Pump noob needs help with pressure & flow control 3

[ditch182]

Hello all,
I've inherited a test system for cross flow filtration membranes. Basically it pumps de-ionized water from a holding tank across a membrane. The retentate & permeate both are returned to the holding tank, as we're testing the mechanical strength of the membranes & not really filtering anything. The current set up is a basic single pass system with a gate valve on the back side of the membrane to control pressure.
I've been given the job of designing a new system that will allow higher pressures as well as higher flow rates across the system. I remember Bernoulli's theorem from school, but that's about it.
My main question is: Is there a way to control both the pressure & the flow rate using a single pass system or will I need a "feed & bleed" type system? I'll need to hit various pressures & flow rates ranging from about 15 to 75 psig, and the flow rates will be variable at a given pressure as well. Any help is appreciated, or even some pointers on where to start my research. Thanks in advance.

 

[BigInch]

A flow control valve or a pressure contro valve is the same valve, just controlled by different signals. Generally you can chose one primary control variable depending on your input signal source; typically from either a pressure transducer or a flow transducer. Once the primary variable is chosen, the other variable will ride to whatever it must be to achieve the set point of the variable being controlled. It is possible to limit the other variable by setting "overrides", but that may cause you not to be able to reach the primary variable's set point. You may also be able to switch primary control variables, if you make the system capable to take signals from two different transducers.

I'd suggest you change the gate valve, as they do not offer vary good control characteristics. You will get almost no control outside a range of 0 to 10% open. Try to find a real control valve, or perhaps a nice plug valve. Check the flow versus valve position characteristics for something giving enough linearity so you can work with it.


Since you are checking strength of the membrane, I'd suggest you consider only a differential pressure control, which allows setting and holding a given pressure difference across the membrane directly.


**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[ditch182]

Thanks for the insight, BigInch. I've already figured out that the gate valve isn't very effective. I'll definitely try to find something with a nice linear response.

You said, "Once the primary variable is chosen, the other variable will ride to whatever it must be to achieve the set point of the variable being controlled." That makes sense to me, but I'm having trouble wrapping my mind around the question of whether I can control both. For instance, can I hit a pressure setpoint of 60 psig, then be able to hit two different flow set points, say 70 & 110 lpm?

I understand (or do I?) that a "feed & bleed" type system allows the system pressure to be controlled by a feeder pump & bleed valve, then the flow rate to be controlled by a re-circulating pump. Is that realistic? and is there a way to do something similar using only one pump?

Thanks again for the help, and I apologize if these are stupid questions.

 

[FMJalink]

ditch182 first analyze what you want to do. BigInch gave you a very good answer, but probably you need some more details.

A specific test membrane element will show a specific differential pressure at a specific permeate flow. This rules out the possibility to have different flow rates at a given (preset) differential pressure over the membrane element. You must therefore be talking about presetting the inlet pressure to the membrane element, but the differential pressure is determining the mechanical strength.

In order to achieve an inlet pressure at the membrane element you will require a specific inlet flow rate and a restriction (control valve) in the retentate discharge back to the tank. You could control this valve by the pressure at the inlet to the membrane element. Note that the retentate flow rate is the inlet flow rate minus the permeate flow rate.

The permeate flow can than be controlled by means of a control valve in the permeate stream back to the tank. Please note that sudden opening of this valve could be harmful to the membrane element. This valve could be used to control the permeate flow rate.

This would be a very rudimentary system, with the danger of rupturing the membrane element with a malfunctioning of the permeate discharge valve. Pending on the required flows, different membrane element sizes, frequency of testing and required automatic process control you could change the system described, e.g. allow for different inlet flow rates from the pump (variable speed drive of by-pass), control on differential pressure instead of inlet pressure, etc.

 

[BigInch]

No worries. There arn't too many stupid questions. The dumbest ones are the ones that were never asked.

No the physics of flow works for and against you.

The simple hydraulic system has a certain resistance which is dependent on flow. Likewise the flow resistance, is dependent on the system characteristics. With given system characteristics, the diameter of pipe, length of pipe, type of pipe material, fittings, membrane type, valves, pumps, fluid viscosity, if you control flow, with a control valve or varying pump speed, to a specific value, you will experience a specific resistance, ie a pressure drop, corresponding to that flowrate. If you try to control pressure drop, ie with a control valve or by variable pump speed, each pressure drop value will have a corresponding flowrate. Its somewhat like solving an X,Y algebraic equation. If you have a value for X, doing the math gets you a value for Y. With a value for Y, doing the math will yield a corresponding value for X. OK, flow is pretty much like that, but instead of doing the math, it does the physics for you. If you have Y as value for flowrate, a system will give you X, a pressure drop, a value for X, gives flowrate. If you set one, you're stuck with the other. The only thing you can do with the other is limit its range to values that give you a range of solutions for Y that you want to work with.

So far we've only talked about pressure drop. That's all flow really cares about. A certain pressure drop will give a certain flow and v/v. You will probably get nearly the same flowrate if a pump discharge pressure is 100 psi or 1000 psi, as long as the pressure drop is the same. That leaves you to determine what operating pressures you will have. You must have a maximum operating pressure at least high enough to give you the ability to reach the pressure drop you need, but it could be higher. Just that it usually doesn't need to be higher, unless you're injecting into a vessel or downstream pipeline with a certain operating pressure already. In a closed loop system, all you need to reach is a discharge pressure that allows you to have the differential pressure you need to reach the maximum flowrate you want.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[ditch182]

BigInch, FMJalink, thanks for the well thought out answers. What you've said makes sense, and fits with what I was thinking. Part of my problem is that I've been given information by people who didn't quite know what they were talking about. Physics wins out every time though doesn't it. I'm going to have to get a better idea of what my end state actually needs to be, and go from there, keeping in mind what I've learned from you guys. Thanks again, and have a happy new year.