What are the best/better test methods for determining oil cut? These wells have high water cut

[PressEquip]

Hello,

I'm not a petroleum engineer, please be understanding of my limit knowledge

My desire is to determine the best test method for measuring oil cut to an accuracy precision of ±0.1% volume. Test needs to be judged from economic perspective too. Needs to be practical and economical.

Situation: 20 to 30 wells flow to a battery. At the battery there is water knock-out, settling tank, treater and other equipment. Most of the wells have high water cut, as high as 99% water by volume, some even higher. Skimming operation.

Current tank test is reasonably accurate but it quite involved.

Typical graduated cylinder test (1L sample) is not accurate enough (~±0.4% to 0.5% precision accuracy). Maybe there is more precise glass ware available, any comments are appreciated.

Industrial flow meters (multi-fluid), are expensive, need to be calibrated and when less than ~5% oil cut typically don't have good enough accuracy (maybe ±1% precision accuracy)

Do you know of standard ASTM, API, AGA or other standards that provide a standard test method for oil/water cut?

What test method might be better or the best?

Thanks in advance for your response

Kind Regards,

PressEquip

A Fan of EngTips, share the knowledge

 

[1503-44]

You have some oil in your water. Look for method to test for oil in water. Testing at hi tolerance <1% will be expensive and/or time consuming.

Why do you need to do this testing? The reason may determine which test method you should use.

https://www.astm.org/Standards/water-testing-standards.html

https://www.law.cornell.edu/cfr/text/46/162.050-27

https://www.spectrosci.com/resource.../whitepapers/guide-to-measuring-oil-in-water/

 

[LittleInch]

Measuring anything to +/- 0.1% isn't easy or cheap, that's 1 part in a thousand.

Do you mean 0.1% of the overall liquid? Or something else?

At 99% water cut as you say this is a skimming operation. So 10 m3 or 10,000 litres will give you 100 litres. Are you trying to see if this 99.9 litres or 100.1 litres or 90 or 110 in 10,000?

Anything better than 1% of the overall volume is really a laboratory job or you need to separate a lot of oily water to measure it.

I would go for the latter.

Fill up a large tank / known volume ( at least 10 m3) and then let the oil and water separate and skim off the oil and then measure it.

Then decide to shut the well in for a while.

At those sorts of volumes you could probably sling it or a side sample stream through a cyclone and measure the total fluid in versus oil out.

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

 

[1503-44]

Does not the cost of disposing the produced water exceed the value of the oil?
For that matter the cost of testing may exceed the value of the oil.

 

[PressEquip]

Hello,

Thank you all for your responses to date.

The main reason for testing is to determine which wells are still economically viable to keep pumping. The company has strategically position itself to be a low cost producer in a certain region.

Yes, the aim is to measure 0.1% by volume of the over-all liquid. The liquid is initially an oil and water mixture. The aim is to separate out the oil and measure the amount of oil volume to determine the percent volume of oil in the original liquid.

Ideally the sample volume is something a person could physically handle. Ideally 1 liter but practically I'm hoping a larger sample volume of 5 US gallon pail (18.927 liters) would be sufficient to use large lab glass ware to obtain a measurement to 0.1% by volume (18.927 mL).

I find LittleInch's idea interesting: At those sorts of volumes you could probably sling it or a side sample stream through a cyclone and measure the total fluid in versus oil out.

The water is disposed of through an injection well near the battery and is not overly burdensome. Since every month there are several hundred tests required, I'm hoping there are economies of scale which allow the test costs to be reduced so the testing isn't prohibitively expensive.

Thanks again for all responses!



A Fan of EngTips, share the knowledge

 

[LittleInch]

Well if you got a 1m tall glass tube, you're talking about measuring +/- 1mm oil to water.

That's just not practical when that interface is likely to be rather indistinct and also it will take a long time to separate.

At that level of water cut you've got to be looking at measuring the amount of oil in much larger quantities to be able to measure the oil to some relevant accuracy or just constantly measuring the total fluids versus the oil out. Then you can get one meter of one size and another meter 1/100 the the size.

Hence the cyclone idea. Run it for an hour and measure total flow in versus oil out. That might work.

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

 

[1503-44]

I'm trying to figure out the 1% recovery economics, but just don't see it working out. No pumping costs anywhere? No trucking costs? No gold? How does it work?

 

[waross]

Compare the volume of recovered oil with the total volume processed.
If you can't recover it, don't measure it.
In addition, the oil recovered determines much of the economics of the operation.
I don't think that your accuracy needs to be much better than your recovery efficiency.
That is if you are measuring to 0.1% accuracy but you can only recover 90% of the oil at those low levels, then there may not be a need for such accuracy in the tests.

Bill
--------------------
"Why not the best?"
Jimmy Carter