LCV - choked flow - how can it control level? 1

[Radionise2]

Hello guys,

Imagine a horizontal 3-phase separator for an upstream offshore service, i.e. inlet flow will vary from time to time. And there is an LCV on the produced water outlet stream.

It just occured to me that if the pressure differential across the LCV is high, for example about 70 bar, hence the flow is choked, how can the LCV control the water level in the separator?

Correct me if I'm wrong but I guess we have to allow for flow variation across the valve in order to control the level.

Please advice.

And thanks in advance!

 

[hanon]

Level control is fundamental. In my experience, bad level control is the main source inestabilities and global oscillation of a plant. A level bad tuned can make the plant to oscillate entirely. It should be tuned mayority as proporcional PID with huge integral time to avoid cycling. In your case I don´t understand what you mean. you should perhaps allow flow variation in the HC level but the water level must be tightly controlled. But I still don´t understand choked flow? I suppose that you mean that sometimes the water LCV is closed completely. This may happen if your loop is cycling as consecuence of the integral time . Try to tune it mainly proporcional and this will prevent you from closing the LCV

 

[Latexman]

You are correct.

To continuously control the level in a steady-state manner it is best to have the ability for the output_water to equal or exceed the largest expected instantaneous input_water, so the accumulation term in a mass balance goes to zero. If not, then you have to have the ability for the average output_water to exceed the average input_water, but this gets a little dicey and complicates your control. If these are not possible, you simply cannot control the water level.



Good luck,
Latexman

 

[mbeychok]

Radionize:

First, we should understand that there are literally thousands of three-phase (gas, hydrocarbon liquid, water) separators in operation the world over that use interface level controllers to control the level of the interface between the hydrocarbon liquid and the water ... and they all work.

The choked flow of a liquid (water in this case) is a much different phenomena than the choked flow of a gas. See

http://en.wikipedia.org/wiki/Choked_flow

and, in my experience, does not occur very often. I have never heard of it affecting the ability of an interface level controller on a three-phase separator to work properly. Perhaps others can cite a specific example of that occurring, but I cannot.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[dcasto]

If the water volume swings greatly, just install a larger LCV or two smaller ones. If you really have huge volumes that could carry over, install a high level control system that closes (throttles) the gas outlet until the water is dumped.

 

[Montemayor]

Radionise:

The following follows the way I interpret your two questions.

First, which “inlet flow rate will vary from time to time...” – the inflow to the Liquid Control Valve? I presume so, if the inlet to the production separator is also varying.

Secondly, engineers normally are only interested and deal in gaseous fluids achieving sonic velocity. I see no practical reason or purpose to impose sonic velocity on a liquid fluid inside a pipe (assuming you could design for, and handle, a sonic-level liquid hammer in your piping). Therefore, the only way that you are going to tolerate sonic velocity (“choked” flow) inside your drain pipe is for the drained liquid to “vaporize” – actually, an isenthalpic flash – across the drain valve (LCV). This may be true in your case, but it depends on the liquid fluid we are talking about. Production separators can handle a variety of hydrocarbon liquids as well as water – it depends on your production characteristics. Only you can tell us that.

Based on a material balance, the liquid fluid that enters the separator must exit through the LCV. I believe that is a basic premise for your continuous operation. There is NO liquid accumulated in the separator. In my opinion, the vapor (or gas) flow at choked condition CAN EXIST at the LCV outlet – if you want it that way or if there is enough gaseous volumetric flowrate through a fixed, small pipe diameter. Normally, this is not what you want – and, therefore not what happens. I can safely control the separator’s liquid level by expanding the liquid content down to whatever pressure I need to drain down to and handle it accordingly - without causing choked, gas flow. If I find that my LCV will be actually flashing, I need to take this into consideration immediately at its sizing and design stage. If this is true you will find that you probably will also inherit other problems: the isenthalpic expansion will cause cooling – perhaps down to the sub-zero range - due to the hydrocarbon refrigerants being flashed. This is exactly the same principle and technique employed in mechanical refrigeration. The expansion (“refrigeration”) valve is nothing more than a fancy LCV acting in exactly the same manner that you describe – except that it’s signal is reversed: it opens more on liquid level depletion in the evaporator. For many years, before the advent of turboexpanders, that’s the way we produced liquid hydrocarbons at low temperature in order to subsequently apply de-ethanizers, de-propanizers, de-butanizers, etc., etc.

Like Milton, I also have never heard of choked liquid flow affecting the ability of an interface level controller on a three-phase (or 2-phase) separator and not allowing it to work properly. In fact, I dare say that anyone attempting to design for choked liquid flow is begging for trouble in the form of a sonic-speed water bullet trying to make a 90^o degree turn at the first elbow it meets. I wouldn't give the elbow much chance to continue existing were that the case. So why tolerate/design for sonic liquid flow?

I've handled/designed dozens of these same applications and never have come up with what is being questioned or posed. Neither have I heard of anyone else coming up with the same. Perhaps I don't understand the questions correctly.

 

[mbeychok]

Radionize2:

I think that, for the sake of clarity, it should also be understood that such three-phase separators almost always have two level controllers.

One controls the upper level of the hydrocarbon liquid phase, and that LCV is in the hydrocarbon liquid outlet line.

The other controls the interface level between the bottom of the hydrocarbon liquid layer and top of the water layer, and that LCV is in the water outlet line.

I assumed that you were speaking of the interface level controller in the water outlet line.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[Radionise2]

First of all thanks for all your reply. The nature of projects that my current company are taking on are 99.99% brownfield engineering for matured oil and gas assets. This is my first time dealing with liquid control valve as most of the time the type of valves that I've dealt with are for gas services throughout my early career since nearly 6 months now.

Secondly, sorry for the lack of clarity about the challenge that I'm currently facing. At the moment and what's actually happening is that we have an existing separator that is going to be operated at higher operating pressure (say about 70 bar) for a higher capacity commitment to receive fluid from one additional field, and at free-flow, i.e. by-passing compression and directly into the discharge manifold together with the rest of the gas and condensate from the other separatorS and down to the onshore gas terminal. The intention would be to get the water to flow out with the condensate stream on the 3-phase separator (and just close the outlet stream of the produced water). However, there would probably be some sand in the incoming fluid albeit a sand screen has been installed at some point further upstream of the process. So the sand may affect the condensate liquid fiscal metering skid or even causing serious erosion problem. So I'm now thinking if we can still get the gas and condensate flowing at high pressure, free-flowing into export while STiLL, routing the water stream down into the desanding unit to remove the sand. the sanding unit pressure is normally approximately 1/7 the pressure of the separator. Just to cut things short, the downstream units such as the liquid surge drum before the pumps and also the desander have to be operated at it's current low pressure because they are also receiving condensate from several other separators operated at lower pressures and for other fields. And because of all these, with just this one separator operating at high pressure, the the LCV of the water stream would have to absorb about 7 times the pressure of the liquid surge drum and/or desander. My initial calculation shows that at such pressure difference, the liquid flow would be choked.

Anyway, with regards to my original question about understanding LCV, I managed to get hold of my principal yesterday to get Initially I thought that an LCV at choke flow would have a fix flow rate out of it, and I was wrong. I forgot that the opening of the LCV, although at choke condition can still vary by controlling it's % opening and hence the flow rate can still increaes or decreases, although the flow is choked.

Sorry for the confusion and my incompetence. I'm trying hard to make myself into a useful engineer here.

 

[Montemayor]

This thread is now getting more confusing. My comments are given relative to the quoted segments of the last post:

“what's actually happening is that we have an existing separator that is going to be operated at higher operating pressure (say about 70 bar) for a higher capacity commitment to receive fluid from one additional field, and at free-flow, i.e. by-passing compression and directly into the discharge manifold together with the rest of the gas and condensate from the other separators and down to the onshore gas terminal.” --- The 3-phase separator is there to separate the phases. There can be no compression of 3-phase well head fluid; therefore, there can be no by-passing of compression – just straight flow into the 3-phase separator for phase separation. I seriously doubt that both gas and condensate are being sent to a gas terminal. The gas should already be separated from the associated condensate before reaching the gas terminal; – otherwise, why employ a phase separator?

“The intention would be to get the water to flow out with the condensate stream on the 3-phase separator (and just close the outlet stream of the produced water). However, there would probably be some sand in the incoming fluid albeit a sand screen has been installed at some point further upstream of the process. So the sand may affect the condensate liquid fiscal metering skid or even causing serious erosion problem.” ---This is just backwards of what can actually happen. The condensate is always the TOP layer on a 3-phase separator. The water is the BOTTOM layer in the same separator. Therefore if the condensate-to-water ratio is low, it is possible that the liquid level sustained in the 3-phase separator would essentially be mainly water (with the condensate being swept out together with the drained liquid through the condensate drain valve) if the bottom water drain valve were plugged or blinded. This action will convert the 3-phase separator into a sand trap and sand accumulator. The sand normally gravitates down with the heavier water and this leaves the lighter condensate relatively sand-free. If the 3-phase separator has reasonable residence time and little or no turbulence (which is normal), then any included sand will establish a permanent residence within the lower, water layer – up against the usual water weir. There should be no sand getting downstream and causing problems, unless the 3-phase separator is grossly undersized.


“So I'm now thinking if we can still get the gas and condensate flowing at high pressure, free-flowing into export while STILL, routing the water stream down into the de-sanding unit to remove the sand. the sanding unit pressure is normally approximately 1/7 the pressure of the separator.” ---This is very confusing. No separator – whether 2- or 3-phase is capable of separating a mixture of 2 liquid phases with a gas phase into a gas+one liquid phase product AND a liquid product. A separator simply separates a mixture into ONE gas phase and ONE liquid phase. How did the gas and the condensate get preferentially mixed after going through the 3-phase separator?


“My initial calculation shows that at such pressure difference, the liquid flow would be choked.” --- Please submit these calculations that demonstrate that the liquid fluid is traveling at sonic velocity. Also show the references where the relationships originated.

“I forgot that the opening of the LCV, although at choke condition can still vary by controlling it's % opening and hence the flow rate can still increase or decreases, although the flow is choked.” --- Again, please show how it can be calculated or proven that the liquid is flowing at sonic velocity (a.ka., “choked flow”).

 

[dcasto]

Great job Mortemayor in trying to sort the facts. I've run a "threephase" separator where we allowed the HC liquids to carry over with the gas, its not common. Can't wait for the "rest of the story".