Size a Rainwater Tank Around Your Driest Spell

Why the dry spell, not the annual average, sizes the tank

Annual rainfall totals flatter every tank. A region can collect a healthy yearly figure and still run eight dry weeks in the middle of it. Water does not arrive evenly, and demand does not pause when it stops arriving. A tank sized on averages will be full in the week nobody needed it and empty in the week everybody did.

The dry spell is the honest unit. It matches the exact window where the household lives from stored water alone. Your longest recent spell, or the one the local bureau publishes as typical, is the right number to enter here.

One honest caveat cuts the other way. Climate shifts have stretched dry spells in many regions beyond the published history. If your local records feel outdated, size against the longest spell you have personally lived through, not the tidy average in a table.

Storage rounds the answer too. Tanks ship in standard sizes, and the next one up costs less than a custom volume. That retail rounding is the final quiet adjustment between the maths and the delivery truck.

Worked example: a three-week spell, step by step

Take a household using 150 litres a day of non-drinking water, facing a 21-day dry spell. Raw demand is 150 times 21, which is 3,150 litres. Now the step most people skip: four days inside that spell still bring useful rain, and the roof catches about 300 litres on such a day. That is 1,200 litres of in-spell harvest.

Subtract it, and the net gap is 1,950 litres. Add 20% headroom for a hot week or a dirty first flush, and the target becomes 2,340 litres. The calculator rounds that up to a 2,500-litre tank, a size most suppliers stock as a standard product.

Skip the rain subtraction and you would have bought 3,800 litres of tank for a 1,950-litre problem. The subtraction is where tank sizing stops being folklore. It is also why a rain gauge and a month of observation beat any regional table.

Feeding the harvest input from your real roof

The one rainy day figure looks vague until you anchor it. A roof catches about one litre per square metre per millimetre of rain. A 60 m² roof under 5 mm of rain yields roughly 300 litres before losses. Multiply your roof area by the millimetres of a typical useful shower, and you have entered the honest version of that input.

Capture efficiency lives inside that same number. Leaves, first-flush diverters and tank evaporation realistically cost a tenth to a fifth. If your roof is overhung by trees, trim the figure rather than trusting the clean calculation.

If you have no idea what a useful shower looks like locally, a cheap rain gauge settles it in a season. The same gauge also protects you from the opposite error, which is assuming in-spell rain that never actually comes.

One more habit protects the harvest number. Note the month the tank first runs low each summer. After two seasons you can compare that date with your entered spell length. The plan either proves itself or shows which input was wishful.

When the maths says a smaller tank than you expected

People routinely overbuy tanks, because stored water feels virtuous. The maths argues otherwise. Every litre of tank capacity is money, footprint and structural load, whether it gets used or not. A tank that sits half full for months goes stale, invites mosquitoes and does nothing productive for its surplus.

A smaller correct tank plus a planned top-up habit beats a large careless one. That habit has real forms: running the washing machine only on full loads, garden watering moved to early morning, and a conscious decision to accept shorter coverage during a record spell.

The counter-case is genuine too. If your supply has total-failure risk, if pumps depend on grid power you lose in storms, or if a remote household cannot buy water quickly, that is a resilience argument for extra headroom. It is a deliberate choice, not an averaging error.

Hot-tub thinking also inflates demand. Households that plan to refill pools or run pressure washers from the tank should treat those litres as a separate line. They arrive in bursts that a weekly indoor average never predicts.

Accuracy, assumptions and limitations

This model is deliberately simple. It assumes steady daily use across the spell, uniform-ish luck in where the in-spell rain lands, and a tank that starts the spell full. It does not model seasonal peaks, evaporative loss in heat, or the slow leak you have not found yet.

It also assumes your non-drinking figure is measured rather than imagined. Households guess wildly here. A metered month of usage, or at least a count of toilet flushes and laundry loads, will move this input more than any other, and it costs an afternoon to obtain.

Treat the rounded-up retail size as the floor of your decision, not a ceiling. If two sizes sit close, walk the actual site: the truck route, the crane reach, the foundation and the bypass pump-in day will outvote the spreadsheet more often than you would expect.

Finally, the model prices nothing but water volume. Tank cost, pump power and council approval are inputs that live elsewhere. The one number that must be right is the gap between rain events. That gap is exactly what the dry spell input asks you to know honestly.