Variable Freq. Drives

[rey-man]

Guys, a client bought a 3000 gpm pump without asking. now, they don't need that...can we suggest to use a VFD to automatically reduce the speed so that it would not pump out 3000 gpm but less which is required?

 

[mpoulton]

How much too big is it? Pumps are most efficient at their design point, and become very inefficient at very low flow rates. If they only need to adjust the flow by 25% or so, they could just use a gate valve in the discharge line. This has an added benefit of reducing cavitation by increasing discharge pressure (if they were having any cavitation). If they need a greater reduction in flow, a VFD might be good for them. However, if the flow is going to be reduced below half of rated output, they will be better off in the long run getting a correctly sized pump instead of paying for the inefficiency of a big pump running slow.

 

[Jraef]

I agree, somewhat.

The energy a pump uses is tied to the flow though it; reducing the flow reduces the energy. A VFD or a valve is a good choice if variable flow is beneficial. Valves, while a lot less expensive than VFDs, involve a pressure drop which represents an energy loss across it. In the long run if variability is needed, using a VFD will avoid that pressure drop and save a slight amount of energy compared to a valve. The comparable amount it saves will increase as flow decreases.

But for a lower amount of a constant flow rate, an even better solution might be trimming the impeller a bit; the method of choice if the additional capacity is not going to be needed. Trimming the impeller decreases the flow without creating a pressure drop. The downside is that if by chance you need that flow someday, you have to replace the impeller to get it. If they think they might need it someday, they can keep the original impeller as a stand-by upgrade and use a trimmed impeller now.

Talk to a good pump supplier, preferably the one that supplied that pump or at least one that has the capability (i.e. machine shop) to trim the impeller.

 

[Besoeker]

Guys, a client bought a 3000 gpm pump without asking. now, they don't need that...can we suggest to use a VFD to automatically reduce the speed so that it would not pump out 3000 gpm but less which is required?

What is the pump being used for? Does it need to be running all the time?

 

[JFletcher]

Unless this pump is positive displacement, there are a number of factors that will affect its flowrate once installed in a system. If it needs to pump a fixed volume (gpm), and it's seriously oversized for the application, a smaller pump would be a better option than a VFD.

More information is needed to determine the best course of action.

 

[hillbilly]

Do you have a copy of the pump curve?
Fluid being pumped?
Fluid temperature and viscosity?
Desired output at a given THD (head pressure)?

If so, you can post them and we'll help answer your question.

steve

 

[Cold Fusion]

Guys, a client bought a 3000 gpm pump without asking. now, they don't need that...can we suggest to use a VFD to automatically reduce the speed so that it would not pump out 3000 gpm but less which is required?

As steve said, you didn't give us much.

I'm going to assume this is not a closed piping process system. Usually those sorts of customers know what they are doing when they buy pumps.

First the easy one: If the pump is a PD -- easily

If the pump is a centrifugal, and the application is not too far from BEP, what you are looking for are called the Affinity Laws:

For small variations is impeller diameter (constant speed):

D1/D2 = Q1/Q2 = sqrt(H1)/sqrt(H2)

BPH1/BPH2 = (D1^3)/(D2^3)

For small speed variations (constant impeller diameter):

S1/S2 = Q1/Q2 = sqrt(H1)/sqrt(H2)

BPH1/BPH2 = (S1^3)/(S2^3)

where (US units):
D = impeller diameter in inches
Q = flow rate in gpm
S = speed in rpm
H - head in feet

As was mentioned earlier, a throttle valve or a valve and impeller trim works pretty good and may be a lot easier/cost effective than installing a VFD.

As was mentioned earlier, a pump curve is good to have. You want to make sure you are not getting into a range where the NPSH is to low - cavitation is tough on pumps

I'd recommend getting copies of Cameron Hydraulic Data and a Crane Technical Paper 410. Those two are a good place to start for DIY engineering.

PS: I have no problem with DIY engineering - As long as your customer is okay, this could be an interesting project.

cf

 

[Besoeker]

As was mentioned earlier, a throttle valve or a valve and impeller trim works pretty good and may be a lot easier/cost effective than installing a VFD.

[/QUOTE]
A throttle valve may well be cheaper for initial supply and installation than a VFD.
But...
the VFD would continue to save on energy costs over its life.
For the UK water industry, this has been the major driving force in the adoption of VSDs. Payback is typically 3-5 years for the VSD compared to valve control. And there are other benefits too.
But each case needs to be taken on its merits and we don't know enough about the application. If it just needs lower flow with no variation, then mpoulton's suggestion of getting a correctly rated pump would seem to be the most pragmatic.

 

[Cold Fusion]

A throttle valve may well be cheaper for initial supply and installation than a VFD.
But...
the VFD would continue to save on energy costs over its life.
---.

Well, I was thinking life cycle cost as opposed to installation cost.

Let's examine your statement from lifecycle cost standpoint.

Application is an irrigation system, 4 hours/day, 3 months a year, power is $0.1 kwh. Money costs 8%.

1. A 125hp VFD with isolation transformer is at best 90% efficient and costs $100K installed. Hydraulic efficiency drops from 72% to 68%

2. The existing pump (already paid for) with an impeller trim puts the pump within 5% of the needed capacity and head. A manual throttle valve does the rest. Trim and valve installed cost is $10K. Hydraulic efficiency drops from 72% to 65%. Additional 5% dropped across the valve.

3. A new 100hp motor with pump with new motor costs $30k. But whoops, it still isn't right - needs an impeller trim to get it right (yep, pretty common) Hydraulic efficiency is 75%.

Q: What is the payback time for options 1 or 3 over option 2.

Never for option 1??. Maybe never for option 3??

You are right we don't know the application. Adding to what I said earlier, if it is a low hour/year operation, or is the pump curve is not far from the application, the existing pump/motor may well have the lowest lifecycle cost.

You're in the vfd business, they are all cool, and I suspect you are a bit preselected toward vfds for all solutions. But they are not necessarily the most cost effective - and, I knew you knew that

cf

 

[Besoeker]

Well, I was thinking life cycle cost as opposed to installation cost.

Let's examine your statement from lifecycle cost standpoint.

Application is an irrigation system, 4 hours/day, 3 months a year, power is $0.1 kwh. Money costs 8%.

1. A 125hp VFD with isolation transformer is at best 90% efficient and costs $100K installed. Hydraulic efficiency drops from 72% to 68%

I respectfully disagree on several points.
On efficiency:
First, why assume a need for an isolation transformer?
Even if you do, it ought to have an efficiency of no worse than 98%. The VSD would typically be 97% or better. So combined 95% at worst.
We routinely have to commit to efficiency figures to within 0.1% at the tender stage. Getting them wrong would incur iniquitous financial penalties. B

On costs:
If we could get anywhere near $100k for a 100 kW VSD installed we'd be quids in. We can't. Our last installation of drive and cabling was for a 225 kW drive, installation, and cabling.At the prevailing exchange tare it would have been around $60k

You're in the vfd business, they are all cool, and I suspect you are a bit preselected toward vfds for all solutions.

Perhaps I should be. But I'm not.
Above all, integrity and honesty matters most.
If I think a VSD isn't the way to go I'd say so even if it means potential loss of business.

 

[weressl]

Well, I was thinking life cycle cost as opposed to installation cost.

Let's examine your statement from lifecycle cost standpoint.

Application is an irrigation system, 4 hours/day, 3 months a year, power is $0.1 kwh. Money costs 8%.

1. A 125hp VFD with isolation transformer is at best 90% efficient and costs $100K installed. Hydraulic efficiency drops from 72% to 68%

2. The existing pump (already paid for) with an impeller trim puts the pump within 5% of the needed capacity and head. A manual throttle valve does the rest. Trim and valve installed cost is $10K. Hydraulic efficiency drops from 72% to 65%. Additional 5% dropped across the valve.

3. A new 100hp motor with pump with new motor costs $30k. But whoops, it still isn't right - needs an impeller trim to get it right (yep, pretty common) Hydraulic efficiency is 75%.

Q: What is the payback time for options 1 or 3 over option 2.

Never for option 1??. Maybe never for option 3??

You are right we don't know the application. Adding to what I said earlier, if it is a low hour/year operation, or is the pump curve is not far from the application, the existing pump/motor may well have the lowest lifecycle cost.

You're in the vfd business, they are all cool, and I suspect you are a bit preselected toward vfds for all solutions. But they are not necessarily the most cost effective - and, I knew you knew that

cf

You should really get out more often.:smile:

• A 125HP drive with filter is $100K? Try <$10K plus installation DIFFERENCE compared to the regular starter $3K.
• Efficiency 90%? try 96% -2% for the filter IF needed.
• Throttling power loss? 30-65% depend on where you may be on the operating curve.
• The pump curve at reduced speeds will be different and so will the optimum operating point and efficiencies. Impeller trim may drive the whole efficiency in the other direction compared to reduced speeds.
• Cost of valve? (You don't even know the size and if it is a control valve it quadruples the pricing as a minimum.)
• Not only the reduced energy cost but the reduced maintenance cost and the prolonged life of the motor and pump should be taken in consideration for lifetime cost analysis.

There is a vast amount of energy to be saved in controlled fluid and gas flow applications by removing the control valves, dampers and replacing them with ASD's.

 

[topgone]

You should really get out more often.:smile:

• A 125HP drive with filter is $100K? Try <$10K plus installation DIFFERENCE compared to the regular starter $3K.
• Efficiency 90%? try 96% -2% for the filter IF needed.
• Throttling power loss? 30-65% depend on where you may be on the operating curve.
• The pump curve at reduced speeds will be different and so will the optimum operating point and efficiencies. Impeller trim may drive the whole efficiency in the other direction compared to reduced speeds.
• Cost of valve? (You don't even know the size and if it is a control valve it quadruples the pricing as a minimum.)
• Not only the reduced energy cost but the reduced maintenance cost and the prolonged life of the motor and pump should be taken in consideration for lifetime cost analysis.

There is a vast amount of energy to be saved in controlled fluid and gas flow applications by removing the control valves, dampers and replacing them with ASD's.

IMHO, it depends on the specific application. If a process is properly designed, and the pumpwork is not of the variable-flow type, you won't need VSD's (ASD's). Do you believe that VSD's (ASD's) are one-cure-for-all solutions? And didn't we hear more installations having problems with power quality, harmonics, overheating, etc. all related to power electronics? I smell snake oil!

 

[weressl]

IMHO, it depends on the specific application.

The OP supplied very limited information so the pursuing discourse presumed a common, centrifugal pump and responded so.

If a process is properly designed, and the pumpwork is not of the variable-flow type, you won't need VSD's (ASD's).

Again, the OP identified a procurment error and lack of design and asked if an ASD possibly be a solution.

Do you believe that VSD's (ASD's) are one-cure-for-all solutions?

Nothing ever is, nor did I say such. You are building a strawmen argument.

And didn't we hear more installations having problems with power quality, harmonics, overheating, etc. all related to power electronics? I smell snake oil!

Of course, we have heard of thoughtlessly and carelessly designed systems AND ASD's put in service by unqualified people and creating all sort of problems besides the ones you mentioned. Nor is the effects of ASD's are necessarily the application engineers responsibility to solve, it is the power system/distribution engineers'. The continous monitoring of the systems parameters is up to them. Usually it is not the addition of the onsey and twosey ASD's that create the problem, but when single large - relative - ASD's or the cummulative effects of ASD creep is noticed by a negative event not by being in tune what is going on with 'your' power system, are problematic.

 

[Besoeker]

There is a vast amount of energy to be saved in controlled fluid and gas flow applications by removing the control valves, dampers and replacing them with ASD's.

There is.
As just one specific example, we have a couple of variable speed fan drives in a cement works. They are fairly big - 2,300kW wound-rotor machines operating directly from the incoming 11kV supply. They are critical to the process so fixed speed bypass (a no cost option for this arrangement) with controllable dampers was seen as useful standby feature.
Unfortunately, it turned out not to be so useful. Under damper control, the extra demand exceeds the maximum demand limit. The financial penalties mean that shutting the operation down is actually cheaper.
These have now been in operation for close to twenty years and replaced a previous generation of variable speed drive. Break even on capital cost was less than five years. Had they replaced damper control, even if that was a workable alternative, breakeven would have been much sooner.
So yes, variable speed drives can yield big savings.

 

[Cold Fusion]

---I respectfully disagree on several points.
On efficiency:
First, why assume a need for an isolation transformer?
Even if you do, it ought to have an efficiency of no worse than 98%. The VSD would typically be 97% or better. So combined 95% at worst. ---

On costs:
If we could get anywhere near $100k for a 100 kW VSD installed we'd be quids in. We can't. Our last installation of drive and cabling was for a 225 kW drive, installation, and cabling.At the prevailing exchange tare it would have been around $60k ---.

And I respectfully suggest your numbers are better than mine:smile: I'm surprised my guesses were within 50%.

I think my premise that a low hour/year machine has a real hard time paying off an efficiency investment is still true.

--- Above all, integrity and honesty matters most.
If I think a VSD isn't the way to go I'd say so even if it means potential loss of business.

And I knew you thought that.

Where I am headed is typfied by wer's statement:

There is a vast amount of energy to be saved in controlled fluid and gas flow applications by removing the control valves, dampers and replacing them with ASD's.

For some it all too easy to say, "Install a VFD - there are huge cost savings from increased efficiency". Yes, I know that is not what you are saying.

My recomendation here is, "Do your home work first - don't guess. Maybe a VFD is a good answer, maybe the answer is a simple impeller trim"

cf

 

[Besoeker]

Do you believe that VSD's (ASD's) are one-cure-for-all solutions?

I don't believe anyone with experience in the field of variable speed drives would believe so. Certainly weressl didn't.
Nor do I, and pretty much all of my working life has been in this field.
From post #10:

Above all, integrity and honesty matters most.
If I think a VSD isn't the way to go I'd say so even if it means potential loss of business.

I'd rather lose a contract than lose a customer.

And didn't we hear more installations having problems with power quality, harmonics, overheating, etc. all related to power electronics? I smell snake oil!

Variable speed drives are one application of power electronics. Non-linear loads from whatever source result in harmonics.
When we bid for a new VSD project, as a rule we need to provide harmonic distortion data at that stage. In UK, G5/4 is usually taken as a maximum permitted level and contracts for VSD systems often state compliance with G5/4 as mandatory and demonstrable. Before and after measurements are par for the course.
A while back, we did a refurb on a pumping station. There were eight variable speed drives and, as contracted to. we undertook the before and after 24-hour measurements.
The before measurements already didn't comply and, of course, I gave that some thought. Distortion was mainly third harmonic which usually emanates from single-phase non-linear loads. It was in a residential area. Domestic appliances, though maybe not high power, make up for it in sheer number.
VSDs are not the only polluter.
So, no I don't smell snake oil. Maybe just a whiff of agent provocateur.

 

[Cold Fusion]

IMHO, it depends on the specific application. If a process is properly designed, and the pumpwork is not of the variable-flow type, you won't need VSD's (ASD's).---!

That's the way I see it. Pump manufacturers have been designing pumps for 150? years, and do pretty well at it.

They take their closest pump, trim the impeller to match the specified process. That's where the pump affinity laws come from - and they work well. As you alluded, sometimes the simplest, cost effective solution is mechanical - not high tech electrical.

--- And didn't we hear more installations having problems with power quality, harmonics, overheating, etc. all related to power electronics? ---

Yes. Wasn't there a thread a bit back talking about how all of the refrigeration vfd controls in a large store were generally turned off cause they couldn't get them to work right - or they would break and the repair costs weren't worth it? Makes one wonder about the life cycle costs.

--- I smell snake oil!

I don't think so. Like everything else, one must do the engineering to get a cost effective, low maintenance, system that meets the customer's spec.

cf

 

[Cold Fusion]

I don't believe anyone with experience in the field of variable speed drives would believe so. Certainly weressl didn't.
---.

I don't know about that. Wer says he can install a 125hp drive for less than $10K. :roll: That would just about make any control valve, regardless of the hours/year of use, a poor choice.

cf

 

[Cold Fusion]

---Yot only the reduced energy cost but the reduced maintenance cost and the prolonged life of the motor and pump should be taken in consideration for lifetime cost analysis.
---

VFD's reduce maintenance and prolong the life of the motor and pump?

Bes - Have you seen any evidence this is true?

cf

 

[weressl]

I don't know about that. Wer says he can install a 125hp drive for less than $10K. :roll: That would just about make any control valve, regardless of the hours/year of use, a poor choice.

cf

You're misconstruing the answer. The question was if ASD's are the cure-of-all. It wasn't restricted to variable load applications. Nor is ASD the cure for ALL of those applications either. The answer is not unequivocal EVEN if it is a new installation, not a replacement.