Rainwater Harvesting Sized to a Drought That the Code Does Not Require
A 120,000 ft² distribution warehouse in a semi-arid region (annual rainfall ~14 in., high summer evaporation) is pursuing a water-efficiency credit. The design team proposes a 150,000-gallon rainwater cistern to supply toilet flushing and irrigation, sized to the largest single storm on record, with no makeup water connection. The site has 60,000 ft² of roof and 40,000 ft² of landscape (mostly native, low-water). Occupancy is 200 people, 250 days/year. The owner wants the credit but is worried about running dry in summer and about the first cost of the tank. Local code requires only that harvested water be used for non-potable applications and does not mandate a specific storage volume.
1. Evaluate whether the proposed cistern sizing is technically defensible, and identify the flaw in sizing to the largest storm.
Read the worked answer
Sizing to the largest single storm is the wrong basis and is not defensible. A cistern sized to a single extreme event optimizes for capture of a rare storm while ignoring the seasonal supply-demand balance that determines whether the system ever runs dry. In a semi-arid climate with ~14 in./yr and high summer evaporation, the binding constraint is the dry-season demand, not peak storm capture.
Rough check: 60,000 ft² roof × 14 in./yr × ~0.6 runoff coefficient ≈ 42,000 ft³/yr ≈ ~314,000 gal/yr of collectible supply before losses. Toilet demand for 200 people at, say, 1.28 gal/flush and ~3 flushes/day over 250 days is modest, but irrigation of 40,000 ft² - even native/low-water - plus summer evaporation losses can exceed dry-season rainfall. A 150,000-gal tank sized to a record storm may be over-built for capture and under-protected for the dry season.
The flaw: no makeup connection plus storm-based sizing means a single dry summer empties the tank and the building loses flushing water - a reliability failure the credit does not excuse. The tank volume should come from a monthly mass-balance simulation (supply = rainfall × roof area × runoff coefficient; demand = monthly toilet + irrigation; storage = buffer between surplus and deficit months), not from a design storm.
2. Recommend a defensible sizing and reliability approach, and state how you would verify the credit.
Read the worked answer
Sizing approach - monthly mass balance, not design storm:
- Build a 12-month (or multi-year) water balance: monthly rainfall capture vs. monthly non-potable demand (toilet + irrigation), with evaporation and first-flush losses deducted.
- Size the cistern to the maximum cumulative deficit between surplus and deficit months - the volume that carries the system through the driest stretch - not to the largest storm.
- Right-size demand first: specify low-flow fixtures and confirm the landscape is genuinely low-water; reducing demand shrinks the required tank and improves first cost.
- Add a makeup/reliability path: a potable makeup connection (with appropriate backflow protection) or a municipal top-up for the dry season, so a dry summer does not shut down flushing. A no-makeup design is a reliability risk the owner correctly flagged.
- Consider staged/right-sized storage - a smaller tank plus makeup is often lower first cost and more robust than an oversized tank.
Verification:
- Confirm the mass-balance model is documented and the tank equals the computed maximum deficit.
- Meter harvested-water supply and non-potable demand separately; trend over a full year to confirm the tank neither overflows chronically (wasted capture) nor empties (reliability failure).
- Confirm the credit's actual requirement (e.g., percentage of non-potable demand met) is satisfied by the modeled and metered performance, and that code non-potable-use restrictions are met.