Collecting rainwater is the oldest water technology there is, and it remains among the most sensible. Rain falls on your roof for free, it is soft and unchlorinated, plants prefer it, and using it displaces treated drinking water that was purified at considerable energy cost to flush a toilet or water a lawn.
The question is how far up the scale to go, because the options run from a twenty-pound butt to a buried tank and a pumped supply, and the economics change sharply along the way.
How much water is actually available
The arithmetic is simple and the answer surprises people.
Roof area in square metres, times annual rainfall in millimetres, gives litres per year, before losses. Multiply by about 0.8 to allow for evaporation, splash and first-flush diversion.
So a modest 60 m² roof plan area in a region receiving 800 mm of rain a year intercepts around 48,000 litres, of which perhaps 38,000 is realistically collectable. That is a substantial fraction of a household's non-drinking demand.
Note that roof plan area is what matters, not the sloped surface, since rain falls vertically. And rainfall is seasonal: in northern Europe most of it arrives when the garden least needs it, which is why storage volume matters more than collection area for garden use.
The simple end: butts, done properly
A water butt is unglamorous and the best value in this article. A few details separate a useful one from a green sludge bin.
Size it against a dry spell, not average rainfall. A single 200-litre butt supplies a garden for a few days in July. Linking two or three, or using a larger tank, is what gets you through a fortnight without rain.
Link butts at the bottom, not the top, so they fill and draw as one volume.
Keep it covered and dark. Light grows algae, and an open butt is a mosquito nursery and a genuine drowning risk to small children. A lid is not optional.
Fit a leaf filter or a diverter at the downpipe, because organic debris is what makes stored water smell.
Raise it enough to get a watering can under the tap, and accept that gravity alone gives very little pressure — not enough for drip irrigation emitters to work evenly, which is why a small pump is usually needed for anything automated.
Provide an overflow that goes somewhere sensible, back to the drain or into a border, rather than against a wall.
The serious end: harvesting systems
A full rainwater harvesting system uses a large tank, usually buried, feeding toilets, washing machines and outside taps through a pumped and filtered distribution.
What it involves, in outline: a filter before the tank; a calmed inlet so incoming water does not stir up settled sediment; a floating suction drawing from just below the surface where water is cleanest; a pump; and a mains top-up arrangement for dry periods.
The critical safety requirement is backflow prevention. Rainwater must never be able to enter the drinking water system, and top-up from the mains must be arranged so there is no direct connection — usually an air gap into a break tank. This is a legal requirement in most countries, not a refinement, and it is one of several reasons this is regulated plumbing work.
Also required: clear labelling of all rainwater pipework and outlets, so that nobody in future plumbs a drinking tap onto it. Typically that means distinctly coloured pipe and signage at every outlet.
Honest economics: payback on a full system is long, often decades at domestic water prices, and the case is stronger where water is metered and expensive, where a large garden or several toilets create real demand, or where it is being installed during a new build rather than retrofitted. If the motivation is environmental rather than financial, that is a perfectly good reason — just go in with the right expectation.
Greywater, and why we are cautious
Greywater is used water from showers, baths and basins — not toilets, which is blackwater, and generally not kitchen sinks, which carry fat and food.
It is attractive in principle because it arrives year-round in proportion to occupancy, unlike rainfall. In practice, domestic greywater reuse is considerably harder than rainwater harvesting.
Greywater contains skin cells, hair, soap and organic matter, and bacterial populations grow rapidly in it — within hours at room temperature. That means it cannot simply be stored: systems either use it almost immediately, or treat it, which means filtration and disinfection with ongoing maintenance. Neglected greywater systems smell and can present a genuine health risk.
Where it works well: immediate reuse, such as diverting bath water directly to a garden the same day, or well-engineered systems in larger buildings with proper maintenance regimes. Where it works badly: stored, untreated, in a domestic setting, forgotten about.
Regulation is also stricter than for rainwater in many countries, with rules about where treated greywater may be used and how it must be labelled. Check before designing anything.
Where automation genuinely helps
Tank level monitoring. An ultrasonic or pressure sensor reporting how much you have left, which turns a guess into a number and lets everything else make sensible decisions.
Preferring stored water and falling back to mains. The core rule of a harvesting system: irrigate from the tank while there is water, switch to mains below a threshold, and notify.
Pre-emptive draw-down before heavy rain. A neat trick where the tank also serves an attenuation purpose: if a large storm is forecast and the tank is nearly full, irrigating or discharging beforehand creates capacity, reduces overflow, and in some areas contributes to local flood management.
Pump protection. A dry-run cut-out, and an alert if the pump cycles unusually often, which indicates a leak or a failing pressure vessel.
Overflow detection. A sensor at the overflow tells you the filter has blocked, which is the commonest maintenance failure.
Winterisation reminders. A butt left full and frozen splits, and an outdoor pump must be drained. A calendar prompt tied to a temperature forecast is more reliable than memory.
Usage accounting. Metering harvested water alongside mains tells you what the system is actually displacing, which is the only honest way to judge whether it was worth installing.
Where to start
- Do the arithmetic: roof plan area times local rainfall times 0.8.
- Start with butts, linked at the bottom, covered, filtered, with a sensible overflow.
- Add a small pump if you want drip irrigation from stored water.
- Fit a level sensor and build the prefer-stored-then-mains rule.
- Only consider a full system with a large garden, several toilets, metered water, or a build already underway.
- Get backflow prevention and labelling right, professionally, if you go that far.
- Be cautious about greywater unless you will maintain it, and never store it untreated.
- Meter what you displace, so you know whether it worked.
Steps two and three cost very little and cover the majority of the practical benefit. Everything beyond them is a bigger project with a longer payback and a stronger environmental than financial case.