and since there is never any liquid water upstream of gas measurement, they report zero water.

Liquid water is no better. The most common technique is to use a turbine meter off the dump line. The latency of a turbine meter is around 4 seconds and the dump duration on most of the separators I see is around 3 seconds--when the dump opens, the turbine over ranges, but the time it is coming back in range the dump is shut. Turbine meters on dump lines are basically very poor dump counters.

Some clients are starting to put dump counters on blowcases (since the blowcase is isolated from the inflow during the dump, the amount of water dumped per cycle is very stable and the volume is excellent. Other folks are using Vortex meters on the dump line with pretty good results (latency is around one second so you are actually measuring liquid part of the time.

Your question seems to indicate that you want to do multi-phase measuring upstream of the separator. There are starting to be meters on the market that claim to be able to do multiphase metering--they can't do as well as +/-20% so trying to use one for a material balance would just be fun with numbers.

David

 

[Mafuta]

Thank you David. That's very invaluable contribution of yours!
Yeah, I heard of fascinating stories about these multi-phase metering and I want to get the professional opinions about it.
I'm about to start a project in which I will have to set-up a mass/energy balance from u/s to d/s for the gas system. The aim is to have the easiest way to detect gas loss along, especially in the d/s where a number of customers are connected and possibly one can temper with the system and fool our meters.
Which way is the best, mass balance or energy balance? I'm confortable with mass balance, but there are opinions around that energy balance is more appropriate. Anybody can share the exprience please.

 

[zdas04]

"Energy balance" is just a mass balance times the specific energy of each constituent. Water goes away. CO2 goes away. Nitrogen goes away. The specific energy numbers usually come from a periodic gas analysis. Well production changes the mix of products many times per day (actually many times per second) and there is so much specific energy difference between methane and hexane (for example) that very small changes in the hexane volume percent make a huge difference in the stream's total energy.

I've done energy balances on field data and the results have never been reassuring. It hasn't worked even with online chromatographs.

I've had better luck with mass balances, but even that has been problematic since well production is so variable over such a short interval (as an industry we like to pretend that monthly averages are "close enough" or "representative", but I've gotten 1/100 second data and seen a glimpse of the real variability and it is pretty big and really fast). The coarser the data, the better chance you have of doing a successful system balance (as long as you define "success" sufficiently coarsely). I've been successful finding gross problems (e.g., what trunk a large leak was in, or a backwards orifice plate in a big station), but small problems get lost in the noise.

David

 

[Mafuta]

I do agree with your opinions. Thank you for your for taking your time to contribute.

 

[dcasto]

I do both an energy and a volumetric balance (ok a mass). with the two balances, you can decide if its a physical measurement issue or if it is a laboratory issue.