Cape Town water planners map the next dry season

The reservoir level that mattered most this year was not the lowest one. It was the one recorded in early autumn, three months before the dry season proper, when the city’s planners look at how much is stored and decide what to tell people.

Cape Town has been unusually good at that conversation since 2018. What has changed is that the conversation now begins from a model rather than from last year’s numbers.

Forecasting instead of reacting

The old planning cycle worked backwards from a fixed target: hold storage above a threshold, and impose restrictions when it drops. It was legible and it was late, because storage is a lagging indicator. By the time levels fall enough to trigger a tier, the rainfall that would have refilled them has already failed to arrive.

The new cycle runs on a seasonal forecast, updated fortnightly, that estimates the range of storage outcomes for the following six months. Restrictions are keyed to the probability of hitting a floor, not to the current level.

Waiting for the number to drop is waiting for the decision to be made for you.

The practical effect is that restrictions start earlier and lighter. That is politically harder — asking people to save water while the dams look adequate requires explaining a probability — and hydrologically far cheaper.

The leak that is not a leak

Roughly a fifth of the water entering the distribution network never reaches a billed connection. Some of it is genuine loss from ageing pipes; a significant share is metering error, unbilled municipal use, and connections nobody has records for.

Distinguishing between those matters, because they have opposite remedies. Physical loss is fixed with capital: replace mains, manage pressure, and accept a decade-long programme. Apparent loss is fixed with administration: audit meters, reconcile the connection register, and bill what is being used.

The city has been running district metering across sixty zones to tell them apart. The early result is unwelcome and useful — apparent loss is a larger share than assumed, which means part of the “leakage” problem is a records problem that no amount of pipe replacement will solve.

Agriculture and the allocation question

The catchment supplies farms as well as households, and the allocation between them is set by an agreement that predates the current climate record. In a normal year the tension is manageable. In a dry year the agreement allocates a share of a total that no longer exists.

Both sides have an interest in renegotiating before the next shortage rather than during one, which is exactly when nobody wants to. Talks have started, quietly, on a formula that shares proportional reductions instead of fixed volumes.

What to watch

Watch the fortnightly forecast bulletins and whether they keep being published when the outlook is good, since a forecast that only appears in bad years trains people to ignore it. Watch the split between physical and apparent loss as more zones report. And watch whether the allocation talks produce a proportional formula before the next below-average rainfall season, because after it they will not.