Modeling Grid Faults in FAST and possibly PMG's

I have been using FAST in order to model a 1.5 MW wind turbine. I am interested in comparing performance of a medium speed and a high speed wind turbine. For this I am using one of the standard tests cases provided with the archive.

I have two questions.

  1. I want to understand the effect of grid faults and its influence on the dynamics of the drivetrain. I am using the high speed shaft torque variations as a measure of this influence. Is there a way of doing this in FAST?
    I am currently setting generator time off to a certain value to simulate the turbine going offline and comparing performance. However, this is not providing me results the way I am expecting them to look. Ideally I would like to spike the value of the gen torque and set it back to baseline value but I can only “play” with TimGenOn or TimGenOf. I have been modifying the drivetrain properties to reflect changes from a HS to MS drive train concept and outputting “HSShftTq” and “GenTq”.
    I have a Matlab script for a torsional model of a drivetrain with almost non-existent aero and generator models (see picture below). I have set some aero torque and generator torque to simulate its effects. I induce a grid fault by creating a disturbance on the torque input on the grid side and the HSS torque variations for a medium and high speed gearbox is shown below. This is what I would to simulate in FAST.

  2. At some stage I want to incorporate a more comprehensive model for generator and probably convertor dynamics. I am specifically interested in modelling permanent magnet generators. Has anyone tried to model permanent magnet generators in FAST? I have seen some work in literature on using standard models available in PSCAD or some other software. Is there a way to link it to FAST for simulating turbines along with models of fully rated converters?

Any suggestions would be of immense help.

Best Regards,


Venkat Ramakrishnan
Graduate Student
Michigan State University


Dear Venkat,

In the current version of FAST, TimGenOn can only be used to model turbine start-up at the beginning of the simulation (without the turbine operating beforehand) and TimGenOf simply disables the generator for the remainder of the simulation. It sounds like neither of these are options you want. You can get FAST to model the generator torque transients you desire, but it will require you to implement your own torque logic in one of the user-defined torque options–that is, either the UserGen or UserVsCont routine or Simulink (it doesn’t matter which one). With any of these options, you could implement torque spikes, periods of zero torque, normal operational torque, or whatever you want.

I can’t really answer your second question. NREL has a team that is currently working on developing improved models of the electrical drive for implementation in FAST. I suggest you contact Mohit Singh and/or Ed Muljadi of NREL for more information. (I don’t think either of them checks this forum.)

Best regards,

Dear Jason,

Thank you for your reply. I can now look into the possibilities of modeling the torque variations by developing my own routines.

Ed Muljadi and Mohit Singh have also been very helpful. Thank you for passing on their contact info once again. I just wanted to ensure I wasn’t missing out on any in built features or previous (similar) work done by others in this forum. Hence, I posted the questions here.

Kind Regards,