08-05-2016 03:35 PM
I understand that when sizing generators for harmonic generating non-linear loads, such as VFDs, it is recommended that the generator be upsized by approximately 2.5x. I recently used SpecSizer to size a generator and found that it recommended the same size generator whether my load was a 6-Pulse VFD, which generates high levels of harmonic currents, or an 18-Pulse VFD, which generally generates lower levels. Both selected a generator that was close to the full load rating of the VFD. Is that correct?
08-12-2016 01:08 PM - edited 08-12-2016 01:09 PM
Well, the best answer really is "it depends". Unfortunately while most manufacturer's sizing programs do a pretty good job these days, there are still a huge number of factors that come into play to properly size a generator when harmonics are involved. the old times 2.5 rule worked pretty well, but resulted in oversized units, and issues like poor long term load performance and cost made trying to size a generator closer to actual more desirable.
The type of loads, the load factors, starting requirements, harmonic mitigtion techniques (if used), number of generators and other factors all come into play. Also the impact of regulatory issues, like emissions and permiting also now affect the selection process as well, at least if you really want to do it right.
Over sized generator ends, multiple harmonic mitigation techniques, and in some cases even things like synchronus condensers added to the electrical system may all be part of the solution process, unfortunately there is no good "one size fits all" solution, at least in my experience.
EGSA ( www.egsa.org ) has recently added a "Sizing to Service" course that appears to help address many of these kinds of questions, as of now I think they are only offered as either part of an Advanced class or in an industry meeting or Seminar, but their website shoudl have more up to date info.
Hope that helps, MikeL
08-19-2016 10:52 AM
So I guess you are saying that most generator sizing software does not include the effect of harmonics when sizing. Is there not a concern then, that people using these programs would end up selecting generators that are not suitable for their loads. If I were to choose the generator that was recommended by SpecSizer for the application with 6-P VFD's, I'm quite sure it would be undersized.
08-19-2016 12:22 PM
The generator sizing programs do a reasonable job of sizing to most applications, however you are sizing a component of a larger power system, and a "free" program can't replace good engineering practise. Inmy own opinion, I would consider a power system with multiple large drives to be a "complex" system, and there are a lot of factors that have to be considered when sizing the generators, switchgear, conductors, etc.
There are softwares, such as ETAP and SKM that can get you further in helping determine proper sizing, fault contribution and stability, and in some cases even those tools aren't enough, so more advanced tools like Matlab and Simulink get used. More experienced engineers can do more with simpler tools, but that is because they have the experience.
I work a lot with the results of companies who have solely relied on sizing programs (from just about all of thegenerator manufacturers around) and unless they are a single or simple parallel system driving fairly "easy" loads, like lights, motors, pumps, the results are usually not what was expected.
So if CAT (amd others) were to add the ability to calculate harmonic capability, just how far would you expect them to go? All modeling requires a good deal a real data to assure the assumptions are correct. Would you be willing to pay for a sizing program that did that? How much? What experience level of the user wold to design too?
Yes, I think there is a concern, however how much does a component supplier want to assume in the design of a power system? That is why good electrical engineers who really know how to size generators in complex power systems are VERY busy and make pretty darn good money. And it is shrinking pool of talent in my experience. But it does help keep me busy.
12-12-2016 11:51 AM
I would add that it is a relatively simple matter to estimate total harmonic distortion based on the nature of loads applied to a generator. However, verifying the generator will provide a stable voltage when powering the load is the first concern, and secondly, someone has to decide "how much distortion is too much" when trying to power the loads.
I'd guess that with most "best AVR" generators the generator can remain stable with a lot more distortion present than the loads will "like".
12-14-2016 08:43 AM
Thanks for your reply. I agree with you that oversizing by 2.5x for non-linear loading does lead to poor long term load performance and cost and I would hope that this would lead users to select a good form of harmonic mitigation to eliminate this need. Since oversizing also leads to much higher installation costs and emissions, harmonic mitigation would have a very quick payback and certainly would be the right thing to do from an environmental perspective. If the Sizing programs could properly reflect this benefit, it would go along way towards supporting this approach. If you have the time, please have a look at the IEEE paper I co-authored that describes an application where harmonic mitigation eliminated serious power quality problems and also allowed for 'Rightsizing' of the generator on a remote oil field pumping application (http://www.mirusinternational.com/downloads/PID369