9 replies to this topic

[Dtan76]

Hi

Recently I faced a headache problem when I realized I do not have sufficient hydraulic / static head to overcome piping frictions loss from the liquid outlet (Test Separator) up to the inlet of Coriolis meter / Mass flow meter from Emerson/Fisher Valve.

One of the requirement of Coriolis meter is it is designed for single phase liquid and accuracy will affected if partial vapor or flashing occur. The accuracy of this type meter is 0.2% mass flow.

I'm facing problem because the liquid inlet to Coriolis meter (after a Y-STrainer) indicate vapor faction of 0.0002 or 0.64kg/hr vapor/gas generate.

Attached please find my HYSIS files, I had done the hydraulic checking using PIPE Segment in HYSIS and also I compare with may manual calc (Both are tally)

Basic data input in Calc:
a) Static Elevation head available, measured from LL of Test Separator at 0.2m to the centre line of meter is 0.758m or 0.0555 Bar Static pressure available (Using ^ P = density x 9.81 x Head in meter / 100,000 to get Bar)
Straight pipe from liquid outlet up to Coriolis meter is 5.645m
c) Total equivalent lenght of all fittings (90 Elbow, Strainer, reducer...etc) = 40m

HYSIS and hand calc / noth shown that pressure drop across the line >> Static head available

Hence, partical flashing of vapor fraction = 0.0002 is expected.

Refer to my HYSIS attachement.

My question here:
a) Any experience Instrument guy can help me on this?
Why Coriolis meter / mass meter not allow 2 phase condition? I try to ask Fisher guy but seen he unable to convince me. The pressure drop contribute by coriolis meter is about 2 PSI.

I'm pushing the client / plant owner to rellocate the meter, at least provide 2 m static head but they are so stubborn and afraid to change. (I see what can I do here.... something changing people mindset is very very difficult .....)

Hope you guy can help

Thanks

Attached Files

•  HI_HJ_HSC_Hydraulic_Study.zip   111.35KB   70 downloads
•  Overall_MEB_HI_HJ_HSC.pdf   614.39KB   74 downloads

[Dtan76]

Hi

I think I'm posting in wrong column..this thread should relocate to Profe/Ind/Simulation forum. Please help.

regards

[Dtan76]

Nirav/joerd/Ark/dijack/Joe/pleckner

Please comment and help if you can.

I need to kill this problem 2 days from now...

[JoeWong]

Danny,
Extremely busy right now. Try to drop some words. Hope it give you some lights...

Why Coriolis meter / mass meter not allow 2 phase condition?
I have "googled" you question and found plenty of stuff out there...Read these may help you

See how it works
http://www.isa.org/I...ContentID=27260

See why not in 2phase flow
http://www.isa.org/I...ContentID=41081


Success of coriolis meter in 2phase flow
http://www.isa.org/I...ContentID=52195

Personally I have seem people used coriolis meter for 2phase flow in replacing a test separator.

Presumably the selected coriolis meter does not meet the accruracy requried. How big is the error ?
You may needs to go back to find out the purpose of well testing...Do you really need high accruracy measurement to understand how the reservoir behave ? Talk to your client's reservoir engineer to see any relazation of accuracy.

I didn't go into details of you calculation, 0.055 bar available to avoid flashing seem insufficient to avoid flashing through piping, strainer, especially you fluid may contains sand, solid, salt, etc. The initial design may not be good...

Check if any chances to increase the LZLL to increase static height.

Got to rush for some meeting. Hope other Jedi can advise...

JoeWong

[djack77494]

I'd like to inject a little bit of what I hope is practicality into this discussion.

First, what do you think happens when you get a nearly imperceptible amount of flashing taking place. Do you think there's a magical instant as you go from 0.00% to 0.01% vapor where the instrument suddenly "fails". More likely, you get a very, very small amount of entrained vapor which has a negligably small impact on the instrument's output.

My second point is that I've noticed in many technical forums that people get way too hung up on simulation outputs. The simulation is just that - it's an educated "guess" as to what the "real process" could look like. Don't take yourselves too seriously. I'm sure by now I've offended many of you, but let me ask, "Do you think that your feedstock will EVER exactly match the composition, temperature, and pressure of your model?" (The correct answer is NO.) Will your columns, reactors, exchangers, etc. ever operate exactly as predicted? (Again NO.) How sure are you of the thermo model you're using? So why would you be hanging on miniscule changes in estimated parameters predicted by a mathematical model that, we hope, bears a reasonable resemblance to reality? OK, it may be the best tool we have at the moment, but we're engineers because we can apply engineering judgement, not because we can blindly run models. (Maybe I'm the one not thinking clearly with that one.) My point is that if it is that critical that vapors be avoided, then you'd best design with a bit of a margin to ensure your criteria are met; if not, then don't go into panic mode if you could detect vapor bubbles with a microscope.

One last point - I believe Coriolis meters have a fair tolerance to the presence of a vapor phase. In fact, I understand that some can even deal with two phase flow. I think that slugs of varying densities might be a problem, but not smooth multiphase flows. (However, I'm no expert in this - ask elsewhere.)
Doug.

[Art Montemayor]

All:

I'm injecting myself into this thread in response to Dtan's invitation and I first want to say: "Doug, my hat is off to you!"

"We're engineers because we can apply engineering judgement, not because we can blindly run models" is a statement that represents what I've been striving to communicate to young graduate engineers for the past 10 years or more. I believe this subject has injected itself into this thread because of a need - not because of a philosophy or an agenda.

I believe I'm one of the first and probably amongst the original admirers and pioneers of introducing and sponsoring process simulation machines since the days in 1958 when I took my first Fortran course. However, I've seen the subject of Simulation get stretched and abused to the point where it has become ridiculous. This movement has the purpose, it seems, of converting professional process engineers into semi-educated technicians who only know one thing: run a soletary computer program and distribute the output as the absolute truth without having the ability to prove if it is or it isn't. To Doug's realistic and practical arguments I would add the following:

If you use Hysys, Pro II, ChemCad, etc., etc. to design and build a $100 Million process plant and it doesn't work, who do you complain to? Or, putting the question in the real-world scenario, who takes the responsibility or the blame and is held economically liable? I don't think it would be Hysys or Pro II, et.al.! I think we all know who would be held responsible for a failure of the Simulation Program to use engineering judgment. This fact is the prime and most important reason for all Chemical Engineers to take stock and remind themselves that there is no quick, down-and-dirty, easy way to come up with a serious and complex design. A human being must make the ultimate decision, not the computer program.

If the main problem is whether the Coriolis meter will measure 2-phase flow, then it should also be noted that it will do so, but with some bit of inaccuracy. However, just how critical is the need for accuracy? Besides, there are straight tube Coriolis meters (Foxboro's for one) that offer negligible pressure drop and 2-phase capabilities. Many are the times that the inaccuracy in the Coriolis meter will probably be less that the inaccuracy in the Hysys simulation output predictions. And that is when human ingenuity will be the most important tool to apply.

Simulation programs are only as good as their algorithms - and we hardly ever get an opportunity to review and analyze the detailed algorithm involved, together with all its inaccuracies and estimates. Like Joe Wong, I would also ask "Do you really need high accruracy measurement to understand how the reservoir behaves?"

[JMW]

Entrained gas can be a problem in any sensor not designed for it and vibrating element sensors are usually vulnerable. On the other hand, quite a bit of interest in multi-phase flow has advanced the knowledge recently.
In vibrating tube systems even small bubbles can migrate to the tube walls and thus, when the tube vibrates, the gas is displaced more readily than the liquid.
Other effects can include velocity of sound errors in density measurement.
In general one might note a loss of stability and a loss of accuracy. In some cases, the instrument may "lose lock" i.e. lose the control of the feedback and have to re start the tubes.
In the Mobrey tube density meter (which was the basis of the first single straight tube coriolis meter) vibration is normally in the second order harmonic and entained air would seriously disrupt the measurement and the calibration would drift.
The entrained gas amplifier allows the tubes to vibrate in a higher order harmonic where the velocity of sound effects are significantly less and where a stable measurement results but with accuracy <1.0%. These will operate upto 100% gas.
However, in recent trials with a 12" coriolis meter by a major oil company (heavy fuel oil) entrained air resulted in mass flow errors of around 3%, which is pretty good, but what the real potential is in any specific application, I don't know.
However, using entrained gas density meters and PD meters one could do much better than this and still better if the entrainment is very small, which might be the case with flashing off by using a standard tube density meter (small amounts of entrained gas will not affect the latest generation tube density meters as much as they did earlier systems due to changes in how the tube is driven).

[Dtan76]

Hi all

Thanks for prompt reply. This few days i'm extremely overload + busy... regarding this problem, i'm going to source other coriolis meter which eliminate this 2p matter.

To doug
I'm confident with my hysis result, I always start manual calc b4 "blindly" date with commercial software, as engineer I believe analytic calc rather than software.

I use simple hydraulic calc to prove static pressure < piping drop

Another way we cld fix it is to have a dump valve function, putting a back pressure control valve may keep the flow in single phase condition.

That's all

[Dtan76]

Anyone interest how i do hydraulic calc can sent me an email then i will forward it.

regards

[T O'B]

Some Coriolis meters do better than others- testing has shown that those that operate at lower frequency, perform much better on two-phase flow (up to 5-10% gas void fraction at better than 1%, assuming the gas and liquid is reasonably well mixed). signal processing is important too.

Why is two-phase flow challenging for this technology though? Because the gas and liquid become decoupled, and the sensing tube cannot get an accurate indication of the fluid momentum.

I think there is some good information on this on Emerson's micro motion website- www.micromotion.com