Adding more panels - DC fed to existing batteries?

[Oldgreybeard]

I think I'm getting closer to a way to install a few extra panels, at a 45° angle, to the wall of our house that faces more or less ESE. I could add another inverter to grid tie the ~2kWp of additional generation, but we already have a G59 system that's using up most of the available/allowable grid export allowance. I have a lot of spare charge capacity on the six US3000C's, though, as I have the inverter/charger (a Sofar ME3000SP) set to charge at 2.5kWp. because of a peak grid capacity issue when winter charging overnight.

I'm thinking seriously about just fitting an MPPT DC charging arrangement in parallel with the ME3000SP. Off the top of my head I can't see an obvious reason why this shouldn't work. I know the Pylontech BMS's communicate with the Sofar, so the master BMS can command the Sofar to shut down when the pack is charged, but would just having a DC charger in parallel with this cause any problem? My gut feeling is that it should be fine as long as the cut-off voltage of the DC charger is set to be slightly lower than the fully charged voltage of the batteries.

Any loss of the absolute maximum energy from the additional panels isn't that important, as they will only help to charge the pack up during the morning. Our main set of panels doesn't really start to produce much until mid-morning, and there are lots of times when we could do with an early morning boost, especially in the gloomier months. Only other thing to tackle would be adding the charge power from the second system to the energy/power monitoring system, but it should be easy enough to do that if I can find an MPPT charger that has a sensible data output.

Anyone see any glaring problems with doing this?

[openspaceman]

I can't help as I am still too ignorant of the concepts but am interested, what is the difference in conversion losses of running the MPPT DC current into the a ~50V battery but simultaneously running the inverter off the same busbar so that most of the VA goes into the inverter without charging the battery?

[Tinbum]

The thing you have to think about is the max charge current that is sent by CAN to the sofar inverter. This will be incorrect as the current from the MPPT won't be taken into account.

You can adjust for this with a Teensy or something similar to intercept, change and resend the CAN data. I do this with my batteries.

[Joeboy]

OGB, I have a set up like this. Will be home Sunday if you'd like photos? Only downside is that once the batteries are full the extra DC direct panels running through a charge controller will switch off. This can be worked around by keeping an eye on the soc but that's a small detail and overall its a great way to generate extra power at no trouble to dno or linesmen.
I use Victron, very reliable.

[nowty]

Its DNO and FIT friendly.

If you do this, the only thing I recommend is set the charge controller charge voltage slightly lower (1V or 2V) than that of your Sofar so they both don't try and whack full power into the batteries together when the batteries get close to full charge.

The only thing is the stats on the Sofar will be wrong, but generally that does not affect its operation.

[Oldgreybeard]

Thanks for the info, always good to get first hand experience, rather than just rely on theory. The plan will be to set the additional DC charge cut-off voltage slightly lower than the Sofar, so it won't even know there is any other charging source. As far as the Sofar data goes, it will presumably still report the PV and grid power, but the battery charge data will be wrong. I think that when it reports SoC it's getting that directly from the Pylontech BMS, so presumably that will still read OK. I can live without the battery charge data, as it's not that useful, main things I look at are SoC, PV generation, house consumption and the grid import/export numbers. If push comes to shove I can get everything other than SoC from the Eddi and Harvi if I add a Hub.

The advantages of adding extra panels facing ESE will be a significant increase in early morning generation, the avoidance of DNO hassle (pretty sure we're already at the max export they will allow) and a slightly better battery charging efficiency by going straight from DC to DC (rather than DC to AC to DC if I were to fit another inverter). Our batteries rarely start to charge before about 09:00. even when the sun's up at 05:30. Being able to cover the house base load and put some charge back into the batteries for an extra two or three hours would be a definite benefit. Chances are that by early afternoon this second system would pretty much stop generating, which would be fine, as the main system peaks in the early afternoon. Topping the batteries up more quickly each day will also allow the hot water to be topped up earlier and leave more spare available for charging the car later in the day.

The current plan is to see if I can fabricate cantilever mounts for the panels using stainless unistrut. Pretty sure I can use a combination of brackets and rails to get around 5 or 6 panels on the wall, so they double up as an awning to provide shelter above the path and also to provide a bit of shade above a window.

[Joeboy]

I did it off the back of Nowtys suggestion a couple of years ago. First year was 6 of 320W panels and a 35A Victron charge controller and a laterally purchased Victron lcd gauge, allows me to see what's Watt as i pass by in the garage. It worked that well that I have since added to the panel count (now 15) and replaced the 35A controller with a 60A. It's THAT !

Literally a built in trick to battery ownership. Not that expensive in comparison to adding a full new second system and seeking dno approval.

My SOC is steady as a rock. The addition won't affect it other than making it move faster.

https://www.bimblesolar.com/victron-mppt-150-60

[Stinsy]

The current plan is to see if I can fabricate cantilever mounts for the panels using stainless unistrut. Pretty sure I can use a combination of brackets and rails to get around 5 or 6 panels on the wall, so they double up as an awning to provide shelter above the path and also to provide a bit of shade above a window.

Why stainless?

Stainless steel doesn’t like aluminium, you’ll get galvanic corrosion of the panel frames. If you’re concerned about the strut rusting, get “hot dip galvanised” and make sure to spray the cut ends.

[Joeboy]

The current plan is to see if I can fabricate cantilever mounts for the panels using stainless unistrut. Pretty sure I can use a combination of brackets and rails to get around 5 or 6 panels on the wall, so they double up as an awning to provide shelter above the path and also to provide a bit of shade above a window.

Why stainless?

Stainless steel doesn’t like aluminium, you’ll get galvanic corrosion of the panel frames. If you’re concerned about the strut rusting, get “hot dip galvanised” and make sure to spray the cut ends.

Or aluminium angle, countersunk s/s allun screws and nuts? Or aluminium and a pal/shop that will tig weld for you.

[Oldgreybeard]

Why stainless?

Stainless steel doesn’t like aluminium, you’ll get galvanic corrosion of the panel frames. If you’re concerned about the strut rusting, get “hot dip galvanised” and make sure to spray the cut ends.

It needs to look reasonably good, and the galvanic corrosion risk is easily mitigated, just a matter of using the right approach. Not much of a risk anyway, my boat has 316 stainless fittings bolted/riveted to the aluminium alloy mast and boom and in 20 odd years it's not once had any corrosion rear it's head. The key is putting things together correctly and making sure the stainless is electrically isolated from the aluminium. For anodised panel fixings just a smear of compound in our relatively sheltered inland location will protect things for longer than I have left on this planet. If I wanted a belt and braces fix I'd just add HDPE isolating washers as well, but they would be overkill.

The rough as rats look of hot dip galvanised unistrut isn't going to be acceptable, and it won't even look half decent if I get it powder coated. I can clean and polish up the stainless unistrut to give a brushed finish, or I may just get it all powder coated satin black. Combined with end caps black powder coat would probably look reasonably OK.

Or aluminium angle, countersunk s/s allun screws and nuts? Or aluminium and a pal/shop that will tig weld for you.

Been talking to a local fabricator about getting something made up, but one issue is that the wall is pretty uneven, so there needs to be some adjustment to get the rails level and straight. Pretty sure I can do this with strut easily enough, it will allow enough adjustment to get things true.