In PTW's arc-flash program, asynchronous motor default contribution time to an arcing fault is 5 cycles ( 0.1 sec for 50Hz system ). Current decrement is not modelled.
A motor current contribution - without decrement - for 5 cycles seems too long. Can anyone comment on what times they are using and why ?

All IEEE 1584 recommends is that all motors 50 Hp and greater be considered. It does not say how. What the standard is attempting to do is point out that motor contribution can affect the final result. The incident energy can go up or down - it depends on how the motor contribution affects the protective device clearing time. It does not really say how to model motors and in reality, this would become a dynamic simulation and much more complex.

What I have done in the past is model the system with all the motor contribution included and see if the PPE rating is valid. i.e. if you select an 8 cal/cm^2 rating, do your calculations fall below 8 cal/cm^2. Next the simulation with 50% motor contribution and then 0 motor contribution. Again compare the result with the PPE rating.

If you can experiment with the motor contribution and the PPE rating is still valid for the various cases, that is about the best you can do for now. This is a similar approach that I use with utility and other data as well.

Hope this helps!

Just curious... Why would a 50 HP motor be recommended for consideration and piece of equipment with two simultaneously running 25 HP motors not be?

Great question!

The 50hp cut off value comes from IEEE Std. C37.010 (also in the IEEE Red Book and a few others) Rather than re-invent the wheel, this value was carried over into IEEE 1584.

I would consider 2 - 25 hp motors as a 50 hp motor for the analysis. I would also look at the PPE rating with no motor load and full motor load just to make sure everything has been covered.

For what it's worth, SKM does allow you to reduce the induction motor contribution to less than 5 cycles.

I don't see a problem being a little bit conservative with motor contribution:

• In some cases, the motor contribution makes devices trip faster. Here, the motor scenarios would be ignored.
  
• In the event that the motor contributes fault current without reducing trip time, then this would represent a higher energy condition.

In either case, the variability of utility fault current and the assumtions that we make to accurately model it, (80%, 60%, 40% etc) seem to have a much greater impact on study results than how we model the 5 cycles of induction motor contribution.

SKM only allows 10 scenarios. These end up consisting of up to 5 utility scenarios paired with motors/no motors. This takes the software long enough to calculate, I can't imagine trying to subdivide the motor modelling if we had 10 more scenarios to work with!

For most small motors under perhaps 500 HP or so a fault current contribution of 3-5 cycles is reasonable. As the amount of iron in the motor increases and the quality of iron increases the residual magnetism takes longer to decay, therefore the fault current contribution from the motor lasts longer.

The 3-5 cycle contribution from a small motor is not nearly constant - the residual magnetism in the motor decays by an exponential curve with a time constant tau on the order of 3-5 cycles. From the maximum short circuit contribution at time = 0, by the 3-5 cycle point the fault current contribution has decayed by 63%.

By assuming motor contribution to be constant from 0-5 cycles then dropping to zero we are greatly overstating the motor contribution. In traditional short circuit studies (performed to determine equipment momentary, interrupting, and withstand ratings) this over-conservatism was OK. In the days of arc flash where higher fault current can mean faster clearing time and lower incident energy, we may be erring on the low side of reality instead of the high side. The magnitude of this error is so low that it will probably not be addressed for some time - there are many more important issues for the standards committees to address first.