synchronous motor starting 1

[farhad47]

dear friends

we have a exaust blower fan which is driven by two
synchronous motor with following specification:
10kv
4500kw
1490 rpm
nominal flow rate for each fan :12500 m3/min


we have to select starting system for synchronous motor from following options:
1)brushless synchronous motor, reactor starting

2)induction synchronous motor , liquid rheostat starter
3)brushless synchronous motor , thyristor soft starter

which of the above items do you recommend?

best regards

farhad47

 

[jbartos]

Suggestion:
3)brushless synchronous motor, soft starter
visit

http://www.hut.fi/Misc/ICEM2000/table2.html

for
Cui J., Wang C., Xia J., Zhou M., Yang J.
969
Overall dynamic analysis of brushless synchronous motor

 

[farhad47]

dear jbartos

thank you for your hint.
I went to that site but found only the title
of proceeding. Do you know any way to access
to text?

 

[Dossa]

Dear Farhad47

My immediate reaction was to say soft start too, but then I got to thinking.

The selection of the type of synchronous motor and the most appropriate starting method depends on more factors than just those in the info you have given. These are quite large motors so it is difficult to be specific without knowing more information. For example:

[ul][li]Are these motors still to be purchased, or are you trying to adapt some existing motors to a new application? [/li]

[li]If they are new motors that you are specifying, put the onus on your suppliers to meet your performance criteria - tell them what to achieve, not how to do it. If you're buying the fan separately, give the fan details to the motor suppliers (e.g., torque speed curves, inertia, etc.) Tell them of your electrical system constraints such as fault level, max permissible voltage dip, intermediate transformer details, etc. If you're buying the fan and motor as a package, the fan supplier will still need to know your system constraints to pass on to his motor subcontractor. [/li]

[li]Have you considered the life cycle cost advantages that you might get from a fully rated variable speed drive? If your process demands variable flow rates from your fans, the improvement in part-load efficiency of a VSD over a constant speed drive can be startling. The incremental capital cost of a VSD vs. a conventional installation can be recovered (in many cases) in two to three years, depending on the amount of part-load running involved. Also, if a fully-rated VSD is an option on this basis, you may be better considering a variable speed induction motor drive instead. [/li]

[li]If they are existing motors, the motor construction may guide you to your solution. If you have synchronous induction motors, you are probably stuck with 2). For reactor starting an SIM, you will probably still have to load the starting winding, which will need the starting resistor. If you just short-circuit the starting sliprings and brushgear, you will likely damage your brushgear and sliprings due to the high current that will flow. You might be able to apply a soft start to an SIM, but you will likely not get away from brushgear totally. [/li]

[li]If you have a conventional brushless synchronous motor, you could probably use either reactor or soft-start. For reactor starting, you would have to study whether you can get enough motor torque to accelerate the fan using the current limitation allowed by the reactor. If you can't get enough torque, you might try autotransformer starting as another option. [/li]

However, my immediate reaction these days would be to say soft start, though you would probably have to modify the brushless exciter (to replace the exciter dc field stator with an induction motor type stator) to allow motor field current to be available at standstill. Again, you might consider a full VSD conversion, but you would need to study the effects of the VSD on the motor (e.g., harmonic currents increasing heating, harmonic torques exciting shaft system and fan critical frequencies, need for separate vent fans) and the supply system (e.g., harmonic currents/voltages causing power quality issues for your other plant and your network service provider). [/li][/ul]
There is no hard and fast rule that says one solution is better than another for any motor of this type and size. Each application is different, and it really does depend on more than just the motor size as to which is the best solution. You need to address all of these peripheral issues, or have your supplier address them before homing in on your final solution.

PS I do not work for a motor or drives supplier, so I'm not touting for business here - I've just been there and done that in the past.

 

[swgrmfg]

Farhad:

Are you sure that these are synchronous machines?
SM's do not slip, and their rated speed (at assumingly your 50Hz system) should be 1500RPM, not 1490.
If ther are, here are a few more pointers:

1. Varying the speed of a SM is more complicated than that of a squirrel cage induction machine. A normal VSD is not enough. You would need a unit called 'Load Commutated Inverter' that controls line voltage and frequency and also controls the field. Unless your fan application requires speed varying (e.g. induced draft) I'd stay away from the LCI.
2. Starting your motor, brush-type or brushless depends on many mechanical and electrical factors. Here are some:
a. Back power (available MVA) to consider for voltage drops. Will your motor starting collapse your bus voltage?
b. Amortisseur winding design (this is the squirrel cage induction part of the SM), are they designed for full or reduced voltage start?
c. Your coupling to the fan drive, can it handle full starting torque? The coupling will actually twist during start!
3. Your starter design will depend on your motor design. Brushless being the simplest and only requires an external minimal DC supply source, normally up to 125VDC at less than 10Amps. But the brush-type requires a large field discharge resistor, a large DC power supply, and a large DC contactor with make before break sets of contacts, along with optimum phase angle field application circuitry.
4. The most economical starting method is a reactor start. You get more motor torque for less line current draw with an autotransformer. For more cost savings, and if your motors are on the same electrical bus, you can have a starting bus and a running bus in your switchgear.

 

[jraef]

farhad47
With all due respect to Dossa, I will assume that you have already ruled out a VSD for some reason. In direct answer to your question as posed, the best prospect of the 3 choices presented is the electronic soft starter. I must say here that I work for a manufacturer of such equipment so my opinion is naturally biased, but I will support the idea.

1) A reactor starter will be a 2-step starting process. Although the average starting current over starting time will be lower, the current and torque spike that occurrs when transitioning from start to run mode is just as high as it would be across-the-line. Some amortisseur winding designs can be severely stressed by that transition spike, not to mention the power train mechanical components. The same is true for Autotransformer starters.

2) A liquid rheostat design is going to be extremely large, cost significantly more money and require a lot of maintenance over the years of operation compared to a solid state starter. You still need a motor starter to begin with, the liquid rheostat just provides the variable reduced voltage.

3) A solid state starter, properly designed, will allow complete flexibility of starting requirements, years of trouble free operation and considerable advantages in motor protection. Anyone who provides them can and will integrate a full synchronous excitation control and protection system. Most of us now have it as a standard option.

In ithe interest of fairness, here are the available manufacturers of 11kV solid state starters in alphabetical order. ABB, GE, Motortronics, Siemens, and Toshiba. There may be others but these are tried and true designs available world wide.
Quando Omni Flunkus Moritati