12 kW System Net-metering Excess Power

[Rob Brunet]

I am installing a solar system in Ontario with about 12 kW of glass and a Deger tracker. The Power-One inverters limit the system output to 10 kW. I am wondering if any of you have considered system configurations (that are legal within the microFIT rules) whereby excess power above the 10 kW limit could be diverted to the household via net metering? My thought process on this involved two potential configurations.

1) Using a Power-One inverter with a diversion load capability, such as the models available for wind turbines, with a smaller inverter connected on the diversion load output feeding power via a separate cable back to the household panel. This would have the added benefit of providing full PV power to the household during periods of power outage when the diversion load is fully activated; the corollary is that the household panel would have to be fitted with an automatic transfer switch to prevent back-feeding the grid. My recollection is that these switches are expensive (~$1600?). I am also not sure offhand whether the diversion load is capable of handling 12 kW or more of power. The alternative to a transfer switch would be to equip the diversion load output with an inverter having UL 1741-2 anti-islanding capability.

2) Employ a larger inverter capable of 12 kW of output and use a current limiter to prevent the output from exceeding 10 kW. A shunt would open once the current value was satisfied in order to provide excess to the household panel connected in parallel. I am not sure whether this configuration is permissible within the microFIT rules.

Any thoughts on these circuit configurations or other possible configurations?

[Paul Bailey]

The reason you are at 12 Kw in the first place is to Maximize overall output as close as possible to the 10 Kw rated ouput of the inverters( system label rating ) as long as possible in daylight hours. You are already getting about 30% more with the trackers.To work the diversion of that differance in power ( if and when) would cost alot and potentially be an integration issue not worthy of persuing. Maybe you need to just add a 1-2 array and net meter that only aside from your fit contract with the trackers if this is permissable.You could also just add a smaller 500 watt array and Enphase it to merge with house power to ease the load ( more watts if you have a bigger load) I'm sure Rob will chirp in here as he deals and Knows Aurora inverter capability and settings. and he knows the contract rules and I do not. That being said I don't blame you a bit for trying to get more output when you have more power... but once the system is inspected and in Service I don't think you can change it .. Paul:idea:

[Rob Brunet]

The system has not yet been installed and inspected, so there remains some design flexibility. Also, I am not in agreement with the supposition that the excess output is insignificant. I have examined the data logs for several systems for the period of August to December of 2010. The systems were both fixed and tracking configurations. Of particular interest were two systems that were equipped with almost exactly 10 kW of glass and over-sized inverters to 12 kW capacity. The microFIT rules stipulate that either the glass or the inverters must be limited to 10 kW of total output, and these systems chose to over-size the inverter in order to capture excess production about the 10 kW STC rating. The systems were equipped with Sanyo HIT glass and were installed in regions of Southwestern Ontario with excellent insolation values.

Upon reviewing the data, I was surprised to find that the systems in fact produced more than 10 kW on about 1/3 of the days during the fall period. On some of those days it was only for 1 hour, on other days it was for several hours. The name plate rating of the panels is determined for STC test conditions of 25 C and 1000 W/m2; with tracking for maximum exposure and lower temperatures during the fall months, the panels can readily outproduce their name plate rating, sometimes by as much as 20%. Also, these panels are STC rated with a tight positive only tolerance, which provides for upside potential in average actual output.

In the system I have specified, there is already a 20% over production factored in on the STC rated capacity. Once temperature, etc. is factored in, the panels could produce 10-20% more than the STC rating, or up to 14.4 kW at certain points. Based on my fall sample, this could happen on a third of the days (perhaps more often in the summer and less often in the winter, but not a bad average estimate). I think it is therefore worth considering capturing the extra energy.

You may wonder why I chose to over-size the glass. There are two reasons. First is that it allows the curve of output power versus time of day to reach a plateau at 10 kW (inverter limited output) more quickly and for a longer period of time than for a system with less glass. Second is that, with degradation rates likely resulting in about a 15-20% reduction in output over 20 years, the system will still be sized to deliver 10 kW at end of contract life. I sought the lowest installed cost per PTC Watt and the maximum amount of glass area that could be handled by the tracker.

My apologies for this lengthy off-topic diatribe. Any further comments on capturing the excess output over and above the microFIT contracted amount of 10 kW? Can I simply program an appropriate MPPT curve for the solar panels I've chosen into the Aurora wind inverter and use the diversion load in the suggested manner? It seems like a fairly low cost and straightforward approach to me.

[Ralph Day]

Rob,
You will find on reading the microFIT rules that there is a 10kw per connection point limit. I found that meant that I could not net meter my house solar and wind production. Each connection point must be separate (transformer) and not only a different civic address, but a different tax roll number! Crazy.

So, if you want to net meter your microFIT plus what capability for the net meter system, if all on the same connection point must be 10kw or less. Rules are not made to be broken in this case.

My system clips the power curve when 10333 is reached, but that's only when it's cold. During the rest of the year we might peak at 10kw, but there's no flattening of the curve (it's a longer day, with less peak because of the heat lessening efficiency). I think it would be a waste of time and money to try and salvage what little power you're ''missing'' for only 6 months of the year (especially with the connection point limits).

Re-cap: unless you can connect at separate points, on separate tax roll numbers, forget net-metering and microFIT-ing side by side. If you want standby power have a battery set and inverter charged with 10cent /kwhr power and sell 80.2cent power. Don't jeopardize your OPA contract.

Ralph

[Rob Beckers]

Hi Rob,

The reasoning for using 12kW of PV and 10kW of inverters is sound: You can easily oversize the PV side to 115% of the inverter and rarely (if ever) run into clipping. This due to temperature derating of the modules, inverter losses etc. In fact, the only times you are likely to see clipping would be in spring/fall when the sun is high enough on the horizon to produce good power, while it is still cold enough to limit temperature derating (and possibly on a few really cold days in winter with snow on the ground, clear panels, and blue skies). I wouldn't read too much into new systems where 10kW of PV outputs over 10kW from the inverter; if they are poly-Si modules (most are these days) they'll loose a few percent during the first months of operation (that's normal and accounted for in their rating). Better look a year down the road, and even then, output will continue to slowly decline over time. I don't have the statistics, but don't think we often see solar radiation in excess of 1,000 W/m^2 at our lattitude, you would need that and/or cold temperatures to get the panels to produce over STC. Putting much effort into utilizing excess power would seem of limited use (to me at least).

Having said all that, your system is a bit over 115%, so it'll likely clip more often. Unfortunately I would not know of a way to utilize the excess power. The Aurora wind inverters have the same hardware as their solar version, but the firmware is very different. For wind they use a voltage vs. power curve, and that works for wind turbines to do MPPT. Solar MPPT works completely different: Solar cells are (largely) current sources and for solar MPPT they are operated right at the edge where that breaks down. That point depends very much on temperature of the cells, and lighting conditions. It is certainly not a simple voltage vs. power relationship. The excess power will not happen at a fixed voltage either; it too depends on temperature and light.

Neither the solar or wind grid-tie inverters will work when the grid goes out (even if you diverted power). It's not about anti-islanding, it's fundamental to the way these inverters work: They behave like current sources, they'll raise the voltage as high as it takes to pump back a set amount of current (determined by the MPPT algorithm) into the grid. When there is no grid there is no reference voltage for them to work from. Battery based inverters function as voltage sources: They try hard to keep their output voltage constant regardless of the current that is being drawn. Very different.

The only way I can see that you could in theory funnel off excess power is by putting electronics in between the modules and the inverters. It would have to replicate the MPPT algorithm and load up the modules optimally, present a 'virtual PV array' to the inverter so it can draw off what it can, and pump the remaining current into another load. It can be designed/made, but it is not something easy (and it doesn't exist as far as I know). You'll then spend a king's ransom to get it CSA listed etc.

-RoB-

[Rob Brunet]

Rob, your explanation of the different implementation of the MPPT curves in the wind and solar inverters makes perfect sense. The wind turbine is essentially a variable voltage source influenced by the amount of current draw, whereas the solar panel is essentially a fixed voltage source (at least above a certain minimum insolation threshold) that only drops in voltage once overdrawn. So, I can see your logic now that an inverter designed for wind use would not be suitable.

I am not certain that I agree with the assessment that the amount of un-ultilized power due to inverter clipping would be insignificant. I think it is practically quite difficult for most system owners with systems like mine to determine how much excess capacity is not being utilized, since the MPPT implemented by the inverter does not draw current above the 10 kW threshold and therefore there is no way to determine what "might have been" available to draw. However, I do agree that output will decline over time, making this a declining benefit investment. Also, given Ralph's comments, it does not seem like an especially wise avenue to pursue from a regulatory perspective.

That being said, there is another hypothetical system configuration that I thought of. One could utilize a battery charging solar inverter with MPPT, such as the Outback or Xantrex, sized for 12 kW, then connect that output to a "dumb" inverter sized for 10 kW. There would still need to be a current limiter on the DC side that determined when the primary inverter was satisfied in order to open a shunt to a second inverter sized at 2 kW. I believe the use of batteries upstream of the inverter is offside in the microFIT rules, so the connection between the Outback and the regular inverter would have to be direct. Anyway, just a thought.

[Rob Brunet]

Ralph, I am surprised to hear about this complexity in the microFIT rules. I was hoping to install a wind turbine at my site for net metering to the house and sell back power to the grid via the microFIT solar system. Do you think that this prohibition against net metering is merely for cases where one wishes to utilize a Hydro One approved net meter to actually store "credits" for withdrawal during low wind periods, or would it also apply if one simply connected to the panel and used a battery system to level the power output? It seems to me that the second option has nothing to do with Hydro One, so they shouldn't be in a position to object. However, given their current zealousness to find reasons not to connect projects, I don't want to risk anything and would ask them before installing.

[Ralph Day]

Rob
Check the OPA application (I think) and you'll see the provision for applying for microFIT or net metering. Same app for both. I didn't recieve any reason for not allowing both except that is was in the rules. You have to apply for net metering, and if/when they see that there's a microfit of 10kw on the same connection point it would just be dis-allowed.

I'm hoping that in a year or so they'll see that it's not an attempt to rip off the utility or the province and allow net metering on the same connection point. It would require a new inverter for me, so it's not really in the cards right now.

If you're going to try for both then try and get clarification on why the microFIT and net metering can't exceed 10kw.

Best of luck with it.