GreenFields

The inverter can only use 5000W as AC output, but it can produce up to 6500W power if you include DC-charging to the battery. It's not just marketing fluff, but without adding enough batteries (and/or using the power directly as it generated), the additional panel Wattage could easily go to waste. Personally I'd go with to 6 panels per string, but that's not necessarily the right decision for all.

Has anyone else done this? Any thoughts on whether it makes sense? My Deye inverter used to have an incorrect measurement of the power taken from the grid, for example when in "Zero Export to CT" mode with Solar Sell activated. The core issue I've understood as being a manual extension to the CT coil leads going to the inverter. Some threads speak of doing these extensions with shielded CAT data cables, and the proper solution is probably to install a power meter. Basically I've always known precisely how much my household plugs consume, but for the geyser and stove, the Deye/Solarman readings of non-essential loads were always wholly inaccurate, by around 70% but always predictable. I tried it out to just adjust the CT coil ratio, from the usual 2000:1 ratio, through trial and error with running the known 3kW geyser load, to 7500:1, ie. just changed the setting on the inverter screen. Why I was less worried was because I'm trading with the grid through a 4-quadrant meter, so tripping up a prepaid meter is a non-issue for me. I also didn't care about 100% accurately knowing how much of my exported power is actually going to non-essentials versus finding its way past the municipal meter. I thought 95% accuracy should do okay for my needs. Is there anything seriously wrong with this approach that I'm missing though?

No. That function will only work as long as the grid is present. As soon as Eskom goes off, everything on non-essentials will go off. You can test this by turning off the main switch to simulate a grid outage.

If your geysers are connected as non-essential loads, there is no setting that will work. You would have to re-wire them either as essential loads or as AUX/Smart load, or with a change-over switch, bearing in mind how heavily that will impact your inverter and batteries' power output limits.

What is your peak current draw from the battery? If you are hitting close to 7 or 8kW during peak periods you could set the power limit to for argument's sake 4-5kW at night. This might help to compensate for a deeper discharge, if you treat it more like a 0.25C battery. Basically, you're going to run out of power overnight anyway, may as well do it slowly and go easy on your batteries.

If it's a choice between the AM-5 and AM-2, then the AM5. Out of the other batteries that are on the Victron inverter compatibility list, the Dyness seems to be a good value option to investigate if the intention is to buy multiple batteries to build up a bigger bank. Not saying it's the best or that others are not compatible, just that Dyness is listed by Victron as compatible (according to their website), and that prices seem to be among the best currently for 5kWh and 10kWh lithium batteries of a "recognized" brand name.

The strings should have the same Voltage (or very close) to place them in parallel, ie. you could do two strings of 8 in parallel on the 1st MPPT with 2 inputs, and put the other 7 panels on the single-input MPPT. But not mix a string of 8 and 7 in parallel. If the panels are facing generally in the same direction it should be fine, and a 2-degree difference in angle will not make a significant difference to the panel Voltage. Otherwise drop a panel and put 7+7 in parallel on the roof if in doubt.

I just don't know where in SA this is still permitted, assuming you are locally-based. More common around here is some fixed portion, asymmetrical crediting, and/or TOU. But technically the export to grid is more efficient than storage, if you just ensure to still run your batteries down regularly, don't keep it at 100% permanently.

Not enough info I think to give a definitive answer. Pushing back to the grid is not storage, it's trading with the power utility. Whether it makes financial sense depends on the terms and conditions in your location, but energy efficiency is good. Battery storage efficiency depends on your batteries and inverter, and number of conversion stages before the battery.

Maybe bear the specs in mind. I think the Victron is 5kVA as opposed to 5kW, with only 4kW real power rating. On the other hand the pass-through of 50A is quite good compared to the Deye's 32A or thereabouts. Also with the low-frequency vs high-frequency design, also the AC-coupling vs DC coupling conversion to storage, it's not a straight 1-to-1 comparison. Your application could be a deciding factor.

I would have expected to see a shift in the marketing approach, ie. fewer ads promoting the product, technical specs, price and the loadshedding benefits, and more ads evaluating the investment case, electricity per-unit savings, payback period, projected Eskom increases, etc. Perhaps some attractive finance options to make the decision easier? Am I just looking in the wrong places?

Do you have non-return valves installed that could prevent a reverse-flow thermosiphon from being set up? Can't make it out on the images.

Something else is happening here. You're showing in 2022 you were averaging 44.6kWh per day, and now 12.313kWh per day. This drop in consumption cannot be explained by 4x550w panels, especially in Winter. Those panels would generate approx 10-12kWh per day on average during the year, and less this time of the year. Could it be that you just became more frugal?

I think you have the gist of it, though you might struggle to do solar charging at midnight.

Repeating a bit of the above. Okay, so this is looks to be the main source of your issues, that your battery is the limiting factor of your power output. You could double up with a second battery so that you can make better use of the total 5kW power capacity of your inverter during loadshedding. It's very easy to exceed the 2.5kW limit of one battery, the geyser alone will do it, but it could be as simple as a kettle and a toaster together to trip your house. It's still not a good idea though to let your geyser chow through your battery capacity. It is better to connect your geyser as a non-essential load, so that when power goes off, that the geyser also switches off. The other alternative is to put your geyser on as a Smart Load, so that you could run the geyser as long as there is sufficient sunlight/solar power, but if that drops away, and the battery gets too low, the geyser will then again switch off. To find an installer, maybe ask directly to @Powerforum Store for the guys that usually work in your area. Some usernames come to mind, but I'm not sure who works where. Might also be a fair choice to use your first installer. Doesn't look like he's done anything obviously technically wrong except fit a system in a configuration that doesn't meet your needs or expectations.

Can you confirm what battery(s) and battery settings you're using? The graph doesn't show the geyser or any loads exceeding the inverter's rated capacity of 5kW, and it also doesn't show the battery ever exceeding around 3kW output. Could still be anything else, but I just want to eliminate one of the obvious possibilities like a 0.5C discharge rate.

Thanks @TaliaB, was not aware, and I humbly stand corrected/educated.

My 2c' worth. I'm going to see it the other way around, and start with the maximum charge and discharge currents on the inverter. For the 5kW inverter it's I think 115A, and on the 8kW inverter it's 190A, so it's based on the power ratings of the inverter at approx 48V. So under normal operation except for transient surges you should not be exceeding that Amp value significantly, and accordingly they've specified the fuse rating at 150A and 250A respectively, with wiring sizes to match. No matter how much battery capacity you install, for me it's the inverter and its rated power that should dictate the maximum discharge limit in the event of a fault, even if the batteries could comfortably push out more current. The magnetic ring you're referring to - if you mean the CT coil - is not intended to be fitted to a battery cable. It must go onto the incoming AC mains wiring. I think you need to re-read that section of the manual carefully.

What is your battery Voltage at the time of testing? Possibly below the value in Setting 12, ie. 24V "back to utility" Voltage? Considering you may have discharged to 21V cut-of Voltage overnight.

Agreement, as above, to create clearance on the sides and behind the inverter for airflow. Battery cables away, maybe against the side walls, and the inverter better centralised in the cabinet. Is the airflow upwards? I'm assuming the inverter's cooling fans move air upwards. If so, I would add, to move the surface-mount DB's away also, ie. possibly against the side walls of the cabinet also, else maybe on the other side of the wall? Regarding the vents, to have two, the lower vent to allow cooler air INTO the cabinet, and an upper vent to remove warm air FROM the cabinet back into the next room. Don't create suction below the inverter. Keep cables from obstructing the space just before the vents. No guarantee this will work to keep the inverter inside the cabinet. Best maybe just to move it after all.

Tending to like the idea of a hybrid as well. From the point of view, for the amount of power, 30 panels, so let's call it 15kW-16kW, I'm struggling to find a suitably-sized Solis inverter that's single-phase. By the time one's bought three of the 4.6kW grid-tieds to connect in parallel, you're virtually in the price territory to start looking for bargains on the 16kW Deye or maybe an over-panelled 12kW.

See if you can get this panel from Herholdts: JAM72S20 445-470 MR 35边框 (herholdts.co.za) JA Solar (herholdts.co.za) The specs are much closer to your original panel, and I think the pricing is reasonable. Pmax=460W, Vmp=42.13V. Imp=10.92A, Voc=50.01V, Isc=11.45A That gives you some better confidence to be able to mix strings if you must, but still try to keep one type of panel on one MPPT in a balanced number of panels in parallel pointed in one direction if you can. It's not a problem if Isc exceeds your maximum operating current. The Isc limit is much higher for that inverter.

Thanks. If I may ask, what's the nominal Voc and Vmp based on the panel spec sheet for your string? I guess the values at 25 degrees, and based on STC and/or NOCT, before doing temperature adjustments? Trying to get a sense whether the temperature issue is overblown for our SA conditions. I mean, will you ever get a case where the daytime temperatures hover around freezing, and the sun then emerges behind a cloud at noon? Sutherland's looking promising for the week, but I don't expect a massive wave of inverters failing coming from that area.

One of the recurring solar topics here is what the impact is of temperature on MPPT Voltage, and choosing your panels' Voc with some spare gap in theory to the inverter's maximum input limit. I'm just wondering whether the current cold snap could bring a surge of failures in some parts of SA. For example, Johannesburg's forecast for Tuesday morning is showing freezing temperatures overnight, with early morning sunshine around 7am-9am. Are there any folks around those areas with longer strings of around 450Voc that could share their real Voltage measurements and/or experience in these conditions?

Put in the CT coil, even if you think you won't need it now. It will allow the inverter to send battery power to the non-essential loads as long as the grid is present. So if you can buy cheaply at night and reduce dependence on the grid even if it is there. Zero Export to Load is all you need for now, but that might change. Energy pattern generally set to "battery priority" but actually doesn't really matter until you get solar panels. Grid peak shaving doesn't make sense to use. Your problem is not that you have a low-power rated supply line, your problem is that you sometimes don't have any power at all. Don't use this function, or set it to a high value, so that as long as the grid is available you will draw as much as you need and not let the inverter place a limit on what it will take from grid. It might still make sense to use the timer if you set all time slots day and night to 100%, but only tick the grid-charge boxes against the cheaper time slots at night. That way you will reduce the amount of cycles from the battery, discharge it deeper during the day before recharging at night. To be weighed up against the risk of running out of power before night falls, and depends on your usage. Which is to say, it will only discharge during grid outages, but not discharge any further once the grid comes on again.