I want to add some context to this discussion. I am talking about 3ph 11kW pumps because 2 of these pumps are going to be in my future. They will only run during the day and run full tilt when they run. ( Star/delta starting).
OK, I am veering substantially off topic here, but hopefully the content of this debate justifies it having a more detailed written reasoning.
Firstly, I can’t find costs related to your PFC solution or even a reference to it being used at this size load. That is not to say it doesn’t exist, it is rather an invitation for you to add input that I am ignorant of.
That said, I will try and break down my research into layman’s terms for the wider audience as as I am sure you already know the components involved.
What I have found out is VFD doesn’t limit the surge it breaks it down into little DC bits, making for a very nasty waveform.
It tries to be on a bit and off a bit so the average overall effect is less of a surge.
(But it is still a bunch of high current surges).
A 6-pulse VFD unit is the cheapest, with pairs of pulses trying to reconstitute a 3ph sine wave.
My research suggested that VFD units should be substantially over-rated to their intended load.
(For context we are in the 1000$+ range already).
A 6 pulse introduces over 40 % THD ( total harmonic distortion). This can be improved by using a 12 pulse unit , an 18 pulse unit or a 24 pulse unit.
More switching pairs means smaller and smaller dc bits adding together making a smoother sine wave shape.
The THD improves every time but the cost goes up as every 6 pulses doubles the hardware and more than doubles the cost.
Even at the 24 pulse stage though, we still would need filtering to reduce the THD to acceptable levels.
However, the conventional inductive filter solutions are as expensive again. ( I understand this to essentially be a 1:1 400V 3ph 11kW transformer).
And this is why THD is a concern with an off-grid inverter and not so much with a grid supply.
Because a normal grid supply is already through a transformer which dampens the effect of the harmonic distortion on upstream generation.
This distortion is mixed in with everyone else’s non-distorted sine wave load, so the generation side has an easier time dealing with it.
But to be clear ESKOM can insist that your THD is <5%.
An off-grid inverter is already trying to reconstitute a sine wave by adding little switched DC bits together. ( With an HF design having less in-built inductive dampening than an LF design, but I don’t know if that’s significant). It also contrasts the grid in that this distortion is now a substantial portion of a far smaller total generation.
The other design weakness I see with this solution is that a VFD is a load-side solution.
I would need a VFD and whatever filter at pump 1, and I need to spend again for pump 2’s VFD.
A generation side solution can start pump 1 and then start pump 2 sequentially.
The cost breakdown there:
I need enough inverter surge capacity for, say, a 70 kW surge.
Which will be a 3-ph blend of AC-coupled and hybrid inverters.
I’d need 30kW ish anyway to run both pumps and other loads anyway, so an extra 40 kW of inverter surge capacity would be required
Battery -wise:
4 X 18s 304A Eve cell banks, I estimate starting current to be around 1C or less. Not cheap, but I would have had a minimum of two banks anyway, as this is a completely off-grid system.
So the inverter side solution roughly doubles the cost of the system, which may or may not be cheaper than 2 VFD’s + filters, but I think there would not be that much in it.
So to be fair, both solutions are expensive, but I’d rather have a bigger battery and be able to start whatever I liked with fewer components to go wrong.
