DC dump loads for grid-tied inverters

System · November 2010

I have lurked on this forum for some time and appreciate the thoughtful comments and opinions of its members. I now have a specific question of my own that I am hoping the community will be able to help me with.

I am located in Ontario Canada and am installing a grid tied system comprising 50 panels of 240 W STC each, for a total of 12 kW STC of potential capacity. The panels are provided on two dual axis trackers, each of which is connected to a 5 kW Power-One Aurora inverter. The reason for this mismatch is to comply with Ontario rules for receiving $0.802/kWh under the microFIT program, which stipulate that either the panels or inverter must be limited to 10 kW.

Even with a realistic PTC rating of 11 kW for the system, there are likely to be times in the summer when the system has excess capacity that it can't deliver to the grid, due to the inverter limitation.

I have used the Power One inverters before in wind turbine applications (Google Biro Air Energy) and know that the wind version includes a DC dump circuit that is activated when the turbine voltage exceeds a pre-set value. This is provided to prevent turbine over-speed in high wind conditions by diverting excess power to a DC resistive load, effectively providing an electrical braking mechanism, while at the same time preventing damage to the inverter.

My question is whether or not Power One solar inverters (or other inverters, for that matter) come with this same capability and, if so, whether that DC output could be wired to a 2 kVA auxiliary inverter that could deliver the excess power directly to the household panel? There would be no ability to store “credit” for excess power, since that would require a separate application for net metering to our utility, but the household baseload in summer should be sufficient to absorb any potential excess. Since I would not be seeking $0.802/kWh for this power, I don't think it poses a regulatory conflict with the microFIT contract.

Your thoughts on the technical and regulatory aspects of this approach would be appreciated. It would be great if the excess capacity could be utilized and not wasted.

mike95490 · November 2010

Re: DC dump loads for grid-tied inverters

GT inverters throttle themselves to their internal safe programed limit. You can feed a 5KW inverter with 18KW of panels, and it will only put out 5.0 KW ( or whatever it's programmed limit is) No dump loads needed, it will only pass thru what it needs.

Cariboocoot · November 2010

Re: DC dump loads for grid-tied inverters

What Mike said.

Wind must have a diversion load, solar doesn't need it. I can well understand your desire to harvest every last potential Watt, but there is no practical way of doing this with a grid-tie set-up as there's no stand-alone charge controller.

A hybrid system such as the Xantrex XW series works something along these lines, although it charges the batteries first and then sells to grid. Perhaps the new high Voltage charge controllers just being introduced may offer a potential solution?

This is why the panel/inverter match-up is based on average potential output. Sometimes the panels will put out maximum - 100% or even more of their "nameplate" rating. Usually it's less, especially in Summer when it's hotter (panel efficiency can drop drastically).

Run the PV Watts program http://rredc.nrel.gov/solar/calculators/PVWATTS/version1/ for the proposed set-up and try a couple different configurations to see just how much "surplus" power you're likely to miss out on.

jimwilks · December 2010

Re: DC dump loads for grid-tied inverters

I have a new Ontario MicroFIT ystem setup which largely mirrors yours with the exception that I have just a single 10kw inverter.

The solution I am toying with is to switch via DC contactors 1 of the 7 strings of panels. The 1 string will go to the 10kw inverter when output is under 10kw and go to a new smaller inverter to backfeed house loads otherwise. This will require "smarts" to determine when to switch the string, but I'm up for that sort of challenge.

This method could also be expanded to cover more than 1 string if necessary to do it in stages as the day progresses.

Jim

Ralph Day · December 2010

Re: DC dump loads for grid-tied inverters

Welcome Jim
I think you'll find that switching from fit to household would cause more problems that it's worth. If your production hits 10kw then some of it drops out it'll be below 10 kw and re-attach...and so on. I find my 12kw of pv driving 10kw of inverters (Enphase) clip at 10330 or so, and that's just when it's cold out.

In the summer the production might peak at 10kw but with the warmed up panels not producing as much there's no ''extra'' power to use elsewhere. The oscillation from connecting and dis-connecting 1 string might not be liked by your inverter(s). What might the repetative surging do to either inverter. Maybe just invest some of the microFIT money in an inverter and battery charged off the grid for backup?

Ralph

jimwilks · December 2010

Re: DC dump loads for grid-tied inverters

Ralph:

You raise some good points. Clearly I need to document how often my inverter is clipping my yield before I do anything rash.

I've no doubt that I could eliminate any switching surges and any "hunting" as you described with good engineering practices.

One other factor is that I'm "stuck" with 11 extra panels, 8 of which could make another string that won't meet Ontario content next year. Added to my setup, they could improve both MicroFIT and backfeed gridtie yields.

Jim

stephendv · December 2010

Re: DC dump loads for grid-tied inverters

This question has come up a few times in UK forums because the newly introduced FITs over there will pay out more for PV energy you consume yourself, so it pays to always use 100% of the generated power.

One of the problems is that you'd want to only use the exact amount of power that you're over-producing and no-more. So simply turning on a relay to turn on a heater won't always work, e.g. you have a 1kW heater but are over producing 700W.

Someone has built a DIY solution using a bunch of 500W immersion heaters and relays and then using 2 current sensors to monitor load and output to switch them on incrementally. Would be nicer if the output was completely variable in a given range.

There is a commercially available unit from these guys:
http://www.coolpowerproducts.com/uk/emma.html although bl00dy expensive.

Some discussions on the topic:
http://www.navitron.org.uk/forum/index.php/topic,11790.15.html
http://www.navitron.org.uk/forum/index.php/topic,9072.0.html

Ralph Day · December 2010

Re: DC dump loads for grid-tied inverters

Jim,
Do you mean to say you have the full 10kw of microFIT and are net-metering?

''One other factor is that I'm "stuck" with 11 extra panels, 8 of which could make another string that won't meet Ontario content next year. Added to my setup, they could improve both MicroFIT and backfeed gridtie yields.''

I got gyped on net metering. One connection point of 10kw. That means full microFIT for the 10kw or a blend of domestic wind and solar (3.1kw) and the rest microFIT (6.9kw). Not with the microFIT payout! So I have 2 meters, one for revenue, one for load...battery charging mostly.

Do you currently have batteries and inverter? If not, why not use the remaining panels to start a critical loads backup...some batteries, charge controller, inverter...

Ralph

jimwilks · December 2010

Re: DC dump loads for grid-tied inverters

Ralph:

I've 12kW of panels now on a 10kW grid-tie inverter. I have a parallel connection with 2 meters.

While I was waiting for Hydro One and ESA to get together on a cutover date, I unofficially "tested" my setup with a backfeed to my house breaker panel. The one conclusion I came to was that the new "smart" meter does not support full net metering. Specifically, I could slow down and stop the meter, but it will not go in reverse.

I don't have off grid batteries and the like. Instead, for backup power I have:

5kW UPS with new batteries.
5kW generator

All of the above are permanently wired to critical loads via transfer switches to cover me in the event of power outages.

I've only had my PV system running since the end of October, but I've had peaks of 9kW. My thoughts are that surely if I can do 9 kW in November, I can do vastly better in the prime summer months? If I produce over 10kW, the inverter will clip it. That seems like needless waste and thus is a challenge to me. Diverting the excess PV power to a backfeed is the perfect solution for a fairly minimal cost?

Finally, I have enough "spare" solar panels to add another 1600w string to the inverter. That should help improve the low output days and also increase the number of 10kW clipping events.

As mentioned before, I'm mulling the concepts over now and this forum has been a great source of 2nd opinions.

Jim

mike95490 · December 2010

Re: DC dump loads for grid-tied inverters

jimwilks wrote: »

I've only had my PV system running since the end of October, but I've had peaks of 9kW. My thoughts are that surely if I can do 9 kW in November, I can do vastly better in the prime summer months?

Diverting the excess PV power to a backfeed is the perfect solution for a fairly minimal cost?

A) summers, you may have better sun angle, but als higher temps may reduce wattage.

Backfeed ? to what ? You are already "back feeding" your meter right? (NetMeter, GTI....) Are you looking for additional loads to power from your HV DC ? Once your inverter hits 10KW, you can feed any more, you need another GTI inverter, best way to use your spare panels.

Ralph Day · December 2010

Re: DC dump loads for grid-tied inverters

Jim,
Smart meters that will net meter say Net on them (check the difference between your two meters). I talked to the meter teckie that came by my install shortly after commissioning. H1 had sent him out because my load meter ''wasn't working''. They put me on an estimate of 100kwhr per month and I hadn't used any by that time. We got that straightned out with a phone call (just use H1 as a charging source...225kwhr since June).

So, net metering only happens with a smart meter that says net on the face. microFIT is essentially net metering as far as the smart meter ''knows''.

Ralph

System2 · April 2011

Re: DC dump loads for grid-tied inverters

There are very good comments, so I'll just add that we frequently build 11.5 to 11.8 DC KW PV panel rated systems in Virginia with Fronius 10 KW UNI inverters grid tied AC. Clipping occurs very infrequently in winter, almost never in warmer months. We often use Sanyo or other very high PTC rated panels (PTC/Rated Watts as percentage of production value). If you are not using premium panels, I would expect even less peak production.

I cannot see how it is economically feasible for you to create a string switching mechanism. If it caused grid tie interruption, that would cost 5 minutes to reset. The equipment, even if you tried to capitalize on the excess value, has got to be more expensive than the extra couple of KW hours . Good Luck!

DC dump loads for grid-tied inverters

Comments

Categories