Load Shedding, Explained
Why the lights go out on a schedule instead of at random, why it happens to everyone at once rather than to whoever plugged in last, and why "just build more power stations" is a longer sentence than it sounds.
A grid is one shared pool
The single most misunderstood thing about electricity is that it is not stored anywhere in the system. It is generated and consumed in the same instant. Whatever the country is drawing right now is being produced right now, and the two have to match continuously — not on average over the day, but from second to second.
What holds them together is frequency. The whole grid turns at 50 Hz, and every synchronised generator on it is physically locked to that rhythm. Draw more than is being produced and the frequency sags, because the load is literally braking the turbines. Let it sag far enough and protection equipment starts disconnecting generators to save them from damage — which removes supply, which drops the frequency further, which trips more generators.
That runaway is a national blackout, and recovering from one is not a matter of switching things back on. A grid restarted from nothing has to be rebuilt section by section from stations capable of starting without external power, matching frequency and phase at every step. Estimates for South Africa run to two weeks.
So load shedding is the alternative to that
Load shedding is a controlled, deliberate reduction in demand to keep supply and demand matched. It is not a fault. It is the mechanism that prevents the fault. When there is not enough generation available, the system operator sheds blocks of load in a rotating, published schedule — which is why you get a timetable rather than a surprise.
The stages are simply how much has to go:
| Stage | Load removed nationally |
|---|---|
| Stage 1 | 1 000 MW |
| Stage 2 | 2 000 MW |
| Stage 4 | 4 000 MW |
| Stage 6 | 6 000 MW |
| Stage 8 | 8 000 MW |
Each stage is a further 1 000 MW off the national load — roughly the output of a large power station unit, or the household demand of a small city. The numbering is not a severity scale invented for the public; it is arithmetic.
Why it hits everyone, not just the newest suburb
This is the part that feels unjust and is unavoidable. Electricity does not travel from a particular station to a particular house. It goes into a shared pool and comes out wherever it is drawn. There is no meaningful sense in which your suburb is "supplied by" one plant.
So a shortfall is national, and it is shared. A city that has done everything right is shed alongside one that has not, and a factory that has paid every bill on time goes dark alongside one that has not. The pool has no memory of who contributed what.
Why building more stations is slow
Generation cannot be conjured, and the lead times are the real constraint:
- Solar: 8–14 months from decision to first power. Fastest by a wide margin — but it produces nothing at 7pm, which is exactly when the national peak occurs.
- Coal: 6–10 years, and Medupi and Kusile both ran vastly over both time and budget.
- Nuclear: 8–15 years. Whatever is decided today arrives in the 2040s.
And then there is the part that catches people out completely: generation is useless where there is no transmission to carry it. The Northern Cape has world-class solar resource and has had projects held up waiting for lines to reach them. A power station connected to nothing supplies nobody, however many megawatts it is rated at.
There is also a maintenance dimension that gets less attention than it deserves. A coal fleet running flat out to avoid shedding is a fleet not being maintained, and deferred maintenance produces more unplanned breakdowns, which forces more shedding. The Energy Availability Factor — the share of installed capacity actually available on a given day — is the number that really matters, and it has spent years well below where the fleet was designed to sit.
What it costs a city
Beyond the obvious inconvenience, the compounding costs land on the things a city is built out of. Industry cannot run shifts it cannot power. Cold chains break. Water pumping stops, so reservoirs run down and high-lying suburbs lose pressure hours after the electricity comes back. Traffic lights fail. Security systems fail. Every business that can afford a generator buys diesel, and every business that cannot, absorbs the loss.
The deepest cost is investment that never happens. Growth requires someone to believe the lights will be on in five years, and that belief is what an unreliable grid actually destroys.
How BraaiVille models it
The game does not simulate stages or schedules — it does something blunter and more instructive. When demand passes supply, the entire city stops growing. Not the newest suburb: everything, everywhere. No zone develops a step further anywhere on the map until you add capacity.
That is a fair model of the real relationship. A constrained grid does not slow one neighbourhood down; it caps what the whole economy can do. And because the game makes you connect each station to a consumer before its capacity counts, you get the second lesson for free: building generation and delivering electricity are two different projects, and only one of them turns the lights on.
Watch the ⚡ Power figure in the top bar. It shows connected load against connected capacity, turns amber at 85% and red when you are over. The amber is the part worth acting on — by the time it is red, the city stopped growing some months ago.
Play BraaiVille · Coal, nuclear and solar compared · The Generation Information Panel