Connecting 10MW solar plant to the grid 1

[ters]

A fairly large PV solar plant, 10MW, is now in the feasibility phase. The proposed configuration consists of 10 inverters 1MW each 240V output coupled with a high ratio transformers stepping up to 46kV. All transformers would be connected in a “daisy chain” arrangement on the HV side (DELTA), using a metal enclosed switchgear attached to each inverter. The collector cable would end in a fully featured outdoor type switching station with a breaker, disconnect switches, metering, protection and the like, which in turn would tap into a local utility 46kV line.

I have never seen such a high ratio transformer (4600:240 = ~ 200) and I’m unsure what typical issues it might come up with. Can anyone comment on this?

Also, to implement some sort of line protection in the solar switching in this case seem to be challenging – inverters will have minimal contribution to the line fault as their output current is limited to only 1.1pu before the internal control circuitry shuts the inverter down. So it seems no conventional line protection will see faults on the line. And for the same reason even overcurrent elements could only detect faults on the solar collector system side (faults fed from the grid), but again cannot see faults on the line. Could anyone comment on how to approach protection issues in this case, in general?

 

[redfurry]

Ontario utilities regularly supply residential customers from the 27kV system, and I think also occasionally from 44kV. So, I don't think there's any intrinsic problem with the transformer ratio.

Protection shouldn't be a huge problem either. As you mentioned, there is little infeed from the inverters, so for any fault on any part of the system down to the inverter it will act like any other fault on the system where you trip based on the fault current supplied from the grid. No real issues with complicated fault paths or directional elements needed.

Of course in addition to overcurrent, there needs to be additional protection on the inverters for over/under voltage, over/under frequency and anti-islanding. It seems most large inverters already have built-in circuitry for these.

 

[Skogsgurra]

I have seen solutions where solar panels feed DC motors that drive standard synchronous generators. That makes the connection to the grid very simple and also provides a standard protection both lightning-wise and in other respects. It also reduces EMI problems and the generators are ideal for power factor compensation.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ters]

Thank you both for replies. I just did some research and it appears that 46kV class distribution transformers do exist too, so looks like transformer shouldn't be an issue, although those hanging on poles seem much smaller than 1MW.

But, Skogsgura, I cannot seem to see a noticeable advantage of the system you described, vs. connecting an inverter to the grid. With a DC motor and AC generator, there is more equipment involved, lower efficiency and much more mechanical complexity (whereas there is almost no mechanical equipment when using the inverter only). Finally, how one can control the DC voltage supplied to the motor as it varies with the PV output? And what sort of "governor" one can use in such cases to control the speed?

 

[Skogsgurra]

The control loop is extremely simple, you just control excitation of the DC motor. PV voltage doesn't change very much, perhaps .8 - 1.0 p.u. and that is easily covered by excitation adjustments.

It is true that efficiency and mechanical complexity suffer. But an inverter needs a filter that is not without losses and the protection against lightning is a lot easier with a generator than with an inverter.

A contributing circumstance may be that the equipment I saw were going to be placed in Africa where power electronics expertise seems to be rare.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[davidbeach]

If delta on the high side, how do you keep the voltage to ground under control during line-ground faults?

 

[ters]

Thank you Gunnar. Well, solar plants should belong to Africa anyway, but one I working on seems to be much close to the North Pole than Africa. Probably because taxpayer living around the North Pole are much more generous than those in Sahara.

The proposed inverter appears to be 95% efficient with harmonic distortion ~ 3% and power factor ~ 1.

Davidbeach, we would use a zig-zag grounding transformer located in the switching station for ground fault detection. At the same time, the assumption is that the utility already has or will have adequate line protection, so using a transfer trip they would disconnect the solar site any time when their line protection operates. Seems that proposed inverters+trafo combinations for some reason come prepackaged that way (DELTA on the high side).

 

[jghrist]

1MW at 240V is a lot of current. I'm surprised that a higher voltage isn't used.

 

[ters]

I'm surprised too. Moreover, inverter input voltage (DC) is in the range of 600 to 900V depending on the model, while the output voltage is for some reason 3 times lower, 200-300V, again depending on the model and also a wider AC operating range is needed to automatically adjust to the line voltage and keep power factor ~ 1.

 

[rcw retired EE]

I'm guessing that the majority of inverters are built for smaller size projects that will tie into the local grid at 240V. There is not much market for 10 MW inverters or even 1 MW. The 600-700V maximum DC input voltage is a limit of the PE panels.

I owuld recommend looking at using ten 1 MVA padmount transformers to collect power from the inverters Both could be located near the panels. Then run 13.8 kV cables to switchgear feeding a 13.8 kV - 46 kV step up transformer. The savings in copper cost might offset the added equipment costs. Also, the design uses more standard utility equipment instead of custom sized equipment. (Cheap commodity item versus expensive custom).

 

[ters]

There is not much market for 10 MW inverters or even 1 MW.

This statement is very true except some political decisions are about to change it in some jurisdictions .

The proposed supplier presently offers 1MW inverters in a configuration 2x500kW with 240V/240V/XkV three-winding step up trafos as part of the package (whereas X is line voltage), and supposedly 2MW units are coming on the market soon.

RCwilson, the 13.8kV collector was initially considered, in which case the pad mounted trafos could come with a bolted on box with HV fuses and disconnect switch. But then, the other option of connecting directly to 46kV was proposed due to possible savings of not having an additional 10MW trafo with associated switchgear and protection. Also, losses would be lower with one trafo only, vs. having two in series.

However, as you indicated, when stepping up directly from 200V to 46kV, the equipment becomes very non-standard, metal enclosed switchgear for 46kV probably doesn't exist, or if it does the cost may be prohibitive, in which case 46kV switches and fuses for all units would have to be outdoor switchyard type, etc.

 

[Skogsgurra]

The biggest problem wold be 10 MW at 240 V. Tens of kiloamps are not easily handled. Switchgear, bus bars everything will be very non-standard - or at least very expensive compared to what rcwilson recommends.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ters]

There will be a transformer in any case, but what voltage level it will it step up is still to be determined. On the low voltage side, a bus duct connected to the transformer to handle 2400Amp may have to be used.

 

[jghrist]

The biggest problem wold be 10 MW at 240 V. Tens of kiloamps are not easily handled. Switchgear, bus bars everything will be very non-standard - or at least very expensive compared to what rcwilson recommends.

This isn't how I read the OP. I thought there would be ten 1 MVA 240V-46kV transformers.

 

[ters]

Yes, that is Option A). Option B), probably more feasible, there would be ten 1 MVA 240V-13.8kV transformers + one 10MVA 13.8-46kV unit.

The DC side is a bit fuzzy too. Depending on the equipment selected, a 10MW site will have several hundred combiner boxes (which combine x number of PV strings in series) and then several tens of such boxes connected in parallel would be served by one inverter. The remote monitoring than becomes quite complex - in order to know that all strings are healthy; each combiner box should be smart, making the SCADA fairly complex.

 

[tinfoil]

I would suggest option b) should use two 13.8-46kV units, unless your project can stand 14-20 week delays waiting for a replacement xmfr.

 

[ters]

Thank you for pointing this out tinfoil. The developer will actually be developing more than one site with the same ratings and capacity, so there is also an option of keeping spare key components in stock.

Use of only one step-up transformer can be found at many small power generation facilities. However, the issue of redundancy you mentioned definitely needs to be considered in this case.

Similar to transformers, there is also a dilemma as what is an optimal number of separate collectors to be used. The present idea is to use two separate collectors (5MW each). More collectors require more interrupting devices, more protection equipment, etc, same as for more transformers.

In terms of redundancy, option A with non-standard equipment and no intermediate transformers is better. But to make use of standard equipment and have some redundancy, seems that additional equipment (and losses) will come with a price as well.

 

[oldnotbold]

I think one way to implement such a solar converter system would be to use a topology like the Siemens-ROBICON Perfect-Harmony Motor Drive system. This uses series connected low-voltage cells, (i.e. each one fed from relevant PV units), in each phase linked together to build up the medium-voltage power output of the system. With this type of topology, the inverter can be scaled precisely for a very wide range of voltage and output power. Additionally, this type of system offers increased availability because of its modular design and ability to bypass any one cell during operation while maintaining the full output voltage.