The kilowatt-hours on a solar proposal and the rand they take off your municipal bill are two different numbers, and the gap between them is set by your household's load profile — what the house is drawing at each hour the sun is up — not by the array's annual yield. A system can hit the yield in the paperwork exactly and still convert very little of it into bill relief, because those units went out to the grid for a credit worth less than the retail rate you pay, or went into a battery that hands back less than it took in. Before you compare quotes on yield, compare them on when your house actually uses electricity.
The Tuesday I finally understood this, I was standing in the passage in front of the inverter's little screen with a mug of tea going cold in my hand. Ten past one, not a cloud, the inverter fan doing its steady whine, my wife's car gone and the kids at school. The screen said the panels were pushing a bit over 3 kW. The house was pulling about 300 W — the fridge, the router, the alarm, the standby lights on the telly and the microwave. Everything else was going out to the street.
Two metres away the municipal meter's red LED was flicking at export speed. That was it. That was the whole thing, in one glance down a tiled passage: I was making plenty and using almost none, and the gap between those two numbers is where the money either stays or leaves.
By half past six that evening it looked like a different house. Kettle, stove, geyser element back on, the telly, the lights, someone's hair dryer. The inverter had nothing left to give, the sun was down, and the meter was counting up at full retail rate like it had never seen a panel in its life.
The yield on the quote is a gross number
Every proposal I've been handed leads with the same figure: annual yield, in kWh. Four thousand-something, five thousand-something. It's a useful number — it tells you whether the array, the roof and the orientation are in the right ballpark — but it is a gross figure. It describes what the panels do. It says nothing about what the house does with it.
Illustratively, and only illustratively: a 4 kW array with five useful sun hours gives you around 20 kWh on a clear day. That is supply-side arithmetic, not a promise about your roof. Now the demand side. If nobody is home from eight to five, the house's base load might chew through a kilowatt-hour and a half of that 20 before anyone walks through the door. The other eighteen and a half went somewhere else.
The annual yield figure doesn't know that. It doesn't know your geyser is already hot at eleven. It doesn't know the pool pump is set for six in the morning and six at night because that's when the neighbour's runs and nobody ever changed the timer.
That is why two houses on the same street, with identical arrays and identical yield on paper, can end up with completely different bills. One of them uses its own midday electricity. The other one manufactures it for the municipality.
Export credits and retail rates are not the same rand
Here is the part the quote rarely spells out. An exported unit and an imported unit are not worth the same. You buy electricity at the retail rate on your bill. Whatever you send back earns the applicable municipal embedded generation tariff, and in the schedules I've read that credit per exported unit sits well below the retail rate you pay per imported unit. How far below is a municipal matter, and it changes, so read your own council's schedule rather than trusting a number someone quoted you in 2022.
The bigger difference is the shape of the scheme. Some municipalities pay a credit against your consumption — the exported units knock something off what you owe. Others apply a rate that isn't a straight offset. Others again cap export, or don't want it at all, or require you to apply through the SSEG process and get the meter and the NRS 097-2-1 paperwork right before any of it counts. There are installations where surplus export earns effectively nothing, and the owner only finds out after the first full billing cycle.
So when a proposal shows a nice round export figure in rand, ask which municipal schedule that rate came from, whether it is a payout or a consumption credit, and what happens in a month when the house exports more than it imports. Those three questions change the value of an exported unit without changing the yield by a single kWh.
A battery moves the units, it doesn't multiply them
The obvious answer is a battery, and a battery does real work. It takes the midday surplus and pushes it into the evening peak, which is exactly when the meter was ticking over in my passage at half past six. That is a genuine shift in value.
But it is not a free move. Round-trip losses mean the energy delivered back to the house is lower than the energy put in — how much lower is on the datasheet, and it varies by chemistry, by temperature and by how hard you cycle it. Illustratively, put five units in and you might get four-and-a-bit out. That missing fraction is the price of the shift, on top of the capital cost of the battery itself.
Which means battery sizing follows the same logic as everything else here: it follows the evening load, not the array. If the house only needs four or five units between sunset and bedtime, a much bigger battery mostly sits full, and the surplus that could have gone into it in the morning still goes out the gate in the afternoon. Buying storage bigger than your evening appetite doesn't raise self-consumption. It just buys idle capacity.
The cheapest battery in the house is a timer
Before anyone sells you storage, look at what can simply move. A geyser on a timer, running its element in the middle of the day instead of at seven at night, does more for the bill in a lot of homes than an extra battery module — and it needs no inverter upgrade, no additional SSEG paperwork and no new firmware.
The pool pump is the same story. So is the dishwasher's delayed start, the washing machine, the tumble dryer on a sunny Saturday, the ironing, the vacuuming, charging the drill batteries. Every one of those loads is a unit that lands on site instead of going out to the street at a fraction of its worth.
I'm not going to tell you what percentage of your generation you'll self-consume, because nobody can know that from a quote. What I will say is that the number is yours to move, and you move it by changing when the house works, not by changing the panels.
Measure your own profile before you believe anyone's model
A monthly municipal bill is the wrong instrument for this. It gives you one number for thirty days and hides the only thing that matters: the shape of the day. Weekday and weekend look different. Summer and winter look very different once the geyser and the aircon get involved.
If your meter gives half-hourly or hourly data, pull a couple of weeks of it and look at what the house draws between ten and three. If it doesn't, a clamp meter or a small energy logger on the main feed for a week or two will tell you the same thing, and it will tell you honestly. I did this with a borrowed logger and a notebook, and the picture it drew was far more useful than any yield figure I'd been shown.
What I ask before signing anything
When a quote lands on the table, the yield number gets ten seconds of my attention. Then the questions start. What does my house draw at noon on a Tuesday in October, according to your model? What share of that modelled yield lands on site rather than being exported? Which municipal export schedule did you use, and is it a payout or a credit against consumption? What round-trip efficiency did you assume on the battery, and what evening load did you size it against? Did you model the geyser on a timer, or did you leave it where it is?
An installer who can answer those has actually looked at the house. One who can't has looked at the roof.
The array is the easy part. It's a known quantity on a known roof and the yield follows from the sun. The house's rhythm is the messy part, the part that decides what the yield is worth, and the only part of the whole system you can change without spending a rand on hardware.
