High pressure line, low pressure analyser bypass?

[jmw]

I just had this one floated by me.
We need to put an analyser in a bypass but we have a maximum pressure limit on the sensor below the maximum pipeline pressure when the high pressure pump is running.
I'm thinking that a downstream pressure reducing regulator would drop the pressure to the analyser. But I now have to do something with the sample stream and all I can think of is to put in a booster pump downstream of the analyser to bring the pressure up enough to return the flow to the pipeline.

I am concerned that the regulator may introduce a lot of pressure drop which may give me some gas problems but I then flow into a circulation pump. Then the analyser and then the booster pump.
Trouble is, both pumps would have to be PD and I now have another problem which is matching their flowrates, unless I use a pressure relief bypass on one to spill excess flow back to the inlet. I'd hate for this to be the low flow rate and have the booster pump cavitating.

Pipeline pressure can go to 2000psi. The analyser has a 1400psi limit.

But is this feasible?
How do others deal with this problem?
What do I have to look out for?

JMW

http://www.ViscoAnalyser.com

 

[BigInch]

Could depend on what the product is, but that aside,
Does it really need to be reinjected? Can it be stored for disposal, or return later to point of origin. Is it a discrete sample, or continuous?
Can you route the analyzer discharge back to main pump suction somewhere.
What sample volumes, or flowrate are we talking about?
Or could be that just a small chemical injector pump is needed.


What would you be doing, if you knew that you could not fail?

 

[jmw]

Hi Biginch,
I already considered a collection tank and all that it entails.
It is certainly a solution but I don't yet know how remote this is from the pipeline pumps or any storage facility.
It will be my luck that this is a remote monitoring point and they won't want to have regular road tankers emptying the collection tank.
This is a continuous sample flow of around 20 l/min. This flow is important for a number of reasons.
I have queried if they are close to the pumps as a return to the pump inlet was one of the first things I thought of, the collection tank second.
I then queried if the high pressure were a default condition in which case I'd just automatically isolate the sample loop and go without measurement but they have two pumps one of which is responsible for the high pressures and when it runs they do need the measurement.
It's crude oil by the way, API 15 to API 18.
It's not the Canadian oil fields because there they have a density and viscosity limit and cut the crude with distillate before putting it in the pipeline and I don't have these problems plus they usually are at a pump station.


JMW

http://www.ViscoAnalyser.com

 

[BigInch]

Do you think the oscillomatic pump might do it? Ask them about the gravity/viscosity thing. Also be careful with the possible temperatures that they might be pumping that crude. It is possible that even with a diluent it must be heated to move it down a long pipeline, so it could turn into a brick in a cold temperature shutdown... maybe.

What would you be doing, if you knew that you could not fail?

 

[MortenA]

Maybe you should look for chemical injection pumps? Some of them have quite high head.

Best regards

Morten

 

[BigInch]

Didn't I say that already.

What would you be doing, if you knew that you could not fail?

 

[jmw]

So basically, there is nothing wrong with my proposal except to consider the pump types and how to match flows between them?

JMW

http://www.ViscoAnalyser.com

 

[BigInch]

What you don't match, you'll have to store temporarily.
Be sure that the pump can handle the worst viscosity.

What would you be doing, if you knew that you could not fail?

 

[jmw]

I am not happy with the idea of relying on there being sufficient pressure in the suction line to the booster pump - I am concerned about cavitation and/or bubbles in the sensor. Hence I want to be sure I can kill them off/ minimise using a circulation pump.
Not sure I can rely on the pressure downstream of the pressure reducing valve as at low pressures I will be entirely dependent on the booster pump suction.
The lowest pressure seems to be around 100psig and the temperatures range from 130[°]F to 160[°]F.
I was thinking of marginally oversizing the circulation pump and setting the PRV to spill back excess flow.
I was also wondering if a centrifugal pump would do what I want as the flow rate will be controlled by the booster pump.



JMW

http://www.ViscoAnalyser.com

 

[BigInch]

Can you vent off the bubbles in some kind of small separator. If it is as critical as you think, I'd be hesitent to rely on a mixing pump.

Come to think of it, if continuous flow, storage isn't going to work well either. Better match the pumps if you wind up with two.

What would you be doing, if you knew that you could not fail?