Frequent Starting of Low Inertia Loads 2

[gordonl]

I have some 500 & 1000HP water pumps which I am considering changing over to an automated start stop system. These motors pump cooling water which is only required between about 1 minute every 7 minutes, to 1 minute every 18 minutes depending on the process.

The motors start in about a tenth of the allowable locked rotor time. The motors are NEMA design, 5800 series westinghouse frames.

My thought to date is to allow a max 6 starts an hour with the motor running for the required 1 minute if the process is at a slow rate, and the motor running 8 miunutes on 6 minutes off, if the process is paced at 7 minutes. Operation should be considered as 24/7.

I'm already looking at the life of the starters, but I'm not sure what sort of life expactancy I can hope for from the motors starting them this frequently.

 

[electricpete]

Even with the low inertia, this seems way too much for that size motor.

I'll look to see what the limits were on our comparable size motors. By the way, what is the speed? One thing is that the off-time required for the motor to cool between two starts increases significantly if the motor is stopped (vs running) for the majority of time between those starts.

 

[electricpete]

For medium size motors,
NEMA MG-10 Table 2-3 gives parameters labeled "A", "B", "C", as follows:
A - Max # of starts per hour, regardless of inertia WK^2
B - Max Product of [starts-per-hour] times [inertia=WK^2 in lb-ft^2].
C - Minimum time between starts (in seconds) to allow motor to cool sufficiently to allow another safe start.
Items A and B are used to attempt to limit long-term degradation due to cuumaltive damage from repeated starts with adequate cooling between starts, and item C ensures the motor is cooled sufficiently for another start (prevent short-term abuse which will cause immediate damage)
Here are some representative values:
_______2-pole_________4-pole________6-pole
_______A____B____C___A____B___C___A___B___C
100HP_2.6_92___220_5.2_441_110_5.9_1181_97
250HP_1.8_210_1000_3.7_1017_500_4.2_2744_440
As expected, these parameters are more limiting as inertia increases, full-speed increases, and horsepower increases.
An example for 250HP (186KW) 3590rpm induction motor we limit the number of starts per hour to 1.8 or or 210/WK^2. whichever is less. If inertia WK^2 > 210/1.8=117 lb-ft^2, then the second number will be more limiting. If inertia were less, then we would use 1.8 starts per hour (long-term average). You could certainly have more than 1.8 starts in one hour as long as you limit the time between starts to 1000sec ~ 17 minutes. The 1.8 starts per hour (or 210/wk^2) is based on limiting long-term cuumulative damage due to an average number of starts per hour applied continuously 365 days per year, 24 hours per day. The 1000 sec between starts is intended to allow the motor to cool and prevent short-term abuse.

The number of expected starts in a motor lifetime decreases dramatically if you don’t allow the cooling between starts. For motor operated with proper cooling between starts we will expect tens of thousands of starts before failure. If motor is not allowed to cool sufficiently between starts, the total number of starts before failure reduced dramatically. EPRI produced a report which estimated a specific motor lifetime in starts as a function of initial rotor temperature (prior to the start). For every 40C increase in initial temperature there was a 10-fold decrease in expected lifetime (# starts).

The table stops at 250hp. Perhaps this is a recognition that motor designs get more individualized for higher motors. For 250hp 2-pole motor they are expecting 1000 seconds off minimum between starts. Larger 2-pole motors would require even more off-time. Your 6 minutes=360 seconds between starts sounds insufficient. It is not clear to me whether the NEMA table assumes motor running or secured between starts.

 

[electricpete]

I would be surprised if you could reach 6 months lifetime using this scheme. Just my guess for what it's worth.

 

[electricpete]

As I'm sure you know, vfd would of course be an option. Either to make the starts less severe, or to eliminate the starts.

 

[GusD]

Hi gordonl

Electricpete already covered the subject quite well.I would like to add that the starting scheme is extremely severe for motors this size.I can't rmemeber any similar applications where motors were subject to such a control.
Of course the VFD would manage this quite well,but so would closing discharge valves and reduce flows.(I assume they are velocity pumps).
If the flow can be reduced,why not just use either smaller motors or redude the number of motors.
Gordonl,for all I know,this system has been operating for a long time and has beed controlled manually.
If that is the case,than you should have a good history on the Reliability of the motors.If they have operated well on Manual control,switching to an Automated control should not change running times or # of starts.
I have a feeling I missed something in the message.


kind of control


GusD

 

[jbartos]

Suggestion: Normally, projects include design criteria, e.g. visit

http://www.energy.ca.gov/sitingcases/magnolia/documents/applicants_files/afc_cd-rom/14_AppendixE.pdf

etc. for more info
It may be a good idea to check the electrical design criteria how they were done in the project engineering and design phases. Considering the size of the motor, 1000HP, there should have been some documentation in the project file that addressed the motor starting frequency.
Incidentally, a motor may be designed for very heavy duty service. E.g. consider plugging, jogging, etc. besides frequent starting. The frequent starting causes the motor to overheat. All what is needed is to engineer and design a motor that can normally withstand such heavy service duty.

 

[motorman]

I agree with Pete. Motors of this size range would normally be limited to a max. of 8 starts total per DAY. The design limit or first weakness encountered will be in the rotor with the rotor bars failing first. But as jbartos suggest, since motors of this size are not "stock" designs but custom designs to match precise applications, they may have been originally designed for such hard starting duty, such as what is done on jet pump motors. The motor designer would be faced with coming up with a motor that has a whole lot of thermal capability meaning more materials thus larger size costing much more money in terms of dollars per horsepower.

This problem is not all that uncommon even with new equipment when say the pump OEM did not pass on to the motor manufacture the need for multiple starting. Some possible modifications that could be done to the existing motors to achieve more starts is to allow the motors to run at no load between the starts, allow them to coast or add things like force air cooling that would come on when the motor is idle and/or coasting. This forced cooling would more rapidly remove the heat generated because of starting. I have even seen air conditioners added to machines for this purpose.

 

[Marke]

The problem is essentially a thermal one. You must be careful that you do not overheat the motor by doing to many starts. If the motor is rated for a particular starting time, then you can increase the start frequency as you reduce the start time, however you must look also at the ability of the motor to dissipate the heat that is being generated during the starts. If the running time after each start is very short, then the effective cooling of the motor will be severely compromised resulting in excessive temperature rise. The cooling of the motor is significantly less when it is not running and so you may need to allow over five times as much OFF time to dissipate the heat, as running time. The actual ratio depends on the particular motor design.
If in doubt, ask the manufacturers.
I would normally suggest that if the total starting time, total running time and total OFF time per hour is equal to that recommended by the manufacturer, then increasing the frequency of starts will not create problems. If however the Total OFF time is increased, the the start frequency would need to be reduced.

Best regards,


Mark Empson

http://www.lmphotonics.com

 

[jbartos]

Suggestion: In some cases, the motor may be started more frequently, if lightly loaded. However, this would require careful experimentation and manufacturer tech support. Also, the shaft load would have to be added over some clutch. If these expenses are compared to the more expensive motor built for heavy-duty service, the more expensive heavy-duty service motor may become the solution.

 

[tmahan]

This raise some additional questions to my mind as to how engineers (Users and designers) come up with "rules."

The real issue is heat inside the motor. Rather than rely on a one size fits all standard why not use RTD or PTC to monitor the motor temperature directly, and then tie this value back into the starting logic?

Motorman cites the 8 starts per day (I do agree that is a pretty fair guideline). I can easily come with realistic situations where that would either be excessive and result in potential damage or overly conservative and overly limit the motors usefulness.

Temperature measurements inside the motor is cheap (relatve to the motor) and accurate. It should be used more.....
Or am I missing something??

 

[Marke]

Hello tmahan

Yes, thermal sensors do add additional protection, however the major problem area during start is the heat dissipated in the rotor rather than the stator. Stator heat is a problem, but rotor heat is a greater one and is difficult to protect with sensors.
Best regards,

Mark Empson

http://www.lmphotonics.com

 

[jbartos]

Suggestion: The rotor heat is not a problem if one has a motor with superconductive winding. Also, if the shaft of the rotor is cooled by a suitable coolant.

 

[Marke]

If you had a rotor with a superconductive winding, it would have zero resistance and it would not dissipate any heat dureing start. The problem is, that it would never start as you need resistance in order to develop torque under high slip conditions.!

Regards

Mark Empson

http://www.lmphotonics.com

 

[electricpete]

Good point Mark. Starting torque is proportional to rotor resistance. Maybe need a double-cage winding with part-superconducting and part resistive. Sorry Gordon, just couldn't resist throwing in my two cents however non-practical.

 

[jbartos]

Comment on Marke's posting. The superconducting material does not necessarily mean the zero resistance. It just means very small resistance. Therefore, the motor will start. Also, notice that a small induction motor has several times higher rotor resistance than a large induction motor. Both motor types start o.k.

 

[Marke]

Hello jbartos

I think we've been down this one before, but whether the bars are made of superconducting material or not, is not really relavent to the question during start. Once the machine is up and running, there will be an advantage because if the resistance is very low, then the slip losses will be reduced. During start however, we have established in a previous discussion that the energy dissipated in the rotor is at least the full speed kinetic energy of the driven load. I say at least, because as we also established previously, if the acelleration torque is restricted, then additional dissipation will occur in the rotor.
So, assuming low work torque during start, the energy in the rotor is directly proportional to the load inertia. This is why some motors are rated with the maximum inertia of load that they can handle.
If the inertia (or start time) of the machine is significantly less than the rating of the motor, then there will be less energy dissipated in the rotor, allowing for more starts provided that the running time is on a par with that used for the motor ratings.
Best regards,

Mark Empson

http://www.lmphotonics.com

 

[gordonl]

The manufacturer has said that the motor can be started upto 10 times in an hour (20 second thermal limit/2 second starting time), but the frequent starting will quickly kill the motor, their answer without pulling design data. (Exceed lifetime starts) (I'm now asking them for a proposal to perform an engineering study to have them model the motors and provide actual starting ability, upgrades, new design.)

Unfortunately the motor manufacturer is trying to sell me soft starters as well,so I'm leary of taking their word for it.

Can I actually expect to get longer life out of the motors if the thermal loading of the starts is the same but the starting current is lower? Stated over, will the reduction of starting current from 6X to 3X increase motor life by reducing motor/rotor torque? (Our machines are generally copper rotor bars, and these probably are as well)

 

[electricpete]

The question that I think you are asking is whether soft start makes the start easier on the motor. This one has been hashed many times before on this board and I'm not sure if there is simple answer.

Here is a link which provides some math discussion to get it started:

http://www.geocities.com/pschimpf/RotorHeatingSymbols.htm

The reduction in voltage has two effects:
#1 - It increases the total heating over the course of the start. The total rotor heating is equal to kinetic energy times a weighted average value of Te/(Te-Tm) over the course of the start. That weighted average does not play much role if Tm<<Te but will play an increasingly larger role as voltage is reduced and Te decreases by V^2.
#2 - It increases the time over which that total energy is produced.

Item #1 makes the start worse, item #2 makes the start better. I think that Gus mentioned that for VERY high inertia loads where start is prolonged over many minutes, item#2 can be important. I think in most other cases #2 is not important. Likewise if voltage is reduced by a small amount and Te still >>Tm than item 1 has little effect. Keeping discharge valve of radial flow centrifugal pump throttled back as much as possible (without endangering pump) reduces item #1 by reducing load torque.

 

[gordonl]

Putting the thermal justification aside, which I beleive I can do for now, I'm interested in the mechanical aspects of the motor life.

Will the reduced torque on the rotor bars, caused by the decreased staring current of the soft starter, give an increased motor life when compared to an equal amount of DOL starts?

The manufacturer is suggesting the mechanical life of the motor is in question, and that the thermal heating is not an issue because of the quick starting time.