Arthur, I have one last (?) thought...

Do I understand you correctly? Is the goal of the "retrofit exercise" to reconfigure the SWRO plant so that the six 1,000 kW machinery trains will be reduced to just three 1,200 kW ones?

If so, then how can you justify relocation, modification, and the operational complication associated with the tandem installation of three existing 1,000 kW motors, vs replacement of three with new 1,200 kW motors?

A second thought! Perhaps this is a job for the HV-LV xfmr/LV motor/VSD arrangement. Definitely expensive!

Either way, this is a perfect case for the application of the forgotten discipline... "Value Engineering!"

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Clearing up some misconceptions of my own:

1) Wed, 09-Aug-11:38 am, response to Dan Boudreault - I mixed USA and European practice. The mechanical shift equation should have been,
(120°/# polepairs) or (240/#poles).

2) Thu, 10-Aug-4:21 pm, response to Arthur Tua - the phrase in parenthesis should have read, (they are operating subsynchronously).

3) Same communique - The last sentence, "There is no practical...", referred to the winding wire size, not the feeder core size.

Information Requested
a) For Arthur, typically, motors of this size and/or voltage class are furnished with winding temperature TC's or RTD's. Are these motors so
equipped?

b) Thu, 03-Aug-3:20 pm, reply from Kim L. Ground - can you provide prorata share (in percent) for each motor contribution to the shared load? Were they connected to the same power source?

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Responding to Johan Bengtsson's 10-Aug, 8:24am comments....

Yes, you did miss something. There is no "self adjusting" feature when the two rotors are locked at some arbitrary angular displacement of their shafts.

An aside.
Referring to Donald Pittendrigh's 14-Aug, 2:56 pm observation regarding two motors (out of 4) "hogging" the load.

Question to Donald: When you say the fluid couplings are fully engaged do you mean "locked-up?" If so, and if the 4 drives have identical characteristics, and they are fed from the same
power source, then the overload of two could be caused by the phenomena I addressed earlier, ie, their respective rotors are improperly aligned
when locked-up! Thus there is "give" as was the case that Kim Ground described. There is no "slip", excuse the pun, between drives. The
problem is exacerbated if the motor speeds vary between sub- and hyper-synchronous speed.

Returning to Johan's comment. If you have ever miswired a multiple-speed motor, you will know what I am talking about. The motor will sound like
it is "growling!"

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Phil,

As this list is concerned with automation & control and as discussions tend to have an electrical engineering perspective, I did not go into the details behind the retrofit exercise.

The current facility consists of 6 units, each containing a pump, motor (1000kW) & hydraulic energy recovery turbine as well as the associated
reverse osmosis membranes. The motors are sized to provide the power required by the pump less the power provided by the turbine.

We are evaluating the retrofit of a high efficiency work exchanger, a positive displacement energy recovery device, which transfers the pressure in the reject stream to part of feed stream. The unit being considered is
called a DWEER (short for dual work exchanger energy recovery) and its use completely chnages the process set up from the pumping and energy recovery perspective. For full details visit http://www.desalco.bm/d-html/papers.htm
and have a look at the paper covering Work Exchangers

To utilize as much of the existing equipment as possible we are combining two sets of membranes, which will now only require one of the exisiting
pumps. This time however, the motor driving the pump will not be assisted by the hydraulic turbine and needs to supply the total pump power requirement, hence the need to go to a larger motor.

From 6 units requiring around 900kW, we are going to three units requiring around 1200kW. The energy savings are quite extraordinary. Even with this order of magnitude for energy savings, investment costs are important and the payback period has to be appreciable less that the remaining term of the contract.

Hope the above paints a better picture of what we are evaluating.

Again, thanks for your assistance and ideas on the matter.

Reagrds

Arthur Tua
Tua Engineering Ltd.

 

I actually don't understand this and am very interested in understanding what the problem might be here. And I don't have any motors to test it with either, that would have been interesting...

The normal squirrel cage asynchronous motor is working by a set of rotating magnetic field from the three phases feeding the motor. In the simplest case, a two pole motor you will have one magnetic field going more or less straight thru the rotor and rotating one full turn each period of the AC source. This will make the magnetic field cross the windings in the rotor and induce a current in them resulting in another magnetic field. This magnetic field will follow the field from the stator off by some angle depending on the load (no load = no angle, breakdown torque = approx 90 degrees for a two pole motor, approx 45 degrees for a 4 pole motor and so on). The magnetic field in the rotor does not follow the rotor, but the magnetic field from the stator.
To put it in another way: If you would measure the magnetic flux at some point in the rotor it would not be constant, it would change with a frequency equal to the slip.

Now, there is nothing, as far as I can see it, saying that the rotor should be turned in one exact angle given the angle of the three phases for an asynchronous motor, this would mean that
there would be no direct conflict between the motors (except maybe during startup, but I don't think even then).

The above does not apply to a synchronous motor where the magnetic field(s) of the rotor is fixed to a specific direction compared to the rotor.

Could you direct me to at what stage the error is located?

/Johan Bengtsson

----------------------------------------
P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]
Internet: http://www.pol.se/
----------------------------------------

 

Responding to Johan's Mon, 28 Aug, 2:02 pm comment... I will proceed on the basis that my Thu, 24 Aug, 4:42 pm explanation, "Motors in Tandem" is still unclear.

Consider the equivalent circuit for motor "A". A voltage, Es, is impressed on the stator's phase "A". A voltage, Er, is produced to drive the rotor circuit. It is equal to the stator input voltage, Es, times a very complex impedance, Ze, formed by the stator circuit, magnetizing circuit, and rotor circuit impedances, Zs, Zm, and Zr, respectively.

Er then produces a rotor current, Ir, equal to EsxZe/Zr. Then, Ir^2xZr produces electrical losses as well as torque. Furthermore, Ira lags Es
by the "load angle", alpha.

Consider a second motor "B". If identical to the first, then its rotor current, too, has the same "load angle". And, alpha lags its own stator
voltage Es. Now physically link the motors, either by coupling their shafts to the same load or jack shaft (or any non-compliant mechanical
transmission system), or by connecting one motor's shaft to the second motor's extended shaft.

Remember, both torques are being produced at the same "load-angle", alpha. And both are referenced to "A" motor's Es. However, if, motor "B" is inadvertently connected to the power supply so that its Es "lags" or is displaced with respect to "A" motor's Es by 120 degrees, then the rotor current introduced in "B" motor's rotor produces a torque that is 120 degrees of phase with respect to the rotor torque produced by motor
"A". Thus, while motor "A" produces 1.0 per-unit of torque, motor "B" produces a negative 0.5 per-unit. The net torque is 0.5 per-unit.

Anyone interested in how I learned about this phenomena?

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Phil,

If you treat torque as a vector, it acts at mutual right angles to the force lever and the force applied - so along the axis of the two motors in this case. Yes, if the two motors are identical and R phase of the supply is connected to X winding in motor 1 and Y winding in motor 2 ( phase rotation XYZ being correct in both) the magnetic flux vectors will be displaced relative to the phase centres by 120 ?. But if the motors are not identical, how does this relate to the
orientation of the windings relative to each other in space? As far as I know (and I'm not a motor construction expert) there is no requirement that winding X is always top dead centre.

I'm interested in your story - aren't we always? - but don't think you can justify your argument by reference to torque vector effects.

Cheers,

Bruce.

 

Responding to Bruce's Tue, 29 Aug, 3:37 pm comment... Different size motors.

The mathematics are similar if their speed-torque curves are identical, ie, full loads occur at the same slip. The combination can then be modeled by equating the larger motor's electrical characteristics in per-unit of the smaller one.

Before I tell you my story, I would like to ask two questions (starting requirement aside)... "How many applications have you seen where
ordinary induction motors were tandem-connected to the same load?" Also, "How do you explain the strange behavior cited by others on the list?"

I have been involved with bascule bridge drives where two motors were used, one on each side of the bridge, but they were slip-ring type. Dual-lane, leaf-span bridges always have separate drives, and where they move together then the drives were slip-ring. Also look at tandem-drives
associated with crane travel. They are done with DC, slip-ring, or VFD. Never with ordinary squirrel cage motors.

And now, the rest of the story:
After engineering school I started my career with US Gypsum, inventor of "Sheetrock" or wallbord. A slurry of gypsum was poured, formed, and air-hardened, on a 1,000 foot long belt-conveyor. It was driven by a jack-shaft, powered by one 60 Hp motor. Other machines were connected to the same jack-shaft for sychronizing other operations. Variable speed was provided by an eddy-current clutch between the motor and the jack-shaft.

Ten days after I bgan work, the drive motor failed. Repairs would require four weeks. However, I noticed that there was a fixed motor
mounting base installed at both ends of the shaft. The "Chief" electrician said that they were for the original slip-ring drives replaced with the motor/clutch system, about a year earlier. Of course the old motors were discarded. Being an "injuneer" I suggested to management that we install two 30-Hp motors. I was suddenly thrust into the limelight. I became an instant hero.

I thought nothing of the sly smile on the Chief's face, as he nodded to the electricians to give me all the help I needed. When the motors were
hooked up, and rotation tested, I accepted all of the plaudits. Well of course my elation disappeared when both motors were started. Much to my chagrin, it didn't work. The motor I was closest to "growled", its ammeter showed over-current, and fortunately, its breaker tripped.

After countless hours of checking, the Chief called me aside. He said, "Son, disconnect and then reconnect the coupling, so that one-half lags
(or leads) the other. The coupling was a 6-bolt unit, each separated by 60 degrees. Well the first attempt didn't work. However, the second
attempt did. Currents were normal, and equal.

The lesson... listen to the "old" Chief. He may not have had book learning, but he certainly was "motor-smart!"

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Phil,

Interesting story - no, I haven't had anything to do with tandem motors. I still can't see how the exact orientation of the rotors relative to the
stators would produce that effect, though. I guess one possibility is that the torque produced is not steady, but has an alternating component due to the finite number of rotor bars - if the
two alternating torque components reinforced, fivne: if they opposed, not fine. But I'm speculating.

Did BOTH the motors in the gypsum case overload, or only one?

I love it when the real world refutes the theory we have all been taught - but it's not so good when my students find out and ask for explanations!

Cheers,

Bruce

 

Responding to Bruce Durdle's Thu, 31 Aug, 7:40 am, comments.... overloads?

Yes, both motors showed different overload magnitudes. The mathematical model shows that stator current is sensitive to actual loading.

Perhaps the following will clear up some of the murky waters. Generally, windings are not randomly installed in the stator slots of
assembly-line produced motors. Thus Motor "A" and Motor "B" will have their Phase R start in, for example, the 3-o'clock slot. Thus both motors will produce torque at the same mechanical angle
relative to the stator.

However, position of the key-way is not apt to be located in the same position for each motor. That is because the installation of the shaft into the rotor is more likely to be a press-fit, w/o regard to a relationship between the two parts.

Perhaps the phenomenon is better visualised if both stators "influenced" the same rotor, as might be found in a dual-speed motor.

Anticipating another question:

Q) How about the torque relationships of a double-cage motor (NEMA D)?

A) One cage is designed for starting, where s=1. The other cage is designed for the normal operating slip range!

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Exactly my point(s)

Since more than one rotor bar is "active" at the same time, those for the moment closest to the poles (of the rotor) are those carrying most current, but even those at some distance helps in making the magnetic field in the rotor, and they do help each other. There is an alternating component when the current passes from one bar to the next, but that oen is not changing the direction of the torque, only the amount of it.

Besides, some motors have the bars somewhat twisted to make this transistion smother, it is done to make the torque more constant and to bring down the sound emitted from th motor.

I hope someone can explain this otherwise I can't teach people the right thing....


/Johan Bengtsson

----------------------------------------
P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]
Internet: http://www.pol.se/
----------------------------------------

 

Responding to Johan's Tue, 5-Sep, 3:35pm comments... I appreciate your dilemma with your students. Therefore as an aid to understanding the
phenomena, try this:

Consider two stators driving a common rotor. Stator 'a' develops a rotor emf, Ea, producing rotor current, Ia. It lags the rotor emf by some
angle. Simultaneously, stator 'b' also develops its own rotor emf. Eb, producing rotor current, Ib. Both Ia, and Ib, produce rotor torque. As
long as both stator inputs are "in-phase" then their torques are in phase.

An analogous situation would be connecting the secondaries of two transformers together, but their primaries are supplied from out-of-phase sources. The resultant secondary voltage is a vector addition to which a common load is connected. The same is true of the motor case. The two generated emfs, hence rotor currents (actually flux fields) are also added vectorially.

Keep it simple. Ignore "real motor" parameters such as magnetizing current, rotor bar and stator slot skew, stator to rotor slot ratios, etc. Just model two stators, each having input voltages that are out-of-phase.

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

Ok, a common rotor with two stators would behave like this since you would need to have the same current in both ends of the same bar the magnetic field inside the rotor have to be aligned along the entire motor (ie both stators).

So, the next step of the trouble:
why, when the rotor is split in two, would the magnetic fields still be aligned? the only connection they have is that their axis are tied together. Even if the axis is able to carry some magnetic field it don't think it would carry enough to actually align the two rotors magnetic fields to each other.

Could there be some issue with some leaking magnetic fields? It should probably be possible to measure the magnetic fields generated inside a motor just outside it (not enough magnetic shield to prevent it). Could this be corrected by simply putting the motors further apart and/or putting a magnetic shield between them? (Not that connecting the motors in another way wouldn't be an easier solution, just to complete the theory of why)

Could someone come with comments if this sounds like the right track of the problem or not.

/Johan Bengtsson


----------------------------------------
P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]
Internet: http://www.pol.se/
----------------------------------------

 

Responding to Johan's Sun, 10-Sep, 3:10 pm comments... we're getting closer:

Now reconnnect B stator so that it lags (or leads) A stator by 120 deg..

The mathematical model I used is the same as that used for the "hot" terminal problem. However, for this application let Ea represent motor A's phase-to-neutral source, Eb represent motor B's phase-to-neutral voltage source, and "short" out Ec. Substitute the stator impedance, Zs, for the original motor impedance Zm, and substitute the rotor impedance, Zr, for the original fault impedance, Zf.

Both rotors are in parallel, therefore their impedance is one-half of the original. First calculate the rotor emf, E2a, and the resultant
rotor current, I2a, for the A motor. Then do the same for the B motor. Remember that it is out of phase with respect to A motor. Vectorally add the two rotor currents I2a and I2b. It will become obvious, that the resultant rotor current will be significantly reduced while the two stator currents will be substantially increased.

Regards,
Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)

 

What?

Ok, I agree that if the currents in the two rotors are forced to be the same, you get this. Still: what is forcing the currents of the two individual rotors to be the same (ie the same vector)? As far as I get it the two rotors currents will be out of phase by the same amount as their stators, I still don't see any problem with that.

Could you repost your mathematical model, my memory of it might be slipping somewhat and I can't find it?


/Johan Bengtsson

----------------------------------------
P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]
Internet: http://www.pol.se/
----------------------------------------

 

Recently I had such requirement in my application and simple galvanical separator inserted into loop solved the problem.
Regards
Piotr Kowalski

 

try action instruments or m-systems

 

Just wire the two devices into the loop in series. As long as you power supply can drive it you can put as many devices in the loop as you want.

 

To accomplish what you have requested, Moore Industries has their ECT-DIN Signal convertor with two 4-20 ma outputs. I have used an earlier model of this device with great success. See
http://www.mooreindustries.com/products/sg_signal.shtml .

An less expensive alternative to splitting the signal, especially if one signal is for control and the other for data acquisition, is to insert a
loop isolator into the 4-20ma control circuit. Moore has their models SIX and ECT.

Best of luck on your sdearch.