The Real Cost of a Load-Shedding Hour in 2026: A Rand Breakdown for Homes and Small Businesses
A full rand breakdown of what one hour of load-shedding costs a WFH household and a small café in 2026, and how pricing downtime cuts solar payback.
Published 2026-06-01 · Updated 2026-06-29
Why an hour of darkness is never just 'an hour without power'
Most South Africans price load-shedding the wrong way. They look at the Eskom or municipal bill, see a small saving while the power is off, and conclude that an outage is roughly free. The bill is the least important number. The real cost sits in the things that don't show up on a statement: the hour of work you couldn't bill, the litres of fuel a generator drinks, the milk that turns warm, the wear you put on a R30,000 inverter or a R15,000 generator, and the customers who walked past a dark shopfront.
Once you add those up, an hour of load-shedding has a rand value — and for most households and small businesses it is a lot higher than the few rand of electricity you didn't draw. Putting an honest number on it matters because that number is exactly what backup power or solar is buying you. If you underprice the hour, every inverter, battery and panel looks too expensive. If you price it properly, the same kit often pays for itself in a third of the time you assumed.
This guide breaks the hour into its five real components — lost productivity, fuel burn, spoiled food or stock, equipment wear, and lost trading revenue — then works two full examples: a work-from-home household and a small café. We finish by showing how electricity inflation, fuel inflation and a rand value on downtime reshape a solar payback calculation. The figures here are illustrative and not personalised advice, but the method is the point: build the number from your own inputs.
The five components of the hour
Lost productivity is the biggest line for anyone who works from home. If you earn enough to sit in the 31% bracket, a realistic net value of an hour of your output is around R250. You rarely lose all of it — some work shifts to before or after the slot — so a fair haircut is 50%, leaving about R125 of genuinely unrecoverable output per shed hour that lands inside your working day. For a two-earner household both stuck offline at the same time, double it.
Fuel burn is the loudest cost for generator users. A 5 kVA petrol generator running a partial home load drinks roughly 2 litres an hour; at about R23.50 a litre for inland 95, that is R47 per hour, every hour, paid in cash. A café's 8 kVA diesel unit under real load burns closer to 2.8 litres an hour at around R22 diesel — about R62 an hour. Equipment wear hides behind this: a generator needs servicing every 100–150 running hours, and at roughly R1,500–R2,500 a service that is another R15–R17 an hour you are spending whether you notice or not.
Spoiled food and stock is lumpy rather than steady. A household freezer survives a normal two-hour slot fine, but a bad week of stacked stages can cost you an R800 freezer clear-out; spread across the dark hours of those weeks, that is roughly R16 an hour of expected loss. A café carries far more risk in its fridges — dairy, prepared food, cold stock — so an R1,500 spoilage event across a run of unprotected outages works out near R37 an hour. The last component, lost trading revenue, is where a business hour gets brutal, and we price it in the café example below.
Worked example 1: the work-from-home household
Take a household where one earner works online during the day. During a shed hour that lands in working time, the breakdown on battery or inverter power is small because the kit just keeps the laptop, router and a light running: about R125 of unrecoverable productivity (the 50%-haircut hour) plus roughly R16 of expected food spoilage. That is around R141 for the hour — and most of it is the productivity, not the power.
Now run the same hour on a petrol generator instead of a battery. Add R47 of fuel and R15 of service wear, and the hour climbs to about R203. That is the hidden tax of going the generator route for daily, short outages: you are paying R62 in fuel and wear to avoid a problem a quiet battery solves for almost nothing. Generators earn their keep for long, deep outages where a battery would run flat — not for the 1–2 hour slots that make up most of a shedding year.
Scale it up and the annual picture is stark. At an average of 1.5 hours of shedding a day — about 548 hours a year — the share that overlaps the working day (say 35%) is roughly 190 productive hours at R125 each, or about R24,000 of lost output a year for a single earner before you count a cent of fuel or food. That R24,000 is the number a backup or solar system is really competing against, and it is why the bill-only view of load-shedding is so misleading.
Worked example 2: the small café
A café feels load-shedding through its till, not its laptop. Assume it turns about R1,200 of revenue in a peak hour at a 60% gross contribution margin — so R720 of contribution is on the table each peak hour. With no backup, the lights, coffee machine and card terminal die and roughly 60% of that trade walks out the door: about R432 of lost contribution. Add expected stock spoilage of about R37 an hour across unprotected outages, and a dark trading hour costs the café roughly R470. That is before the owner has even thought about the customers who now assume the place is unreliable.
Run that same hour on an 8 kVA diesel generator and the maths inverts. Fuel is about R62 and wear about R17, for roughly R79 an hour — but the R432 of trade and the R37 of stock are both saved. The café is spending R79 to protect roughly R470, a net benefit of about R390 for every shed hour it keeps the doors open and the fridges cold. For a trading business in peak hours, running the generator is almost always the right call, and a quieter, cheaper solar-plus-battery system that does the same job without the fuel bill is better still.
The lesson across both examples: the rand value of a shed hour depends entirely on what the hour interrupts. For the household it is dominated by productivity (R141–R203 an hour); for the café it is dominated by lost trade (R470 an hour without backup). A R720-contribution peak hour and a R250 work hour are simply not the same hour, and your backup spend should follow the bigger number.
Electricity inflation vs fuel inflation: why the picture keeps shifting
The two prices that drive your shed-hour cost move at very different speeds. Headline CPI is running near 3.1%, but Eskom and municipal electricity tariffs have for years climbed at double-digit rates — model 12% a year and you are being conservative by recent standards. Fuel is lumpier and tracks the oil price and the rand; a 6% long-run pace is a reasonable working assumption, but it swings hard in either direction. Wages, which underpin your productivity cost, drift up roughly with inflation at around 5%.
This split matters for which backup route wins over time. A generator's running cost is chained to fuel, so it inflates at the slower, more volatile 6% — but you pay it in cash on every single hour, forever. Grid-tied solar and battery, by contrast, lock in today's hardware cost once and then ride the 12% electricity-tariff escalation as pure upside: every year the grid power you are displacing gets more expensive, so the same panels save more rand. The faster electricity inflates relative to fuel, the more decisively the long-run answer tilts away from 'just buy a generator' and towards solar plus storage.
There is a crossover worth naming. For short, frequent outages, a battery's near-zero running cost beats a generator's R47–R62 fuel-per-hour easily. For rare, very long outages, a generator's ability to run indefinitely on fuel still wins, because a battery runs flat. Most of a 2026 shedding year is made of short slots, which is why the centre of gravity has moved to solar-plus-battery with a generator kept only as a deep-outage backstop — and why pricing the hour correctly, with inflation built in, changes the verdict.
Putting it together: pricing downtime changes your solar payback
Here is where the whole exercise pays off. Take a household installing a hybrid solar-and-battery system for about R120,000. Looked at on the electricity bill alone — say R1,400 a month saved, or R16,800 a year — the payback is roughly 7.1 years. On that number alone, plenty of people decide it is not worth it. But that calculation prices the shed hour at zero, which we have just shown is wrong.
Add the downtime value the system actually delivers. The recovered work output across the productive shed hours is about R24,000 a year. The generator fuel you no longer burn — say half an hour a day averaged out, 182 hours at R47 — is another R8,500. Stack those on top of the R16,800 bill saving and the system is delivering roughly R49,000 of value a year, which pulls the payback down from 7.1 years to about 2.4 years. Nothing about the hardware changed; only the honesty of the inputs did. Over five years, the bill-only view credits the system with about R107,000 of saving, while the full view — with 12% electricity inflation, 6% fuel inflation and 5% wage growth compounded — credits roughly R287,000, well past the R120,000 cost.
That is the case for running your own numbers rather than trusting a rule of thumb. Your shed hours, your hourly output, your fuel burn and your tariff are all different from the examples here. The Solar & Backup ROI tool lets you put a rand value on your downtime, set your own electricity and fuel inflation assumptions, and compare an inverter, a battery and a generator on payback rather than sticker price. Price the hour honestly, then let the tool tell you which backup actually earns its keep — open it and run your own scenario.
Run the numbers yourself
Open the Solar & Backup ROI tool to price your own load-shedding hours
Open the toolFAQ
What does one hour of load-shedding actually cost a household in 2026?
For a household with someone working from home, a shed hour that lands in working time costs roughly R141 on battery or inverter power — mostly lost productivity (about R125) plus a little expected food spoilage. On a petrol generator it rises to about R203 an hour once you add roughly R47 of fuel and R15 of service wear. The bill saving from not drawing grid power is trivial by comparison, which is why pricing the hour by what it interrupts matters far more than the tariff.
Is it cheaper to run a generator or a battery during load-shedding?
For short, frequent outages a battery wins easily, because its running cost is close to zero while a generator burns about R47 an hour (home petrol) to R62 an hour (café diesel) plus service wear. A generator only pulls ahead for rare, very long outages where a battery would run flat. Since most of a 2026 shedding year is made of short slots, the cost-efficient setup is usually solar-plus-battery with a generator kept only as a deep-outage backstop.
How does pricing downtime change a solar payback calculation?
On the electricity bill alone, a R120,000 hybrid system saving about R16,800 a year takes roughly 7.1 years to pay back. Once you add the value of recovered work output (around R24,000 a year) and avoided generator fuel (about R8,500), total annual benefit jumps to about R49,000 and payback falls to roughly 2.4 years. The hardware is identical; only the honesty of the inputs changed, which is why ignoring downtime makes solar look far worse than it is.
Why does electricity inflation matter more than fuel inflation for backup decisions?
Electricity tariffs have been climbing at double-digit rates — model around 12% a year — while fuel inflates slower and lumpier at roughly 6%. A generator's cost is chained to fuel and paid in cash every hour forever, whereas solar locks in today's hardware price once and then rides the 12% tariff escalation as growing savings. The faster electricity inflates relative to fuel, the more the long-run answer tilts from generator towards solar plus storage.