Flow control centrifugal pump with frequency drive or control valve? 9

[CARF]

Dear all,

I want to control to a constant fluid flow of 20 ton/hr with a centrifugal pump motor attached to Danfoss frequency converter. My collegue says this is NOT a good way of controlling flow and this should ALWAYS be done with a control valve.

We have a magnetic flowmeter to measure the flow.

Who of us is wrong? Anybody experience with both methods?

Many thanks for helping and sharing knowledge,
MVD

 

[TD2K]

There's no reason you can't use a variable frequency drive rather than a control valve. Both essentially do the same thing. A control valve 'matches' the pump curve to the system resistance curve by taking a pressure drop across it (the difference between the pressure you need to get the fluid through your system and what the pump produces for that flow).

A VFD changes the pump speed to match the pump curve to the system resistance curve. A VFD avoids the pressure loss across the control valve and the associated energy that represents. On the other hand, a VFD is more expensive, has its own electrical inefficiencies and has potential impacts on your electrical distribution system due to harmonics depending on the size of the unit.

 

[d23]

MVD:

TD2K is right (again.) The way I read your post you already own the VFD. If that’s the case then your answer is simple. You own it use it!

I am not a fan of VFD's. If you use a VFD the pump can operate over a wide flow range and stay very near its BEP. If you operate a pump at a comparatively low speed it would be reasonable to expect a longer bearing life. Bearings are rated in terms of total revolutions. The slower the pump runs the longer it takes to make this total number of revolutions.

Being honest if I had a blank piece of paper to start with and could design what I wanted it would be the control valve. The main reason is that everyone understands valves and restrictions. With a VFD anytime you have a problem with it you will need a highly specialized person to repair it. The valve will cost less to repair than the VFD too.

Good Luck!

 

[MortenA]

I know that many non european engineers (especially US engineers) dont care much about energy savings - but its a fact that 10-12% of the worlds total energy concumption are used by pumps and that up to 50% of this could be saved by the use of VLT (not deducting for the additional energy required to make the VLT).

VLT are more often favourable in Europe due to high energy taxes.

Best Regards

Morten

 

[CARF]

Hi all,

Yes we already have the (Danfoss) VLT. Exellent thing; worked in one go and gives information about Amperes and Torque as well. It also has got a PID controller inside.

I guess we are saving energy by avoiding pressure loss over a control valve. Another thing is that the VLT actually gives the upportunity to overrun the pump motor to higher speeds without any problems!

Important for us as well is that avoiding a control valve will give us a cleaner process (food process).

So yes we go for the VLT.

Q. What is BEP?

Thanks for thinking with me,
MVD

 

[25362]

BEP: Best Efficiency Point.

 

[MortenA]

Best Efficiency Point?

Just guessing

Best regards

Morten

 

[tonyh]

For good use of VFD you need to consider the system curve. Ideal systems are those where you wish to pump against vee squared pressure drop down a line. Then as you change the flow the pressure required changes nicely. Be cautious in using VFDs where almost all the head for the pump is static head as a small change in speed will make a big change in flow. In these systems you need to have a reasonably steep pump curve or otherwise you are trying to control lots of flow over a very fine speed range.

 

[ipsys]

Tonyh is right regarding high static heads and VFD. You can check the pump's operating point and efficiency at different speeds by plotting pump curves at different speeds (using affinity rules) together with the system characteristic. Or use software to create the plot.

 

[CB2]

MVD
Generally your application will work well. I have done similar applications for feeding DI filters, primarily for energy savings.

 

[ccfowler]

MortenA,

I must take exception to your comment, "many non european engineers (especially US engineers) dont care much about energy savings." This is an overly broad statement, at best. Energy savings are valuable in the USA as well as elsewhere in the world for both direct and indirect benefits.

In most discussions of variable frequency drive applications, one can readily perceive that there is a significant amount of misunderstanding about the complex interaction between an adjustable speed drive, the pump, and the associated piping and equipment system. All too often, it is somewhat casually presumed that spectacular energy savings can be realized in all applications. This is quite simply not correct. In some applications, the savings at part-load operation can be very great, but in other applications (especially retrofit ones), the energy savings can be very modest.

Each application must be properly evaluated on its own merits. Merely saving a relatively trivial amount of energy while expending substantial physical and economic resources does not represent sound, responsible engineering on any continent.

Local taxation and regulatory policies may cause some differences in engineering criteria, and this can be expected to result in some differing regional trends. Such apparent variations should not be casually presumed to reflect a significant difference in the professionalism and ethics on the part of engineers practicing in different regions.

 

[MortenA]

ccfowler

Energy saving is naturally only a part of a technical solution. I would not recommend a certain solution if i though it was uneconomical.

Generalising across an entire continen as i did will ofcourse cause many people to be wronged. I apologise for this.

Its just that since enrgy prices is so much higer in europe compared to the US - the incitament to save is higher in Europe.

You actually also said: "You own it so use it." That seems like a fairly econmocally sound advice to me. However your preference to valves in favor of VLT (due to suspected higher maintenance costs) might be valid in the US but not in Europe. Since i dont know where MVD lives your advise may be the best.

Dont get me wrong: Im not claiming that VLTs will save the world or that they are suited for all aplications. I do however think that conservative approach to engineering often stop them for beeing used where they would be suited.

The DANFOSS units are usually quite compact and easy to use. They are relatively cheap and selfcontained (at least the small units). The technology has progress a long way since the first VLTs. with modern communication protcols etc.


I will however still claim that even though the high energy prices in Europe (due to taxes) might actually harm the contries in term of total progress i am however certain that it does promote the development of less energy consuming equipment (just look at cars).

 

[CARF]

Goeiedag (good day),

First of all thank you for thinking with me and I hope my question gives good information and food for thought to other engineers throughout the world as well.

I live in the Netherlands and the application is for a plant in sunny Greece; both Europe.

I'm certainly no expert, so here my simple learnings / understanding. Flow control with a control valve seems like running a car at full speed and adjusting to the desired velocity using the brakes ; ).

The VFDs we use are cheap, almost comparable prices to a good control valve. VFDs do not require any maintenance, control valves do because they foul and wear out and loose settings.

An advantage of a valve is that a control valve may give faster control (dynamics) and (looking at static head) better controllability.

The Danfoss VLT 5000 we use is plug and play, very easy to use and with many features, like current and torque information. Good stuff.

Greetings from Holland, thank you all for this discussion, thank you eng-tips,
MVD

PS. I don't have shares in Danfoss ; )

 

[dadfap]

MVD

That is the greatest explanation of using a control valve I have ever seen.

WELL DONE!

all the best

dadfap

 

[PUMPDESIGNER]

Something interesting here.
I once picked on a US Government engineer for statements I thought were wrong about VSDs.
But he explained that I totally missed his point. The Department of Energy was attempting to get plants to use constant speed pumps finely tuned to operate at just the correct point by trimming impellers, changing gear or belt ratios, piping, etc., BEFORE they even considered VSDs. VSDs were not considered good options if simpler more economical methods could be found first.
I agree. Problem is that engineers are asked ahead of time to specify and design without sufficient information, so flexibility and margin for error and unknown is a necessary part of the design.
Then that same engineer replied that is why the DOE was pursuing plant operators and not design engineers, because plant operators can play, and they can more closely calculate the actual true operating conditions.

PUMPDESIGNER

 

[MortenA]

If you know your exacts operating point and its constant then why both with an expensive piece of hardware that increases energy consumption?

You are totally right of course pumpdesigner. But in some case you know from the onset that you wont have a steady flow.

I resently bought a progressive cavity pump for a produced water drum. The expected flowrate of produced water is very uncertain and will change. In this case a VLT was a good choise. Also a valve does not work so well with progressive cavity pumps.

Best regards

Morten

 

[CARF]

Hi,

Of course I was talking about AUTOMATIC flow control (with a flowmeter and a PI-controller) all the time. This is a totally different game then tuning a pump system to a 'fixed' operation level.

Despite various distubances (as changing visosity, heat exchanger fouling, pump wear out and changing tank levels) this automatic control loop will keep the flowrate to a constant level by automatically adjusting the pump speed.

Automatic control has the advantage to go to different operation levels (changing the controller set-point / desired flow value) as well.

You gave me some good ideas on for other applications.

Regards,
MVD

 

[jmw]

The assumption here is that you are using a centrifugal pump. Here VFD has benefits. For example, noise reduction. Running the pump at its rated speed and throttling back using a control valve you will find a noise increase as you close the valve.
However, running the pump at constant speed will not guarantee constant flow. With a centrifugal pump factors such as changes in back pressure or viscosity will affect flowrate. SO where you have a flow meter, you have flow rate measurement and can set up a PID loop to maintain the flow rate constant. You can use a control valve or you can use the VFD. i.e. change the back pressure or change the pump speed.
I am very wary of people (especially me!) who make absolute statements. "ALways" needs to be hedged. You can't simply use a control valve to control a PD pump unless you install it in a feed back loop i.e. feed some of the pump output back to the inlet. Here the control valve will affect the flow rate but who wants this added engineering complication?
If you have a PD pump then pump speed controls the flow rate and the VFD is a very neat solution to flow rate control. (I just commissioned a system using an Alweiller PD pump with Siemens Micromaster VFD control and found it works a treat, once you pay the extra $20 or so for a basic program panel... why aren't they supplied as standard?) and an especially nice way to "guarantee" the flow rate i needed without a flow meter, something i couldn't do using a control valve in a feedback loop. VFDs can let you program all sorts of parameters such as ramp up and down so you can soft start and soft stop the system.
Just because control valves are understood is not by itself sufficient justification for continuing their use.
Since much of service and maintenance is subcontracted out these days it little matters what on site skills are available, what matters is the reliability of the system chosen. Should we expect more problems with electronics or with mechanical /electrical/pneumatic devices? The first batch-controllers i worked with used post office uni-selectors and relays. People love relays because they can inspect and clean the contacts, i was told. Now we all accept microprocessor based control without a second thought.
Technical competance isn't the issue for the user it once was, it is for the supplier and his service support. These guys should also be able to advise reliability figures for their equipment.

 

[MortenA]

Allweiler and Seepex are ofcourse competitors. How did you find the service of allweiler (for my own future reference)? I was very happy with the Seepex performance - allthough they have given me some grief re. final documentation.

Best regards

Morten

 

[jmw]

Morten,
Can I fairly answer your question?
Probably not. I don’t have a history of consistent use of one manufacturer to be able to say if they always perform one way or another. Because I design systems for use globally, I have used different manufacturers in different installations. Usually I choose the equipment type and then choose the manufacturer from the clients approved vendor list so that I don’t introduce “aliens” into their service and maintenance programs. All the pumps and VFDs, I have experience of, seem to have been trouble free especially once installed and commissioned. The problems I have had have been only with this last system and it would be unfair to infer either a distributor or manufacturer as being the source of the problem nor that it is typical.
This latest process viscosity measurement system was for a pipeline blending application for heavy fuel oil custom designed for a refinery in Siberia.
I use PD pumps and VFD control…my choice and for reasons similar to those given by MVD.
I don’t consider the energy costs as the main concern is serviceability. It is important to design systems with minimal downtime, minimal call for special skills and to cause minimum disruption to the end users existing service capability.
The point is that any energy savings might be more than consumed in down time and lost production if things go wrong. If things do go wrong then specialist support can come from the manufacturer or his distributor.
In this case equipment manufacturer choice was based on availability with appropriate approvals for import into Russia.
Fabricated in the UK, but to get all the necessary GOST hazardous area approvals and Russian language manuals, plus the GOST meterological certificate, the Russian agent had to organise all the key equipment item procurement in Moscow. This meant dealing with the Moscow distributors for suitable equipment.
Hence an Alweiller pump (with Siemens motor), with Siemens for the Micromaster VFD.
So much for choice.
Next, problem solving: –31C temperatures [A pipe side installation and for some reason, whether the Alberta oilfields or Siberian refineries (Southern Siberia, fortunately), I always seem to be supervising commissioning these systems in winter!] doesn’t make for an un-biased or balanced view I’m afraid.
Yes, I did have problems (I don’t usually when commissioning); none with the pump and some with the Micromaster, none with the Fisher Valves used for the heat exchanger services.
This means we had to get telephone support from the UK, Moscow, and Germany, all in different time zones and this is not a fair guide to the sort of support you might normally expect. It was less than I hoped for and probably better than I had a realistic right to expect.
The Alweiller pump works fine supplied with a Siemens motor and the Siemens Micromaster VFD, very well in fact.
Q: Would I use them again?
A: No reason not to. If I had found the same problems over and over again with any one supplier I could give you a better answer.
Q: Would I advise the client against any of the vendors specified for this system?
A: No. All work well now they are commissioned.
Q: Would I use VFD again?
A: absolutely.
My only recommendation is that if you choose Micromaster VFD, at the minimum, to get the BOP (Basic Operator Panel), a key-pad module that plugs into the micromaster. The price was £15. Why wasn’t this standard? Why wasn’t it offered? Is this usually not offered or was it just this distributor?
When doing the FATs , downloaded the Micromaster software from the web site, intending to use MODBUS to configure, it proved to be a huge program and we found had downloaded the wrong program. Why are/were they different for different sizes of Micromaster?
Then we couldn’t get the new program to operate over the old.
During commissioning when we returned to this as a programing method we and we found the programs no longer available for download! Was it something we did wrong? Don’t know. By that time we had a BOP sent down from Moscow so we commissioned with this keypad device and left it in place. This is a problem only if you want to be able to control the VFD remotely by comms and only if they don’t have a clean solution and I don’t know that they don’t. We had a problem and we moved to a different solution rather than try to solve the problem.
Sorry for the long answer.