Have You Considered Reduced Voltage Starters When Sizing Generators?

by Visitor bakersj ‎08-12-2010 11:39 AM - edited ‎08-12-2010 03:12 PM



In the context of my job, I sometimes review sizing files to determine whether the recommended gen-set can be replaced by a smaller gen-set. Opportunities for downsizing the specification are possible at times. For example, I had a case in which paralleled 2500kW gen-sets have been replaced by a single 3000kW gen-set. Most recently, a review of a customer sizing file and its specifications, and utilization of some of the competitive sizing options in our gen-set sizing tool, resulted in reducing eight paralleled 3000kW gen-sets to six 3000kW gen-sets.

One of the competitive sizing options that I have used is reduced voltage starters for motors. The following examples provide information on reduced voltage starters and how they reduce the kW/kVA required for a gen-set’s loads. I am interested in applications in which you’ve been successful using reduced voltage starters versus across-the-line starting, and the costs of the reduced voltage starters you specified or installed with your gen-set. As you can see, reducing a customer application by one gen-set can provide a significant financial benefit.

Full voltage motor starting (across-the-line or direct on line) is the most severe starting method, but may be required by specification. A typical motor will draw six times its full-load current on an across the line start. When a motor is energized by a full voltage starter, mechanical and electrical problems can occur. Potential mechanical problems include broken belts, gears or couplings. Potential electrical problems include deterioration of motor insulation or brown-outs causing nuisance problems with other electrical equipment.

A reduced voltage starter helps to lessen these problems by reducing inrush line current and/or starting torque of a motor applied to the drive load. This is done by either reducing voltage applied to the motor during starting or by using only part of the motor windings during starting.  Voltage reduction starters will reduce a generator sets starting requirements (SkW/engine and SkVA/alternator). However, there is a cost-benefit trade-off between the cost of the voltage reduction starter and the sizing of the generator set. Various types of voltage reduction starters and their starting torque % of full voltage starting torque are illustrated below:

Reduced Voltage Starters 


* Variable Frequency Drive (60 Hz)
** Variable Speed Drive (50 Hz)

The Current Limit (C.L.) % is adjustable for both soft starter and VFD/VSD. Increasing C.L.% directionally increases the size of the generator set.

Here is an example sizing of a one-step load scenario with two 420 HP motors and one 150kW resistive load that indicated the following kW/kVA reduction in the gen-set nameplate rating:

Reduced Voltage Starters
1-Step Load Scenario: (1) 100 HP Motor, (1) 100 kW Resistive load





  • What is your field application experience with voltage reduction starters?
  • What are the applications you commonly use for each type of voltage reduction starter?
  • As a consulting/specifying engineer, what type of voltage reduction starter would you recommend to your client to reduce the nameplate rating for the generator set you will specify for their site application?

Please share your thoughts by posting below.

by New member savvymov
on ‎08-19-2010 07:24 AM

nice piece of information. but doesnt that add extra cost of starters for every motor installed.

by Contributor see-deif
on ‎08-19-2010 11:45 AM

Yes it will, Savvymov. You could use one starter for multiple motors with a series of contactors/breakers, but that could be messy and time consuming...you have to start one at a time.  Mr.. Baker has some good suggestions, but those special motors can be very costly, too.

Solid state soft starters can cause problems with KVARS and the exciter. They often have AC filter caps on the input to help filter out the spikes, but they are sized for a worse-case situation and can be too much capacitance (leading VARS).  This can lead to unstable voltage regulation. Then you have to switch caps in or out a different loads / PFs...again, messy and inconsistent with time. As the motor and load changes over time, either the starter has to compensate or the system will need to be adjusted.

If you use inductors or transformers, they can have very high inrush currents and very poor -- lagging -- PF.


There is a cheaper way if you have one gen and one motor or can segregate your bus so one gen can start one motor.


The basic sequence is:

1. start the engine, leave the excitation off. (New regulators have a "Disable/enable" input or use a contactor between the PMG and AVR.)

2. close the output breaker and motor breaker as soon as the engine is off the starter.

3. when it reaches about 70% speed, turn on excitation.

4. the motor and generator will accelerate together.


The electrical and mechanical stresses are less on both the genset and the motor this way. This will even work on synchronous motors--it is a little more tricky, but you can start on the squirrel-cage or amortisseur winding, then once it gets close to sync speed (98% Hz) you can turn on the field excitation. Or you can turn on the field excitation when you turn on the gen excitation, but you may have to delay or temporarily disable some protection functions.


If you want  more info, email me at see@deif.com.



by Contributor wlj1943
on ‎08-19-2010 12:18 PM

I have application experience with all of the methods mentioned, at voltages up to 15 KV class, and motor sizes from 25 to many thousands of HP. The starter selection choice should be application and driven load characteristics specific. There is no one single best solution; the needs of the process and the economic factors all enter into the evaluations. In general, for 400/460 volt (50/60 Hz) induction motors driving pumps or fans up to around 250 HP, our experience is solid state soft starters seem to be the most economical choice combined with the best flexibility in startup tuning, peak load reduction,low maintenance, and minimal space and weight. 


The gold standard is probably VFD's with IGBT based active Var regulating front ends, which can be sized to actually add leading Vars to the system during motor starting or long motor acceleration times, can regulate unity power factor to a whole system, and can be field tuned to minimize power system disturbances to near resistive load levels. Many isolated off grid facilities, such as mines, large cruse ships, container cranes, and cement kilns are some instances where the extra costs associated with fully regenerative active front end drives can be quickly recovered via smaller generator sets and increased efficiency with better fuel rate and lower emissions.  For smaller motors, use the lowest cost solution that achieves the desired result; sometimes just a PLC controlled sequenced starting routine can achieve the required results. For many complex industrial processes, the regenerative capabilities can significantly reduce peak loads and increase overall system operating efficiency significantly.


on ‎10-15-2010 05:36 PM

I believe it's important to pay attention to start timing.  Solid state soft starters have become so affordable, we see them on every lift station we do.  The cost justification versus oversizing the generator to compensate is truly a no-brainer.  The added bonus of these units speaks to the leading VAR issue noted above.  Many of these units (as well as VFDs) have time delays integrated into them to prevent restarts for a certain period of time after power is restored.  Therefore, if you stagger your motor starts - even by 15 or 20 seconds each - you keep that VAR level at a minimum while doing your generator and distribution system a measurable favor.

by Contributor Tearhun
on ‎12-07-2010 03:48 PM

What are the guidelines for sizing of a generator which is to energise a transformer?  For instance, if a gas generator 3512 or 3516 was operating with an output voltage of 3.3kV, in what way would the maximum size of a connected transformer be determined?  The problem is the momentary inrush / magnetising current when the transformer is energised and the resulting possible voltage drop. 

There are two possible scenarios which would be of interest.  The first would be if the generator was energising a unit transformer.  In this application, the generator would be rated at say 2MVA and it would be energising a 5MVA step up transformer 3.3/33kV with no other load on the generator.  The second application would be if the generator was on-line with a partial load, say 1MVA with a feeder transformer to be energised.

What are the guidelines for determining the maximum size of transformer which may be energised?  What may be done to AVR or protection settings to enhance successful energisation?

by New member TBates
on ‎12-10-2014 03:23 PM

Are capacitors for KVAR reduction (if properly placed, sizes, and switched) OK to use with a G3516C ?

Post a Comment
Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Help us grow the Caterpillar Community: Invite a Friend

Meet Our Bloggers