Industrial Rainwater Harvesting: Yield Modelling and Storage Sizing

How much water can an industrial roof actually yield?

The starting point for any feasibility study is the annual harvestable volume. It is governed by the catchment collection equation, which nets rainfall depth against the losses from surface wetting, evaporation, splash and pre-tank filtration:

Use the plan area of the roof, not the sloped area: rain falls vertically, so it is the horizontal projection that intercepts it. For UK sites, R comes from a local long-run annual average (roughly 550–650 mm in the drier south-east, over 1,200 mm in the west and uplands). Sizing on the annual mean gives the resource; sizing storage, however, needs the monthly or daily distribution, because supply and demand rarely coincide.

What runoff coefficient should you use?

The runoff (or yield) coefficient C_r is the fraction of incident rainfall that reaches the tank after depression storage, wetting and evaporative losses on the catchment itself. Hard, impervious, steeply pitched roofs shed the most; textured, flat or green surfaces retain more. The values below are representative design figures consistent with BS 8515 and standard rainwater practice.

Paved yards and hardstanding can multiply the catchment area but carry hydrocarbons, silt and tyre wear, so they demand more pre-treatment and are usually reserved for the least sensitive end-uses such as dust suppression or irrigation.

Worked example: annual yield from an industrial roof

Take a distribution warehouse in the English Midlands with a profiled steel roof of plan area A = 6,000 m², a long-run annual rainfall of R = 720 mm, a runoff coefficient of C_r = 0.85 and a pre-tank filter efficiency of η = 0.90. The harvestable volume falls out of one calculation:

• Gross incident rainfall: A × R = 6,000 m² × 720 mm = 4,320,000 L = 4,320 m³/yr.
• Apply runoff coefficient: × 0.85 = 3,672 m³/yr.
• Apply filter efficiency: × 0.90 = 3,305 m³/yr reaching the tank.

That is about 3.3 ML/yr, or roughly 9,000 L/day averaged over the year. If the site's non-potable demand (washdown, WC flushing, cooling make-up) is of the same order, the resource is well matched and the project is worth engineering in detail. This yield figure is the ceiling on supply; how much you can actually use depends on storage, which we size next.

How do you size the storage tank?

Storage is not sized on annual totals but on the demand/supply ratio and the temporal mismatch between rainfall and use. BS 8515 sets out two approaches: a simple percentage-of-demand intermediate method, and a detailed mass-balance (behavioural) simulation that routes a time series of rainfall and demand through the tank. The behavioural model runs a daily water balance:

Running the simulation over a multi-year rainfall record returns the water saving efficiency (fraction of demand met from rainwater) as a function of tank size. Efficiency rises steeply with the first increments of storage, then flattens; the knee of that curve is the economic optimum. As a robust first pass, BS 8515's intermediate method sizes the tank at the smaller of 5% of annual yield or 5% of annual non-potable demand.

Continuing the worked example: 5% of the 3,305 m³/yr yield is 165 m³, while 5% of a comparable 3,285 m³/yr demand is 164 m³ — so a tank of order 160–170 m³ (about 18 days' average demand) is the starting point, to be refined by the daily behavioural model. A detailed harvesting system design and mass-balance study will confirm the knee-point capacity for your specific rainfall record and demand profile.

Which industrial end-uses suit harvested rainwater?

Rainwater is soft, low in dissolved solids and near-neutral, which makes it attractive for several non-potable duties — but each end-use sets its own water-quality bar and therefore its own treatment train.

Low hardness is the standout benefit for evaporative cooling: softer make-up allows higher cycles of concentration before scaling, cutting both blowdown volume and chemical dosing. Rainwater harvesting therefore dovetails with wider industrial water reuse schemes rather than competing with them.

Why does first-flush diversion matter for water quality?

The first few millimetres of runoff carry the highest load of dust, bird droppings, pollen, atmospheric deposition and microbial contamination washed off the catchment between storms. Diverting this first flush away from the tank sharply improves stored-water quality with negligible loss of yield. A common design allowance is 0.4–1.0 mm of rainfall per m² of catchment (roughly 2 L/m² for the dirtiest urban roofs), sacrificed on each event.

For the 6,000 m² roof, a 0.5 mm first-flush diversion sheds 6,000 × 0.5 = 3,000 L per rain event before clean water is admitted — a small annual fraction, already partly captured in the η term. Downstream, keep the tank cool and dark to suppress biofilm, screen the inlet, and add UV or dosing where water is stored more than a few days or used where aerosols form (cooling towers). These measures manage Legionella and general microbial risk; harvested rainwater is a non-potable resource and should never be cross-connected to potable mains without an air gap and backflow protection.

What is the payback and water-neutrality contribution?

The economic case is the avoided cost of mains supply and trade-effluent volume, set against capital and pumping/treatment cost. Where harvested water displaces mains that would otherwise be discharged, both the incoming tariff and the volumetric part of a trade-effluent consent charge are avoided.

For the worked example, 3,305 m³/yr displaced at a combined supply-plus-discharge value of about £3.50/m³ saves roughly £11,600/yr. Against an indicative installed cost of £90–120k for tank, filtration, pumps and controls, that is a simple payback near 8–10 years — longer than many reuse projects, but the tank asset life exceeds 25 years and the saving is inflation-linked to rising water tariffs. Beyond cash, the scheme contributes directly to corporate water-neutrality and reduced abstraction targets, quantified as the harvested fraction of site demand. Combine it with process reuse to shrink the overall industrial water footprint, and the blended payback improves.

Sizing an industrial rainwater harvesting system

1. Quantify the catchment. Measure the plan (projected) roof area and assign a runoff coefficient by surface type (0.5–0.9).
2. Obtain local rainfall data. Use a long-run annual average for the resource and a monthly or daily time series for storage sizing.
3. Calculate harvestable yield. Apply V = A × R × Cr × η to get the annual volume reaching the tank after filter losses.
4. Profile the non-potable demand. Sum washdown, WC, cooling make-up and irrigation demand, and compute the demand/supply ratio.
5. Size storage by mass balance. Run a daily behavioural (YAS) simulation and pick the tank capacity at the knee of the efficiency curve, or use BS 8515 5% intermediate method as a first pass.
6. Specify treatment and safeguards. Add first-flush diversion, filtration, disinfection where needed, and backflow protection to any mains top-up.