A PV cell working at peak power generally makes 0.5V per cell. Open circuit they make between 0.6V and 0.75V depending on temperature. So on a normal day, a string of 3 72-cell modules will make 72 * 0.6 * 3 ~= 130V, and once the MPPT starts pulling power it will pull down to 72 * 0.5 * 3 ~= 108V. On a cold morning, the delta will be larger.
You can generally work on a ratio of 5/6. This is in fact very useful when debugging a system. If you log the PV voltage, you can use that information to see if the MPPT is running limited or flat-out. A limited MPPT will show a rising PV voltage.
It has nothing to do with efficiency. If you have 96V to start with (on a spring day), you’ll end up with around 80V once it starts to really work.
There are two efficiency numbers generally bandied about. The one is the efficiency of the modules themselves (around 17% or so). This is how much of the incoming insolation is actually converted to electrical energy. This number doesn’t matter, unless you are severely constrained for roof space. It doesn’t matter how efficient the module is, because you can overcome it by covering a larger area with PV modules. All you care about is the cost per watt.
The second “efficiency” number is more of a rule of thumb for South Africa, which is that generally speaking we don’t get 1000W/square meter, except maybe a few days in the year. Our climate is also so hot, that even when we do get that kind of insolation, the heat kills it (PV modules don’t like to be hot). That means that a 300W PV module rarely makes 300W at noon. When you plan, better work on 80% of the nameplate capacity. So again, it doesn’t really have anything to do with efficiency, more with our weather conditions.
Thank you Izak and Louis, fascinating information this!
Only 17% of solar actually being converted to electricity - eeek! So if any size panel was able to convert 100% of solar rays to electricity, I suppose panel sizes would be a quarter of the sizes they are now?
Somehow I think adding extra panels to my current setup may be a waste of money. As Louis said, may cabling is already done so no further benefit there, and my MPPT’s are limited to 35amps charging capability.
I was hoping that the last 6 panels I added with its own MPPT would help a lot in overcast conditions, but this turned out not to be the case. Unless I again did something wrong like my last screwup with the 2nd MPPT being on the wrong side of the shunt, lol.
When the clouds move in front of the sun, my wattage drops to around 700-800 sometimes and I’m fairly sure this was more or less the case when I only had the one MPPT with 6 panels also.
I expected a little more but suppose this is normal.
We can forget about better numbers. The reason for the low number is “silicon”.
In order to make electricity from light, you need enough energy to kick an electron out of its orbit. There is a certain threshold energy that is required for silicon (1.1eV according to the video I will post below). That corresponds to a wavelength of 1110nm, and anything below that simply makes no energy.
Image below stolen from video lower down. 4% of the energy is in Ultra voilet. 44% is in the visible range. And 52% is in infrared, but IR has relatively low energy and little of that can be used. So straight off the bat we’re limited to maybe 50% of the available energy, before we consider other factors (heat, reflection, etc).
After a small discussion with @JacoDeJongh on friday i found this thread. Ive been doing basic power monitoring on my house for several months with HA. For anybody interested PZEM004-T with a 100A current clamp works great for measurement. I interfave them with sonoff basics as its the easiest way of interfacing with an esp and providing it power.
My average household load is 22KWh per day and 8Kwh of that belongs to my geyser. Im very interested in the BNETA plugs. Might look at using some of them to measure usage on specific devices.
I recently discovered these Lovato Energy meters from ElectroMechanica:
They are DIN rail mounted and only 18mm wide (The size of a small circuit breaker) and don’t need any CTs. You break the live and run it through the meter (it has a shunt inside for current measurment)
This particular model has an RS485 port which supports Modbus RTU.
You can read Active and reactive energy, active and reactive power, voltage, current and power factor.
Very handy if you want to measure some individual circuits in the DB.
The only minor annoyance is that it can only measure power in one direction. If the direction of power flow is reversed it displays an error.
I’ve got an ET112 energy meter connected to a Raspberry Pi running Victron’s Venus OS which monitors consumption for our house. 2 of us with me working from home.
We’ve got a solar geyser on the house (SunTank) and I’ve got the thermostat to its 3kW element set to 55 degrees with a timer to turn it on early in the morning and late in the afternoon.
The house has GU10 lights which I replaced with LED’s when we moved in. We have a gas hob, electric oven and single inverter fridge/freezer with no other high users.
We average out about 12kWh a day, but the heat of the last couple of weeks brought the geyser usage down a bit.
