Two motors assisting each other


Thread Starter

Arthur Tua

Hi folks,

I am involved in retrofitting some equipment which will require an upgraded pump. The current pump is driven by a 1000kW motor, while the new pump will have a requirement of 1200kW. Rather
than buying a new 1200kW motor and throwing away the 1000kW motor we are considering installing a second motor rated at around 200kW to assist the first motor. We will obviously interlink the
motor starters so that if one motor drops out the second one drops out as well. We will also look at the torque vs. speed curves of the motors as well as the pumps to make sure that we can provide
sufficient accelerating torque during start up.

Has anyone heard of this being done before?
What are the implications?
Any specific problem that we should look out for?

The existing 1000kW motor runs at 11kV. A 200kW motor is small for this voltage and therefore expensive so that we are considering using a low voltage 380VAC motor. Does this have any possible
negative implications?


Arthur Tua

Kim L. Ground

I have done this in a different application. Existing machine had a main lineshaft driven by 10Hp. Added stations made the existing motor too small, but no room for a larger motor, so added 5Hp at the other end of the shaft, with the motor starters linked as you describe in you proposal. Everything worked as expected in this application, but I think I gained some benefit from 'slop' in 2 flexible couplings and the torsional bending of the lineshaft tokiund of smooth things out. this was a low starting torque application.

Your application differs in that the motors are (apparently) fairly tightly coupled to the pump and in that you want to run the two motors (or radiaclly different sizes) on two different power supplies. My gut feeling is that if you do not have enough torque in the single larger motor to bring the load up to at least 60-70% of operating speed BY ITSELF, then you will have starting problems. I.e. the combined available starting torque from a 1000kW motor plus a 200kW motor is expected to be significantly less than for a single 1200kW motor -- in actuality the combined starting torque of the 1000+200 combination is going to be only very slightly more than for the 1000 alone. This should not be a problem in a pump application (unless it is an uncommon type of pump like a constant displacement unit).

Have you considered to use two pumps -- the existing pump and motor you have plus an additional 200kW 'booster' pump? It would be much easier to predict the response of such a system under different circumstances and would give you the ability to turn off the booster when it was not needed for maximum output.

Anthony Kerstens

I've seen this arrangement before with motors <30HP. The smaller pump was a recirculation pump feeding a 3-way valve When the larger pump needed to work, the 3-way switched and the recirc pump stuffed the larger pump.

The potential problem with this is that if the smaller pump stops, then the larger pump might get starved. With motors <30HP and centrifugal pumps, this was not much of a problem. However, at 1MW, I'm not so sure what starving the pump would do.

I would need to feel a great deal of faith in the
contactor interlocking. Using monitoring relays with positive guided contacts and redundant coil would help that. Do a risk assessment.

Anthony Kerstens P.Eng.

Johan Bengtsson P&L Automatik AB

I assume you are talking about plain squirrel-cage motors without any speed control or similar:

One thing I got to think about quite immediately is that both motors will have to have the same operating speed (of course), and then not just reffering to the same number of poles as would be obvious but to the actual speed during load. That means the slip have to be the same for full load
for both motors otherwise they won't help each other according to their size but with some other relationship and that *could* mean one of the motors take more than its share of the load.
I am not sure if the nameplate can be trusted enough for the operating speed for this issue, some measurements of actual slip might be required.

Since current drawn is directly depending on the slip in every moment and both motors always will have exaclty the same slip the load balancing will happen automatically thru the amount of the steady-state slip for the assembly. This could make the load balancing unfair to one of the motors if that motor wants a smaller slip than the other for full load.

What happens between no load and full load can be ignored since no motor will be overloaded even if the sharing is unfair.

Another question popping up (someone else have to answer it):
How if efficency changed over time? will the "full-load slip" increase as the motors get older? Obviously if this happens the motors will not change the same (Murphy) they are not even of
the same age right?

Are you sure you don't have any other use for the "old" motor? ....or can sell it to someone...

Conclusion: as I see it it could be done, but be careful about the load balancing between the motors, nameplate might not be good enough for "full load speed" due to tolerances.

/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]
While its not unusual to to have two motors driving one load, its more common in my experience to make both motors identical. This is particularly needed with induction machines so the motors will share torque as the load is applied.

In your case, I'd drive the smaller machine from a variable speed drive so that I could adjust its operating point to load it properly without having it "hog" the load. With the Variable frequency drive available it would be quite easy to make the small motor take its percentage of load by monitoring the larger motor current or Kw. and generating its speed reference signal based on that.
Responding to Arthur Tua's query... some thoughts:

1) Has the 1,000 kW motor manufacturer been contacted with regard to upgrading or uprating the existing motor?

2) It seems to me that the 11 kV rating (are you sure about voltage?) suggests "modifying" the motor's capability.

3) As a result of actual field measurements we were able to uprate the rating of a 14,000 kW motor to 16,600 kW, without modifications.

4) Contact a surplus electrical machinery vendor regarding the motor's worth.

5) Of course the pump performance must be investigated at the new operating point as well.

6) Substitute a new 1,200 kW motor having a lower voltage rating (example, 3.3 kV) and connect it to the 11 kV bus via a Captive Transformer, i.e., one whose secondary is connected only to the 1,200 kW motor.

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

Daniel Boudreault

Hi Arthur,

It is possible if you drive the pump in torque mode. Set the max torque for each motor, and limit the speed once it reaches the operating speed.

Otherwise, one motor will become a generator.

> Has anyone heard of this being done before?


> What are the implications?
> Any specific problem that we should look out for?

Matching the speeds of the motor. I am assuming the both motors will be mechanically linked to the pump with no give. If the drive is telling one motor to turn at 1750rpm, and the other drive is trying to turn the second motor at 1700rpm, then the slower motor will become a generator. But if you are running in torque mode, then if will spin the motors at pretty much the same speed. You should give yourself a margin of extra capacity, since the two motors might be doing a see-saw trying to
meet the torque needs.

I have seen one engineer put two motors together thinking that it would double the output power, but it ended up halving the output power! The second motor had become a generator. This was done using motor starters...

In the power generation arena, they use multiple generators to provide power, and they use phase control to match the output phases....but that is another story...

> The existing 1000kW motor runs at 11kV. A 200kW motor is small
> for this voltage and therefore expensive so that we are considering
> using a low voltage 380VAC motor. Does this have any possible
> negative implications?

What about copper cost? You will have use a significantly larger wire size to power the smaller motor. Also, now you have two different
voltage levels to work with...

Dan Boudreault

Thanks for the info. In my minds I may have over-simplified the matter. We were basically thinking of starting both motors simultaneously direct on
line. We were assuming that the smaller motor would assist the larger motor and both motors would settle down at the same speed providing the torque available at that speed. Bear in mind that the pump will always be coupled to both motors so that there would always be a torque require in excess of the full load torque of any of the two motors.

The existing 1000kW motor does not have a VFD or other form of speed control and this is not something that we want to put in. So that effectively we do not have any way in which we could operate in torque mode and/or limit the


I have researched the matter a bit more and I don't see how you can get into a situation where one of the motors acts as a generator. My
understanding is that for an induction machine to generate power it has to run at negative slip or just above synchronous speed - see attached torque/slip curve for induction machines. Even if the motors had no losses and there was absolutely no load the highest speed that can be achieved is synchronous speed which is zero slip. In reality both motors have losses and there will always be
some slip.

Arthur Tua
Tua Engineering Ltd.
School Street
Msida MSD04
Tel (+356) 313817, 9491286
Fax (+356) 319058
e-mail: [email protected]

Thanks for the input. Items 1,3 & 4 are particularly interesting points.

What would one gain by going to the last option that you suggest?

I need to let you have another piece of information that needs to be considered from the monetary perspective. There are 6 units and therefore 6 motors.

11kV is the correct voltage. It is a common voltage in Europe..

Another though that did cross my mind was to check the actual insulation rating of the motor windings to see if it can be operated at a higher
voltage. If the insulation rating permits, one could get 1200kW by going to 13.2kV. I dropped this idea as the transformer that feeds the motor control centre does not have the tap changer positions to go up to this voltage.


Arthur Tua
Responding to Arthur Tua's 7-Aug comments:

You have not detailed your requirement. As a result, you are getting a number of non-educated guesses as to what/will/could happen.

Assuming that you want to mechanically couple the two motors together so that they "share" the total load, then this is called a "tandem"

Several successful analogous "tandem" arrangements exist. They are:
paralleling different size generators: paralleling different size transformers; and even motor testing. However, their success is very
dependent on the similarity of operating curves - voltage regulation in the case of generators, impedance in the case of transformers, and slip
in the case of motors.

Your case is a classic tale of "the tail wagging the dog!" Subtle, but very important concerns are:
a) How do you control the smaller motor-VSD? b) Is your control capability aways tight enough to insure that both motors are supplying a proportionate share of the load?
c) What happens if either motor experiences "single-phasing."
d) You are complicating starter methodology (as well as requiring a complicated mechanical & electrical interlock) on the same process pump... one being HV (11kV) circuit breaker, vacuum bottles, or fused contactor, the other, LV (380V) breaker, contactor, or VSD.
e) What are the ramifications of the mechanical detail for coupling the three machines?
f) What is the startup sequence - when do you apply power to the smaller
g) How will power sags or interruptions affect mechanical and electrical rtransient?

Because of the great disparity in size, and the safety concerns, I strongly recommend against the tandem arrangement ! ! !

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
I apologize to Arthur Tua, but I failed to address his comment about the "motor as a generator."

Because I prefer Risk Mitigation to Accident Investigation, I tried to provide "food for thought" emphasizing negative reasons. Now,
following are two scenarios with which I hope to deter you from doing what you propose, assuming it isn't already "fait accompli."

Assume that the 380 V source of power to the small motor is suddenly interrupted. I don't mean the motor circuit, but the upstream supply. Then the motor becomes a "generator" until its
residual magnetic field decays. It might be in the order of 10s or 100s of milliseconds. During this period, the transient torque could snap the small motor's shaft.

A similar situation could occur upon loss of the source of power to the large motor. However, in this cse the motor becomes a seriously overloaded one. Once again, its shaft could snap.

I can say that I have never experienced either scenario described above. Probably bcause of the wide disparity in size. But, I have investigated situations where short transients generated by residulal magnetic fields caused great damage. One was related to a tandem-coupled turbine and motor used for a very common application... an air pre-heater for a boiler. (Details available on

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
Please excuse the pun... I don't want to appear to be beating a dead horse-power, but this is in response to Dan Boudreault's cited example of two motors whose tandem arrangement produced only
half of their expected output...

If the motors had identical characteristics, then the problem was caused by improper mechanical "alignment." I'm not talking about the mis-alignment that causes vibration. Assume that motor 1 was connected for an ABC directional rotational sequence, that is, Phases A, B, & C connected to the stator leads T1, T2, & T3,
respectively. Further assume that motor 2 connections were shifted so that Phases A, B, & C were connected to T2, T3, & T1, respectively. This would have provided the same directional
rotation, but their respective "mechanical" displacement angles would have been shifted by 120°/#poles, either leading or lagging.

Like setting ignition points in a car, the "mechanical" displacement angle for tandem-coupled motors must be aligned. It is not always
the "keyway" that is machined into each motor's shaft. This problem could have been easily fixed with a stroboscope.

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
In response to Arthur Tua's 7-Aug, 9:10pm query about the use of a captive transformer...

I understand that, although 11 kV is used in Europe, it is not a common insulation grade for motors of a 1,000 kW. The lower voltages are far
more common, and cheaper. Thus you might have had an appreciable economic incentive to purchase a lower voltage machine. The captive transformer arrangement would have allowed you to reuse the 11 kV system.

Of course, the fact that there are now 6 motors changes the situation considerably. There is an alternative. How about 6.6 kV and an auto-transformer between the motor and the 11 kV system. There are many options, but without knowing specifics, its difficult to suggest a
"good" fix.

Regarding insulation re-rating, Good approach! But, be wary of the motor's mechanical design criteria! However, I don't believe that European manufacturers would actually provide both an 11 kV and a 13 kV grade insulation system. They might from a "sales' point of view, but not from a technical one.

BTW, Arthur, one of the BITE stories I mentioned occurred in Sardinia, Italy, another in La Spezia, Italy, and another in Libya.

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
A simpler solution for Arthur Tua's dilemma...

An increase in voltage will probably affect the iron saturation. Based on the additional information you presented I believe that your most
viable option is upgrading. It can be stated that the single most influential parameter for determining a motor's performance is temperature-rise above ambient. So perform a field test.

Increase the load in say 2% increments above design, up to the motor's rated Service Factor. If there is one its probably 10 to 15 %.
Determine the temperature-rise above ambient and the corresponding increase in kW losses. Then calculate the "hot-spot" temperature for the motor's insulation class.

If exceeded, then consider an auxiliary cooling (Air, N2, even Air-H20 heat exchanger) system sized to accommodate the measured kW loss value.

If uncomfortable with this approach, ask the motor builder to perform a "back-to-back test in their factory, on like-size machines. You will
probably have to negotiate a fee. They may even provide you with a certified "uprated" spec for a fee.

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
Responding to Arthur Tua's 10-Aug 822:pm implied query... "when is a motor a generator?"

Yes, Arthur, the logic falls apart. The perception that an induction motor can only "generate" when operated hypersynchronously is dead wrong. All that is needed is a prime mover and a source of reactive Volt-Amps in the isolated system. Remember, induction motors are major contributors to the asymmetrical component of short circuits (considered as sub-synchronous faults) in industrial plants.

Rate of decay of the "trapped" flux linkages is irrelevant. It does not affect the instantaneous torque transient. The motor's electrical power
output, hence its mechanical input, is related to the "active" power demand in the suddenly disconnected LV system.

In anticipation of a followup question... a motor's contribution to a short circuit seldom jeopardizes its mechanical integrity. The reason is that the "active" component of the current contribution is nil compared to the "reactive" component.

Lastly, please reconsider your proposal to use 11 kV for the 250 kW motor. Firstly, it will negatively impact reliability. The weakest point
will be the terminal connection between the motor winding and the feeder cores. By necessity the feeder wire will be large, but the motor terminal pigtail, will be small. Secondly, and more importantly, an internal winding fault could be catastrophic (using my definition). The reason is subtle, but simple. At 250 kW, 11 kV, the current is about 17-20 Amperes. The winding wire will be so small (1-2 mm²) that it will vaporize for an internal fault, resulting in total destruction of
the motor... even explosion. There is no practical wire size that could withstand the fault current level available on the 11 kV system.

Phil Corso, PE
Trip-A-Larm Corp
(Deerfield Beach, FL)
Further to my 10-Aug 4:21pm response to Arthur Tua... Gut impression.

The scenario you described is not unusual. It analogous to excess energy recovery systems utilized in refineries, chem plants, etc.

It can be shown that reliability is inversely proportional to the number of operating components. Thus it seems to me that any system using twice the amount of "electrics" will be less reliable. I base this on the simple fact the motor's full capacity is used to drive the system until the turbine can carry its share of the load. Then it is reduced.

Furthermore, you say that efficiency, or energy, is the primary concern. Since the maximum electric contribution is 1,200 kW, then:

1) The most "reliable" system is one having just one motor rated 1,200 kW.

2) The second, but still relatively reliable system, is one where the additional short time demand is provided via a "temporary" energy
source. Investigate some way of providing the motor with a temporary increase in capacity... a heat exchanger.

3) The third, but more costly system than above, is one where the additional short time demand is provided via a "disconnectible" energy source. Since the "electrics" are available, then revisit the tandem 2x1,000 kW motor arrangement. But this time, connect the 2nd motor through a "clutch", either mechanical or electrical. It should be
considerably more economic than a VSD rated at 11 kV. A drawback will be the clutch's physical size.

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

Anthony Kerstens

Just a thought. This is all about the economics of the installation. Have you investigated the selling of the old motor to offset the purchase of a new larger motor?

Anthony Kerstens P.Eng.
Arthur, there is another concern to investigate...

Were the 11 kV motor feeders sized for:

a) Short-circuit withstand duty?

b) Earth-fault duty?

c) Motor's present capacity? Including Service Factor allowance?

d) Future motor's capacity?

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

Johan Bengtsson

Did I miss something?

I don't realize why this point would be a problem for a pair of asynchronous motors, since this should be quite self adjusting (the direction of the magnetic flux in the rotor follows the flux from the stator, it is not fixed to a specific directon compared to the rotor). If we are talking about a pair of synchronous motors it
is of course true, but I don't think that is the case here.

/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]