Electrical Engineering Software

[controlnovice]

Different question from thread238-157322:

Do any of you feel it is required to verify Electrical Engineering Software short circuit, load flow or arc flash analysis, such as those from SKM, ETAP, EDSA, etc., with hand calculations? Is the software mature enough (years of being in service) to not perform hand calculation verification?

If you had to stamp (PE) the computer model results (versus no stamp required) would you verify by hand (short/quick calculation) that the computer results were within 'reason'? Is there a difference between the two?

Does it depend on the project size?

Any common practice out there?

______________________________________________________________________________
This is normally the space where people post something insightful.

 

[dpc]

I always do a "sanity check" of any calcs done by computer program (especially spreadsheets). Most of the time any errors are mine (data entry). But the computer program won't know that the 500 MVA transformer is really a 500 kVA transformer. It's not so much the computational algorithm that I'm concerned about as it is the software's lack of ability to find my mistakes, read drawings, open tie breakers, etc.

 

[davidbeach]

Always useful to do hand (or head) calcs as a sanity check as mentioned by dpc. It is always necessary, and particularly when using a computer for the calculations, to have a rough estimate of the results. If the results then don't fall in the range of the rough estimate it is necessary to find out why. Without some means of checking, you'll never catch stupid data entry mistakes or fundamental flaws in the calculations. Eventually it becomes possible to do the computer calculations and look at the result and decide that the results make sense, but until then, some calcs are useful.

 

[aicon]

I fully support the sanity check on software because of the garbage in garbage out issue, but more importantly if you are doing fault calculations which are determining switchgear duty you are potentially putting someone at risk. If you get it wrong, switchgear is liable to fail explosively if required to break currents greater than it is designed for. Similar points can be made about calculations for protection settings in that if fauts are not cleared then plant may be put at risk.
Getting the parameters of plant and/or the network configuration wrong are the most likely sources of errors. For voltage and powerflow calculations real measurements should correlate closely with calculations, however, it has been my experience it is hoghly possible that all other things being correct that the results from measurement may be less accurate than the calculation. This points again to having a good understanding of what you are calculation and explaining any differences.

There are potential difficulties with most software in carrying out hv network calculations wrong from the point of view of voltage and circuit loadings because you have not set generators up correctly so that they are supporting voltage way beyond their capability. e.g. set a voltage control set point without placing limits on the VAR capability particularly on the slack bus.