Motor Starting Power & Power Factor 1

[X49]

I am sizing a gas genset to start a 600V 45kW motor driving a centrifugal pump. Motor characteristics are as follows:

Starting Current: 405A
Rated Current: 54A
Total Moment of Inertia: 0.42 kgm2
Speed: 1775rpm
Load at pump operating point: 32kW

Motor starting power is the limiting factor for the genset in this application. My genset sizing software assumes a starting power factor of 0.36 and a running load (input) of 51.7kW, giving a starting load of 87.3kW. This assumes a "low inertia" load and NEMA Code H motor.

Is a starting power factor of 0.36 and starting load of 87.3kW reasonable for this application? I've heard that the rule of thumb for induction motors is a starting power factor of 0.2 (which would give a starting current lower than the running current in this case).

I may be over-analyzing things here but I would really like to know the relationship between starting power, starting power factor, and load characteristics.

 

[dpc]

I think you have a misconception somewhere. The starting current is still going to be 405 A (at full voltage) regardless of what assumptions you make about power factor. The starting power factor has nothing to do with with running load kW.

0.36 sounds a little high, but the difference between assuming 0.2 and 0.36 should not be too significant.

You need to pick a NEMA starting code in the softwware that matches the actual starting current data for your motor.

David Castor

http://www.cvoes.com

 

[electricpete]

I don't do genset sizing, but I wouldn't have guessed starting power would be limiting a limiting factor.

At any rate, I have a spreadsheet that will build a best-fit model of the motor which you can use to predict (guess) specific performance parameters. The more data is available to use for estimating model parameters, the better model can be built. Do you have a data sheet for the motor. Or do you know
* starting torque?
* breakdown torque
* speed at which breakdown torque occurs.
full load speed (exact, not synch)
Full load efficiency and power factor.
Half load efficiency and power factor.

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

 

[X49]

dpc, I'm not interested in what the starting current will be, since this only affects the generator windings and not the engine, which will be the limiting factor. It is the real power the motor draws during starting that requires work from the engine, not the apparent power.

electricpete, I know the speed, efficiency, and power factor at full and half load, but torque characteristics are not given. They are:

Full load speed: 1775rpm
100% Efficiency: 92.0%
75% Efficiency: 92.5%
50% Efficiency: 91.5%
100% pf: 0.87
75% pf: 0.83
50% pf: 0.73

I would be interested in seeing your model.

 

[electricpete]

ok, I'll take a quick look. Without locked rotor torque it won't be great for estimating starting.

Note I had voiced the same question as dpc. What aspect of the prime mover are you considering is limited by starting power? If it is thermal, remember this only occurs a brief time. If it is torque, I'd think breakdown torque would be more important.

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

 

[dpc]

It doesn't take much power to accelerate an unloaded motor up to full speed. The limiting factor with a gen set is usually the voltage drop that occurs during starting, hence the normal interest in starting current.

Why do you think the engine is the limiting factor?



David Castor

http://www.cvoes.com

 

[electricpete]

I agree with dpc.

fwiw I did a quick fit of the model to the data. Results as follows:

Induction Motor Equivalent Circuit Fit Results

===============Model Parameters Solution================
Name Value Units Description
R_NL 297.4383381 ohms Resistor simulate portion of No-Load losses - connected direct in parallel with the source
R_1 0.244894922 ohms Stator Resistance
X_1 0.443833825 ohms Stator Reactance
R_2 0.095351691 ohms Rotor Resistance refd to stat
X_2 0.443914208 ohms Rotor reactance refd to stator
X_M 14.94061415 ohms magnetizing reactance
FullLoadSlip 0.013890011 none Full Load Slip
BarDepth 0.02603539 meter Equivalent Depth of rectangular rotor bar
Bar Type Aluminum Rotor Bar Material

============Selected Inputs ====================================
VLL 600 volts Line To Line Voltage
SyncSpeedRPM 1800 RPM Synch Speed in RPM (like 1200, 1800, 3600 etc)

=========== Model Performance Against Targets==============
Perf Variable Calculated Value Units Target Value FractionalError Weight Factor Weighted Squared Fractional Error Comment
FullLoadAmps 54.05775614 Amps 54 0.001069558 1 1.14395E-06 INPUT
FullLoadEff 0.919956267 none 0.92 -4.75356E-05 2 4.51926E-09 INPUT
FullLoadPF 0.869728336 none 0.87 -0.000312258 2 1.9501E-07 INPUT
FullLoadPower 44949.06836 watts 45000 -0.001131814 1 1.281E-06 INPUT
FullLoadTorque 241.8211117 N*m 242.094843 -0.001130678 0 0 Calc (redundant)
HLEfficiency 0.915434202 none 0.915 0.000474538 1 2.25186E-07 INPUT
HLPowerFactor 0.730605915 none 0.73 0.000830021 1 6.88935E-07 INPUT
LRC 405.0233091 Amps 405 5.75532E-05 1 3.31238E-09 INPUT
LRT 584.5416491 N-m 484.189686 0.20725754 0 0 NOT USED
NoLoadCurrent 22.56266102 Amps 13.5 0.671308223 0 0 NOT USED
BD_Tq 790.2624248 N-m 605.2371075 0.30570716 0 0 NOT USED
X2overX1 1.000181111 none 1 0.000181111 0.1 3.28013E-09 Thumbrule - split is not critical to model
R2overR1 0.389357569 none 1 -0.610642431 0 0 Thumbrule- NOT USED
X1overXm 0.029711912 none 0.05 -0.405761767 0 0 Thumbrule- NOT USED
Full Load Slip 0.013890011 none 0.013888889 8.07766E-05 0.1 6.52485E-10 Calc
Bar Depth 0.02603539 m 0.015 0.735692664 0 0 No target specified

SWSFE 3.54585E-06
Starting Power Factor =0.55

At 1st glance starting power factor looks unrealistically high. The model was based primarily on running data... the parameters change dramatically during starting. There is a bar parameter built in to try to model this, but it's not representatitve of every bar geometry. Having more data for starting (especially starting torque) would help improve accuracy of the parameters when it comes to predicting starting prediction.

One thing to note: R1 is not dramatically smaller than X1 and X2, so it is not surprising the model predicts relatively high starting power factor. So even though I'm not convinced the model is accurate I'm not convinced it's completely wrong.... maybe this higher power factor is more representative of smaller motors which tend not too have very high effieicny and have relatively high stator resistances?

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

 

[electricpete]

Bar type aluminum was obviously an input (like V=600vac), not a solved parameter. I put that in the wrong part of the output report.

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

 

[rbulsara]

If it helps see faq237-766. Engine usually is not the limiting factor in motor starting, the alternator winding and kVA rating is.

Rafiq Bulsara

http://www.srengineersct.com

 

[waross]

Motor Starting and Gen Sets;
DOL Starting.
Genset/Motor ratio.
3:1
The motor will start. The voltage drop will probably be noticeable.
2.5:1
The motor will probably start. The voltage drop will probably be excessive. A ratio of 2.5:1 is not recommended unless the motor is the main load on the gen set and any other loads can accept severe voltage drops. The engine speed may drop.
2:1
The motor may start. Expect the engine speed and frequency to drop. Added to the current voltage drop will be the Under Frequency Roll Off voltage drop.
Less than 2:1
The engine speed may drop to the point that UFRO drives the voltage lower than the holding voltage of the contactor coils. When this happens, the engine will lose speed and the voltage will drop until the contactor opens under high current. The gen set engine is then unloaded and accelerates. The voltage builds back up until the contactor closes. The engine slows down and the cycle repeats. If the load is an A/C compressor it will be stalled and the contactor will be making and breaking locked rotor current every few seconds. It is often the Automatic Transfer Switch contactor that cycles. It doesn't take many minutes for a set of contacts to burn clear under these conditions.
At ratios of less than 2.5:1 prime mover size is a limit.
At ratios less than 2.5:1 reduced current starting should be considered.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Hoxton]

Gas engines have a very low load acceptance - often 10 - 20% for first step from no load. They are not as good as diesels for motor starting. So may need a large gas engine to start a motor - then find that the running load is lower than the minimum load on the engine. Sometimes you meed several gas engines to start one motor, then shut the spare ones down.

 

[electricpete]

So... it's the transient of suddenly applying a load that the gas engine doesn't like? Perhaps that causes it to slow unacceptably before the controls can respond?

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

 

[electricpete]

Let me ask a different way. What is the definition of "load acceptance". Thanks.

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

 

[electricpete]

Interesting table at the bottom of the page here.

http://books.google.com/books?id=tW...epage&q=motor "starting power factor"&f=false

The table shows induction motor starting power factor by horsepower:
hp / starting pf:
<50 / 0.4
100 / 0.3
500 / 0.2
1000 / 0.15
> 7000 / 0.1


It sconfirms the trend that starting power factor tends to be high for small motors and decreasing for larger motors we expected. Although of course size of motor is not the only factor. (Speed, vintage, design, etc):


My model reported gave 0.54. I went back and examined it and removed the "horsepower rating" from the model because it is redundant with the combination: FLA, Voltage, Eff, Pf (and if you multiply these out there ends up an error in predicting horsepower which tends to affect the model). The result was that the predicted starting p.f. decreased to 0.515. Still seems a little out of whack with the reference linked above which I guess would put this 45kw (60hp) motor somehwere between 0.3 and 0.4 starting power factor. Like most things, if you need it to be exact... the motor OEM would have the most reliable info.

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

 

[ScottyUK]

Pete,

Can't help with the 'definition' but generally it's the relationship between stored energy in the rotating mass of the machine, the acceptable frequency excursion when a step load is applied, and how quickly the governor can act to recover the speed excursion. The relative mass of the machine's rotating parts and its power output obviously has an effect.

Gas turbines - probably not Hoxton's 'gas engines' - suffer badly in all respects because they are relatively light for their power output, so the frequency excursion is deep, and their ability to produce power falls away drastically as the compressor slows down. Our heavy-frame gas turbines couldn't accept more than about 20-25% step load change, especially in the first block from sync idle: we had to study this in a fair bit of detail because of our duty as a black start station.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[waross]

Sorry; my rules of thumb are based on diesel sets. For gasoline they will probably apply, but for gas turbines, see Scotty's comments.
Can you consider diesel?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[antigfk]

I size diesel gensets for motor starting applications daily. Without having manufacturer's full data (speed-torque/current/power factor motor curves plus speed-torque load curve plus moment of inertias of motor, coupling, gearbox, load, etc.) you can't perform an accurate motor acceleration analysis. When we don't have full data, we usually do a static motor starting study (kind of like a load-flow study, but with the motor drawing starting current rather than running current). It's been our experience that with a transient voltage drop of 35%, starter contactors may drop out or chatter. Voltage drop usually will start a low-inertia load, such as a centrifual pump or compressor. A voltage drop of 20% will usually start any motor, including motors driving reciprocating pumps and large fans.

To calculate starting kVA for your motor... 405A x 600V x 1.732 = 421 SKVA.
If you use Waross' rule of 3:1, you would size the generator to 135kW... you'd have to buy a 150kW. A 150kW generator will have an alternator sized at least 187.5kVA, and will have a transient reactance of no higher than 22% usually. Take 187.5kVA divided by .22 = 852 SKVA.
To find per unit voltage, take 852/(852+421) = 0.67. So voltage drop would be about 33%.

If you can get a 150kW generator set with a main alternator over-sized or with a low transient reactance, it will most likely start your motor. To be sure though, you might consider stepping up to a 175kW or 200kW genset.

At any rate, the engine will most likely not be a limiting factor. What you should be concerned about is the motor starting capability of the genset alternator (starting kVA), plus the starting kVA of the motor. This will let you calculate voltage drop, and you can take that voltage drop and look at how it would affect other loads on the system, controls/relays, and to see if your motor will develop enough torque to accelerate its load.

 

[electricpete]

Thanks for the explanation Scotty.

Setting aside the question of whether starting power factor is needed relevant (I am not qualified to judge... will let others figure it out), I was thinking some more about starting power factor.

I remembered we have previously measured starting power factor on a motor with almost the exact same horsepower rating (60hp). Specifically it was 60hp, 3600 rpm, 460vac, KVA code G, TEFC motor and results including power factor were presented and discussed a little here:

thread238-227811

In that case the, measured starting power factor seemed to be 0.32 on one phase and 0.35 on the other two phases... not sure why the difference. But the value fits reasonably close to what was suggested by the linked reference.

There is also a plot of power factor vs time attached to the 1st post of that thread. I remember at the time I noticed the very strange wiggle where power factor started increasing, then dropped back down before increasing to its final value and at the time I couldn’t think of any explanation for that weirdness. Now I think maybe the motor overshots the target speed before it settles into steady state. So for a brief period the real power is reduced, at which point we see a relative minimum in current and in power factor. Would be interesting to see if simulation predicts that behavior as well. I guess if I get really energetic I could run my “model” on that motor to see:..
#1 how closely does it predict the power factor (which has some relevance to you... in order to see if you should trust the number that came out of my model for your motor) and..
#2 to see if similar swing in power factor occurs... maybe this weekend.


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

 

[ScottyUK]

Maybe not the motor overshooting speed target but the generator governor itself undershooting target as it tries to stabilise the speed? A critically damped governor would exhibit this to some degree, and an under-damped governor more so. Most small sets seem to be slightly under-damped to get fast response to step loads, at the expense of stability.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[electricpete]

The trace that I linked was a 60HP motor started in a 1300 MW grid-connected power plant. I am pretty sure the power system frequency didn’t budge.

But the speed overshoot phenomenon is predicted by theory as discussed here
thread237-248895

This is perhaps the first time I personally have seen anything measured that might tend to corroborate the theoretical prediction.

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