1955 is calling and wants it's technology back! 2

[jraef]

Believe it or not, I am being asked to defend the use of VFDs for a large pumping project where the technology being considered is actually Eddy Current Clutches! I have not run into this idea for years. They currently have them on about 1/2 of their systems, the other half are (our) VFDs. The project scope is going to involve replacing everything because of capacity, so since all of the pumps are going to get bigger, they want to go with new speed control systems that are all the same, no more mixed bag. As you might imagine, the MEs are favoring the Eddy Current Clutch concepts, the EEs are wanting the VFDs. The Eddy Current Clutch people were invited in there yesterday for making a pitch, to which the EE manager was invited. He found himself being on the defensive quite a bit, but he doesn't know Eddy Current Clutches well enough to counter some of the claims they were making, chiefly about how they are easier to fix if something breaks and how they are MORE efficient than VFDs!

Mind you I've heard this before, but it was generally being pitched to EEs who knew better. Now the problem is, the MEs are buying it hook line and sinker and they have the power at the moment because the gist of the project is all about the increased flow. As the mfr of the approved standard of VFDs for this entity, it is falling on me to help my Salesmen retain this as an electrical solution. Here's where I need a little help.

My pitch is fine with regard to MTTR etc. We happen to be the PLC vendor as well, so there is no faster fix to get up and running than swapping out our VFD talking to our PLC, because the PLC stores the entire VFD programming and will automatically reconfigure it when replaced; literally making it plug and play (they are in MCCs). But where I'm struggling is on the energy efficiency issue. I find data from the primary ECC vendors that looks suspiciously like smoke and mirrors to me regarding their efficiencies, plus they appear to be comparing themselves to VFDs using performance data from very OLD studies, maybe even based on old Six Step Inverter technology (they discuss how motor power factor decreases the throughput efficiency of the VFD...). But I also found some independent studies, i.e. Department of Energy funded, where they do not appear to have a bias but are showing that at low flows on centrifugal pumps, the throughput efficiencies of ECC system drops precipitously. Fine, I can use that, especially in that the design flow rates will be 33-69% of pump design capacity at this point; in other words it will ALWAYS be low flow. However, looking at the project description I read that the pumps are almost all Axial Flow and a few Mixed Flow pumps! What I'd like from the experiences of this group is whether I am thinking about this correctly.

I'm not an ME but as I have been told, in an Axial Flow Pump the power requirement actually INCREASES as the flow decreases. I believe in fact that I was schooled (scolded) in here at some point years ago in that the Affinity Laws that make the energy efficiency argument so compelling for VFDs do not really apply to Axial Flow pumps. If that's the case, I may be in trouble. But then again from what I know of ECC drives, they may in fact be in MORE trouble in this case because in order to vary speed, they must increase the slip across the clutch. In doing so, they INCREASE the motor-to-driven-load losses. If the pump then needs MORE power as flow decreases, they are going to have to deliver MORE power at the ery time they have MORE losses. VFD's on the other hand, while not providing the same dramatic Affinity Law savings we are used to, will at lease deliver the REQUIRED power to the motor without added losses. That to me would make a case for VFDs being MORE efficient than ECCs in this application. So even though I may have to counter the initial pitch from our salesmen on the Affinity Law savings, I'd like to be the one who tells them the REAL truth about BOTH concepts.

How does that fly with the collective wisdom of this group?

"Will work for (the memory of) salami"

 

[dpc]

I'd go with liquid rheostats and wound rotor motors myself.

So can you give us a rough idea of the hp and speed of the motors? This may be blasphemy, but do they really need adjustable speed for all of the pumps? I have done some very large pumping stations with vertical turbine pumps that used synchronous motors for most of the motors with a couple of pumps running induction motors on VFDs for maintaining wetwell level.

 

[Skogsgurra]

I fail to understand the problem.

The Eddy Current Clutch works on the slip principle. The losses are equal to (input speed - output speed)*torque. That is valid for all makes of ECCs and should be easy to understand.

Ask the mechanical guys to go all mechanical - it should be possible to arrange a clutch with a variable slip - and see what they say about that. I guess they will protest saying that losses will be too large.

Then, you have won half the battle. Just tell them that the losses in an ECC is identical and I think that you will win the war as well.

Getting involved in fuzzy discussions with fuzzy brains and fuzzy facts will never get you anywhere. Just tell them to go 100 % mechanical. They should love that - don't you think?


Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[MikeHalloran]

It sounds like that customer is in an ideal position to do a side by side comparison of what's installed already, and _measure_ the differences between different drive technologies.

Simplest test is to set up two pumps at identical operating points, and measure the heat being rejected by the ECCs vs the VFDs.





Mike Halloran
Pembroke Pines, FL, USA

 

[cswilson]

As an ME at heart (that's what my degrees are in), I'm with Skogs here. Clutches, whether mechanical or eddy current, fundamentally work by creating losses and dissipating those losses as heat. An eddy current clutch would be superior to a mechanical clutch only in that it is not wearing itself out by its action. As far as efficiency losses are concerned, they are the same. MEs should understand that.

The designers of the motors they will be using went to a great deal of trouble to build the stator frame out of lamination stacks rather than solid metal. The reason they did this was to keep the eddy current losses low. Without the stacks, the motor would be its own eddy current clutch. Ask them why they would want to undo this if they cared about efficiency...

Curt Wilson
Delta Tau Data Systems

 

[mikekilroy]

sounds like you are going into the fight without all the info you could use? can you make the MEs give you speed torque curves on all the pumps that will be used? I suspect they have this info?

then you dont have to guess about load vs speed of the actual application?

with this, could you not plot VFD actual power along the same chart by including the motor/drive losses (probably a line just 5% above the actual load)?

and the kicker then, also include the ECC actual power superimposed on same chart (probably a curved line from 5% above load at no slip max speed to 50% additional losses at the lower speeds)?

then no explanation required, just show the chart for the facts?

 

[rhatcher]

Consider this. The input torque to an eddy current clutch is the equal to the output torque. At the same time, the input power to the ECC must equal the output power minus losses. These facts are indisputable.

As an example of what his means, consider a per unit model where the motor inputs 1 unit of torque to the ECC at 1 unit of speed. This equals 1 unit (1 PU) of input power.

For the example, assume that the desired load speed is 75% of the motor operating speed. In this case, the ECC outputs 1 unit of torque at 0.75 units of speed. As a result of the speed reduction, the ECC output power is 0.75 PU. The missing power, 0.25 PU, is dissipated as heat loss in the ECC. This example obviously applies linearly for any speed reduction between the motor and load.

This model does not conflict with the fact that an ECC results in energy savings by allowing the load to run at a lower speed and, therefore, consume a lower amount of power. What this model does is highlight the fact that an ECC inherently produces losses that are proportional to the reduction in speed.

While this may seem bad enough, the situation is actually worse than this ideal model assumes. What this ideal model does not include are the additional losses, when compared to a VFD, for friction and windage when the motor operates at full speed versus a reduced speed and the additional power that is required to operate the ECC.

In comparison, it is my understanding that for a VFD the losses are relatively low and are relatively constant regardless of output voltage and frequency (motor speed in volts/hertz mode). With respect to the motor, the losses for friction are reduced proportional to speed and the windage losses, assuming a TEFC with a centifugal fan, are reduced proportional to the cube of the speed.

I think that you will find this explanation to be simple enough to register both with the MEs and the end user. However, since both MEs and end users can be stubborn, I wish you good luck!!

 

[jraef]

Thanks for the input, I may plagiarize some of the ways you folks worded some of these issues, they are better than mine.

Check this out, from one of the remaining few purveyors of ECC drives

Much of what is being said in this website is what I am hearing from the EE manager as the exact arguments made in that meeting. I especially like how the minimize the losses in their eddy current coupling as if it is insignificant, then immediately say in the next point that one reason VFDs are less efficient because they introduce eddy current losses! Wow, talk about disrespecting the intelligence of the reader...

Ironic also is the fact that this company, once owned by Eaton, was all but killed by them once VFDs became the go-to choice, but they were rescued from the business scrap heap at the last moment by a group of investors (in their "history" page). I can see someone doing that to milk an installed base for a few years with little investment in having to develop a technology, but not as a long term ownership strategy that requires growth. If the technology was so good, I doubt a reasonable company would have divested itself like that.

"Will work for (the memory of) salami"

 

[oldfieldguy]

Interested in seeing this discussion pan out. My own employer uses VFDs for motor/centrifugal compressor units less than 10,000 HP and synchronous motors and Voith variable transmissions (Vorecon) for larger installations. I've helped install several in the 15-25,000 HP range.

old field guy

 

[ScottyUK]

Jeff,

Remember Coyote / Payback from a few years ago? I'm not sure if Wile E. is still running the show but you might find some info on their website. It may have an amount of bias.

If you look back through Eng-Tips and search for keywords eddy current coupling you'll throw up a few threads from the old days. In thread237-107895 the guy whose company builds them acknowledges that they are second place to VFDs in energy savings in the final post.

 

[ozmosis]

Jeff
If I was you I'd sit with these guys, explain you are a member of E-T, as are a number of eminent engineers from around the world, and show them this thread. If they don't believe you, ask them to contribute to the thread with their argument.

 

[Marke]

Hi Jeff
The real issue here is to gain acceptance that the ECC is a constant torque device, not a constant power.
If the speed changing device was a constant power device (ie gearbox or belt drive), then as the speed is reduced, the torque is increased. Apply this to a pump as the speed is reduced, the torque from the motor is further reduced by the speed reduction ratio.
KW = S1 x T1 = S1 x T2. In the case of the ECC, the torque at the output of the motor at the motor speed is equal to the torque at the pump input at reduced speed. Power in = motor speed x torque and the pump power = pump speed x torque. Power in - pump power is loss power and that is significant. If the pump is operating at half speed, the slip losses are equal to the pump power. - 50% Loss!!
If you had 50% loss in a VFD, you would have a major problem.

I agree with Gunnar, approach it from a mechanical perspective and determine the slip losses of the coupling, then show that with a VFD, the slip losses are very small as the induction motor is operating very close to the speed of the VFD output.

It is amazing how marketing license can be used to make inferior products appear superior and educated people actually believe the BS!!

Mark Empson
Advanced Motor Control Ltd

 

[ScottyUK]

Marke,

One of the key things to consider when using an eddy current coupling to drive a fan or centrifugal pump is that they are at their most efficient when the power throughput is highest, and at their least efficient when the power throughput is lowest. Their appalling efficiency at high slip is offset by the rate at which the power demanded by the load rolls off as speed drops. If the losses are quoted in watts rather than percent the figures don't look as bad - try plotting losses v's speed as a percentage and in watts for some arbitrary fan or pump and see how different the curve shapes are.

In some respects VFDs also have efficieny problems at lower speeds because auxiliaries such as the control boards and the cooling fans are constant loads which become increasingly significant as load power reduces.

 

[dArsonval]

The following from the original post caught my interest.

"... he doesn't know Eddy Current Clutches well enough to counter
some of the claims they were making, chiefly about how they are
easier to fix if something breaks and how they are MORE efficient
than VFDs!"

I'll plead ignorant relating to comparing efficiencies of
Eddy Current Clutches verses Variable Frequency Drives.

From a repair perspective, Eddy Current Clutches do differentiate
the Men from the Boys when it comes to fixing them.

These apparatuses have multiple mechanical component fits
requiring additional jigs-and-fixtures to productively dismantle and
reassemble them during the repair process which are costly to fabricate
from scratch if a repair shop doesn't have them on hand already.

Once upon a time; when ECC's were in repair shops more frequently,
competent repair people who had years of experience working on them made
these items a productive, smooth repair.

Today, when an ECC shows up on a shop floor for repair, there can be a huge
learning curve for the younger repair technician to learn about what the item is,
and how to go about "fixing" it.

Just adding another angle to this topic, and enjoying the forum.

John

 

[djs]

I second dpc's comment to go with a mix of fixed speed and variable speed pumps. The most efficient pump is a fixed speed pump. So if the Owner is looking for the most efficient setup, make most of the pumps fixed speed and only a few variable speed. Efficiency is higher and overall cost is lower. Control is a bit more difficult, but that cost is in the noise for most projects.

 

[cswilson]

Maybe you should just ask the MEs how many of them would drive a car with the accelerator always floored, and modulating their speed by working the clutch pedal... How many of them think they would get good gas mileage that way?

 

[jraef]

LOL, great analogy!

I'm still a bit puzzled about how they make their efficiency claims. Seems a bit more voodoo magic than I would expect from an older venerable company like Dynamatic. It's almost as though they are taking a page from the playbook of these "energy saver" scams. Just say whatever you want, most people will not check anyway and for those that do, just start throwing around conspiracy theories I guess...

"Will work for (the memory of) salami"

 

[electricpete]

Good thread. I liked the thread title and the clutch analogy. (Lots of other good info too).

=====================================
(2B)+(2B)' ?

 

[Strong]

Jraef,

What is the pump HP and full speed for the application? I do not believe that an ECC is a constant torque device. Torque is reduced as speed is reduced, and power as well.

Walt

 

[LionelHutz]

Walt - I think you're confusing the torque the load requires with how the ECC transfers the motor torque to the load. The ECC most definitely follows a torque output = torque input relationship. Using constant torque to describe the ECC may be a bad choice.