CEE Relays Ltd.

87C Whitby Road, Slough, SL1 3DR

Telephone: (01753) 576477        Fax: (01753) 825661

Website: www.ceerelays.co.uk

Running a Transient Motor Starting Study

When running TMS studies without a utility infeed, it is important to model the source impedance, otherwise, there will be no voltage drop on the main busbars, even when the motor is started. To get reasonable results, it is best to set the SB generators' driving voltages to give unity voltage under normal running conditions. The easiest way to check that the driving voltages are set correctly is to use DAPPER's load-flow study.

What you need to do is to run the static load-flow study (all motors set to running) with the motor(s) that are to be started in TMS out of service and source impedance modelled. Then, you will need to change the SB generators' driving voltages so that the load-flow voltage is 1pu at the main busbar. Then you will need to run the TMS study (again with the source impedance modelled) after having put all the motors back into service.

Although it is possible to use TMS to check the generators' driving voltages, we find it easier to do so using DAPPER's snapshot load-flow study because a TMS study cannot be run if there are no motor starting events defined.

It's very difficult to discuss motor starting studies with on-site generation without mentioning I*SIM. Anything beyond the initial voltage drop is only an approximation in TMS, because generators are modelled as explained in our FAQs section about SB, PQ and PV generators. Using the generator inertia model with governor and exciter controls in I*SIM is the only way to properly model motor starting with on-site generation.

Problems Converging

There are several reasons why a TMS study might have problems converging. The best things to look for are:

1. Does DAPPER's load-flow converge? If not, run it with the approximate option to find out where the network's overloads are.
2. Double-check the TMS motor and load library models. These can easily be checked by selecting the "Ignore Network Effect" option in the TMS setup.
3. Too many PV generators on a network can simply end up transferring VARs between one another, never allowing the TMS study to converge. Using PQ and SB generators will work better. Don't forget that it's perfectly acceptable to have more than one SB generator in a network.
4. Change the solution parameters in the TMS study setup.
5. If possible, use the double rotor motor model instead of the single rotor model. The single rotor model can only be accurate during either starting or running, but never both. The double rotor model, on the other hand, can be accurate throughout a motor's operation. The double rotor model does not imply that the motor has two rotors, it merely reflects the more accurate equivalent circuit that has two circuits instead of one.

Don't be tempted to use the approximate option to get a solution in TMS. It is used to enable PTW to run an approximate study in order for users to be able to highlight where their network's problems are. These problems should be corrected and then the exact study should be run again.

Soft Starts

The TMS study does not differentiate between "Full Voltage (Square)", "Full Voltage (Square Transient)" and "Full Voltage (Curve)". All three are treated as a full voltage start. The three different soft start options are included because CAPTOR does differentiate between them.