EDGE Expert study materials
EDGE Expert study resources

Rainwater in a 270 mm City

EDGE Expert · Put your knowledge to work in context.

All case studies

Rainwater in a 270 mm City

A 200-room hotel in Amman, Jordan, runs at 65% occupancy with water demand of about 300 litres per guest-night including laundry, kitchens and public areas; there are also 800 m² of irrigated lawn. Mean annual rainfall is roughly 270 mm and the roof area is 2,500 m². The consultant's water package is: rainwater harvesting from the full 2,500 m² roof, conversion of the lawn to native drought-tolerant planting, low-flow fixtures throughout, and a greywater system recycling shower and sink water to the toilets. The client wants the cheapest package that reliably clears the 20% water threshold, and is nervous about maintenance burden in a building he will hand to an operator.

1. Size the annual demand and estimate what each of the four measures actually contributes.

Read the worked answer

Demand: 200 rooms × 0.65 occupancy × 300 L × 365 days ≈ 14.2 million litres, i.e. about 14,200 m³/yr before irrigation.

Rainwater: 2,500 m² × 0.27 m × ~0.85 runoff coefficient ≈ 570 m³/yr - about 4% of demand. And it arrives in a handful of winter storms; a tank sized to capture the whole annual yield would be enormous, so practical yield is lower still.

Irrigation: 800 m² of lawn at roughly 900 mm/yr ≈ 720 m³. Converting to native planting with drip cuts perhaps 60–75%, so ~450–550 m³ - about 3–4%.

Low-flow fixtures: toilets, showers and taps dominate guestroom and public-area use. A well-chosen set can cut 30–40% of indoor demand, i.e. several thousand m³ - roughly 25–30% of total.

Greywater: shower and sink water recycled to flushing might recover 20–25%, but only with dual plumbing, treatment, and an operator who maintains it.

The headline lesson: the measure the client is most likely to think of as 'the green one' - rainwater - is a rounding error in this climate, while the unglamorous fixtures do the work.

2. Rank the measures by cost per cubic metre and by robustness, and propose the package you would take to the client.

Read the worked answer

Ranking:

  1. Low-flow fixtures - lowest capital cost, zero operational burden, performance independent of weather, and auditable from a specification. First and non-negotiable.
  2. Xeriscaping with drip irrigation - modest one-off cost, no new O&M beyond gardening, and it removes a demand rather than adding a system.
  3. Rainwater harvesting - moderate capital cost, climate-dependent, small yield. Worth including for resilience and for the story it tells guests, but it cannot be the basis of the claim.
  4. Greywater recycling - highest capital cost, real O&M burden, operator-dependent, and needs health and plumbing approvals. Last, and only if the operator contractually commits to maintenance.

Recommended package: fixtures + xeriscaping + rainwater, modelled to a comfortable margin above 20%, with greywater value-engineered out unless the operator signs up to it. The client asked for the cheapest reliable package; reliability is the operative word, and greywater is the least reliable element in a building that changes hands.

3. What would you build into the submission to protect the 20% claim over the life of the asset?

Read the worked answer

Three protections:

  • Model a margin. Target 25–30% modelled savings rather than 20.5%. Later value engineering, a substituted fitting, or a dry year should not be able to push the delivered building under the threshold.
  • Specify by numbers, not by label. Low-flow fixtures must be specified with flow rates in L/min at a stated pressure, toilets in L/flush. A manufacturer's 'water-saving' badge cannot be credited by an auditor; a data sheet can. The same applies to irrigation: emitter rates, controller type, and evidence that the planting list is genuinely drought-tolerant.
  • Design the rainwater system for the dry season, and say so. Tank volume should be stated against the dry-spell length, and the modelled yield should be conservative. If the tank is sized for a full annual capture the client will pay for storage that sits empty, and the modelled yield will never be delivered.

Finally, brief the operator in writing on the maintenance of whatever is installed - a system that is credited on paper but abandoned in year two is a reputational liability, not just a data problem.