Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
isolation valve
- Thread starter kdashc
- Start date
You should use a control valve as noticed above, in order to adjust the flow to your process requirements.
There are other strategies available.
1) Use a VSD.
2) Use “by-pass approach”. The pump runs continuously at the maximum process requirement and the excess fluid flow rate is continuously diverted and sent back to the supply.
3) Trim the impeller (when the pump has been sized wrongly, that is oversized, machining the impeller is an option to reduce energy added to the fluid).
There are other strategies available.
1) Use a VSD.
2) Use “by-pass approach”. The pump runs continuously at the maximum process requirement and the excess fluid flow rate is continuously diverted and sent back to the supply.
3) Trim the impeller (when the pump has been sized wrongly, that is oversized, machining the impeller is an option to reduce energy added to the fluid).
- Thread starter
- #5
Isolation valves are generally not designed with flow coefficients that make them particularly good at controlling flow. Consequently, you don't have very smooth or accurate control and you could easily end up with a situation where the control system is not able to find a stable operating point and bounces back and forth opening the valve too much, closing the valve too much, etc.
Other than that, you will likely cause damage to the valve thus negating it's effectiveness for isolation as mentioned by TenPenny.
Generally, just not a good idea but it would really help to know specifically what type of valve you are talking about (gate, ball, butterfly, etc.).
Other than that, you will likely cause damage to the valve thus negating it's effectiveness for isolation as mentioned by TenPenny.
Generally, just not a good idea but it would really help to know specifically what type of valve you are talking about (gate, ball, butterfly, etc.).
No. An isolation valve is either open or closed. You can't do much control with that. You should install a real control valve, or dare I say, have a look at VFDs and see if they're right for your system.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
Another way of saying what rneil and BigInch said is that the linearity of a block valve is not very good. So going from 5% open to 10% open will increase the flow 50% or so. Going from 75% to 85% will not change flow at all. Programing that kind of a valve performance curve into a PLC is pretty difficult (I've never seen anyone get it right). On the other hand, a pressure regulating valve has pretty good linearity and is self correcting.
The most energy efficient way to do pressure control is to never have increased the pressure beyond what you need. That means that a VFD or a constant pressure valve (i.e., a valve that has a pressure regulator that dumps excess flow) are both far better than throttling the full stream. Anything is better than throttling the full stream with a butterfly, gate, or ball valve.
David
The most energy efficient way to do pressure control is to never have increased the pressure beyond what you need. That means that a VFD or a constant pressure valve (i.e., a valve that has a pressure regulator that dumps excess flow) are both far better than throttling the full stream. Anything is better than throttling the full stream with a butterfly, gate, or ball valve.
David
zdas, Please strike your last paragraph and I agree. We don't know enough about this system, or the flowrate modus operandi to tell if a control valve or a VFD would be the more efficient choice, or which would lose less energy if you will. Which is why I suggested he might should have a "look" at it.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
David,
It's easy enough to program the response into a function block to linearise the response, and the electronics are plenty fast enough. Getting a realistically sized actuator with the speed and power to rapidly swing the valve through more than 3/4 of its physical travel for (e.g.) a few percent change in the upstream pressure, and also finding a valve which will put up with that kind of abuse, takes a bit more doing.![[wink]](/data/assets/smilies/wink.gif "[wink] [wink]")
----------------------------------
If we learn from our mistakes I'm getting a great education!
It's easy enough to program the response into a function block to linearise the response, and the electronics are plenty fast enough. Getting a realistically sized actuator with the speed and power to rapidly swing the valve through more than 3/4 of its physical travel for (e.g.) a few percent change in the upstream pressure, and also finding a valve which will put up with that kind of abuse, takes a bit more doing.
----------------------------------
If we learn from our mistakes I'm getting a great education!
Scotty, that just wouldn't work at all. I doubt the turbulent flow would be any help to the sensor you've sending that signal. By the time you filtered out the noise and calculated the new valve position, the vortex would be off the sensor ... Well, I think you'd do much better forgetting about all hopes of linearization and just pulse it open and closed at say what? 10 to 100 kHz, or something. 
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
I completely agree. I've seen it done and the result is a horrible control scheme which is barely stable and wears itself out in a relatively short time. I was just making the observation that it's certainly possible - almost trivial - to linearise a valve characteristic in a PLC, but it's a bad idea to try to linearise a highly non-linear valve because of the effect it has on the field equipment in terms of the large excusions on the valve which result from the linearisation.
Show me a valve which can pulse at 100kHz. I have a lot of money to make...
----------------------------------
If we learn from our mistakes I'm getting a great education!
Show me a valve which can pulse at 100kHz. I have a lot of money to make...
----------------------------------
If we learn from our mistakes I'm getting a great education!
I have one. Its an air turbine powered actuator attached to a "rotary-type gate valve", for lack of better terminology.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
BigInch,
I stand by my last paragraph. It is true that throttling a non-linear valve is less energy efficient than either a VFD or a constant pressure valve. It is not always true that the most important characteristic of a system is efficiency. There are times that the most effective control scheme is to throttle the full stream. Just because something has a better process result does not mean that it has the best energy-consumption result.
ScottyUK,
Like I said, I've seen people try to control systems with a non-linear valve and the results have been poor every time. I think that the reason is that the amount of flow that you can get through a throttled block valve is a function of some really subtle things (like what direction the valve moved last--if you are changing direction then you have to account for gear lash which changes with time) and early in the travel, very small changes in position are large changes in flow.
David
I stand by my last paragraph. It is true that throttling a non-linear valve is less energy efficient than either a VFD or a constant pressure valve. It is not always true that the most important characteristic of a system is efficiency. There are times that the most effective control scheme is to throttle the full stream. Just because something has a better process result does not mean that it has the best energy-consumption result.
ScottyUK,
Like I said, I've seen people try to control systems with a non-linear valve and the results have been poor every time. I think that the reason is that the amount of flow that you can get through a throttled block valve is a function of some really subtle things (like what direction the valve moved last--if you are changing direction then you have to account for gear lash which changes with time) and early in the travel, very small changes in position are large changes in flow.
David
zdas, you did say "energy efficiency", so I thought you really were talking about "efficiency" in the context of classical energy usage terminology, rather than this control effectivness term. In the classical context, I do can give you many examples where paragraph 2 would not hold true in terms of energy used.
Apologies.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
Apologies.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
Ya. But we know Scotty is just having some fun today.
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch
That is a good point TenPenny. When someone says "isolation valve" I hear "on/off valve". The major types of on/off valves (let's say ball valves, gate valves, plug valves, and butterfly valves) all have the characteristic that they will pass 100% of rated flow at something less than 50% of valve travel (sometimes a lot less). The control logic for any of them is similar and very difficult.
On the other hand there are people who do use globe valves for isolation. In that case, the linearity is pretty good and control schemes work well, but the act of throttling is hard on disks and seats and they often wear badly (as you said above) and don't isolate the flow when fully shut.
Bottom line is that a valve designated for "isolation" should not be used for control and a valve designated for "control" should not be relied upon for isolation.
David
On the other hand there are people who do use globe valves for isolation. In that case, the linearity is pretty good and control schemes work well, but the act of throttling is hard on disks and seats and they often wear badly (as you said above) and don't isolate the flow when fully shut.
Bottom line is that a valve designated for "isolation" should not be used for control and a valve designated for "control" should not be relied upon for isolation.
David
- Status
Similar threads
- Locked
- Question