Every number on this platform can be traced back to a published source.
Aquiterum is a pre-feasibility screen, not an engineering study. That makes transparency the whole product: the coefficients, the evidence grades, the peer bands and the literature they came from are all set out below, so an engineer or sustainability director can check our working before they trust a single figure.
The engine is deterministic: identical answers always produce an identical result. Every saved snapshot records the engine and knowledge-base versions that produced it, so an old result can be reproduced exactly, or re-run against today's rules and the difference inspected.
The chain
How a figure is produced
Five steps, in order, with no discretionary adjustment anywhere in between.
Your answers
Twelve adaptive questions describe the property: type, keys, occupancy, climate, cooling system, pools, landscape, laundry, kitchen and any metered data you can supply.
Published coefficients
Each answer is matched to a band drawn from a named public dataset, standard or peer-reviewed study. Nothing is invented, and nothing comes from a vendor.
Range arithmetic
Low, likely and high values are carried end to end. Where uncertainty compounds, the band widens - it is never collapsed into a single flattering number.
Evidence grading
Every figure is tagged measured, reported, modelled, benchmark or unknown, so you can see at a glance which parts of the picture rest on your data and which rest on ours.
Confidence scoring
Grades are weighted into a single confidence score for the snapshot. Supplying meter readings raises it; gaps are shown as gaps rather than filled with assumptions.
Provenance
The evidence grades
Each grade carries a fixed weight in the confidence score. These definitions and weights are read directly from the engine, so this page and the calculation can never disagree.
Measured
weight 1.00Taken from a meter reading or utility invoice supplied by the property.
Reported
weight 0.75Stated by the property. Plausible, but not yet evidenced by a document.
Modelled
weight 0.60Calculated by the Aquiterum engine from other inputs and published coefficients.
Benchmark
weight 0.40Drawn from a published benchmark for comparable properties.
Unknown
weight 0.00Not enough information. Shown as a gap rather than an estimate.
How the grades become one confidence score
The score blends two things. First, the weighted grade of your water balance - counted twice, because everything else is built on it - averaged with the grades of your three largest opportunities. Second, a penalty for how wide the resulting bands are: a well-graded answer that still spans a factor of three is not a confident answer. The result is capped between 8 and 95, because no screening exercise deserves certainty.
Above 70 the snapshot is good enough to brief a specialist. Between 45 and 70 it is directional - worth investigating, not yet worth quoting. Below 45 every figure should be treated as a hypothesis. Supplying meter readings or invoices is the fastest way to move up.
The model
The water-loop model
Every Aquiterum output rests on one model: demand is split by where water is used, outflows are identified as recoverable streams, and each stream is matched only to the end uses it may legitimately serve.
The loop in three moves
Demand
Total consumption is split into where the water is actually used — guest rooms, laundry, kitchen, pools, landscape and back of house — using published intensities for your property type and climate.
Streams
Each use produces an outflow. Six of those outflows are recoverable in a hospitality building; the engine estimates how much of each your property produces.
Matching
Each stream is matched only to end uses it may legitimately serve, at the treatment class that pairing requires. Nothing is matched on volume alone.
A stream is only ever credited once. Where two end uses compete for the same water, the engine allocates to the better-matched use and shows the remainder as unallocated rather than counting it twice.
The recoverable streams
Air-conditioning condensate
CondensateWater condensed out of humid air by cooling coils in fan-coil units, air handling units and split systems, then usually discharged to drain.
Why it is overlooked: It is produced continuously, distributed across the whole building, and every drain run looks like a nuisance rather than a resource.
- Retrofit complexity
- Moderate — some plant and pipework
- Capital category
- Small capital project
- Minimum treatment
- Filtration
Peaks exactly when cooling load and irrigation demand peak — one of the few streams naturally in phase with summer demand.
Condensate is low in dissolved solids but can pick up metals, biofilm and Legionella risk from drain pans and pipework. It is not potable and must not be cross-connected with the potable network.
Permitted end uses, best match first
Near-distilled quality makes condensate an unusually good cooling-tower make-up, reducing both make-up volume and blowdown frequency.
Source: Water Resources Management (Springer) — open access 2025, US Environmental Protection Agency 2012
Low mineral content suits most planting, though it is slightly acidic and nutrient-free; blending or dosing may be advised by a horticulturalist.
Source: Sustainability (MDPI) — Magrini, Cattani, Cartesegna, Magnani 2017, Journal of Physics: Conference Series — Khan et al. 2026
Low hardness means fewer deposits on glass and hard surfaces.
Technically straightforward but requires a separate marked network, so it is usually only viable alongside a wider non-potable scheme.
Source: Sustainability (MDPI) — Magrini, Cattani, Cartesegna, Magnani 2017, Journal of Physics: Conference Series — Khan et al. 2026, Water Resources Management (Springer) — open access 2025
Rainwater harvesting
RainwaterRainfall collected from roofs, terraces and hard standing, stored, and used for non-potable duties.
Why it is overlooked: In seasonal Mediterranean properties most rain falls when the hotel is closed, so annual rainfall totals badly overstate usable yield.
- Retrofit complexity
- Moderate — some plant and pipework
- Capital category
- Medium capital project
- Minimum treatment
- Filtration
Usually out of phase with demand in dry-summer climates. Storage volume, not roof area, is normally the binding constraint.
First-flush diversion and covered storage are required. Roof materials, bird fouling and stored-water stagnation all affect quality; local rules on rainwater use vary widely.
Permitted end uses, best match first
The classic pairing — but in dry-summer climates storage volume, not roof area, decides whether it works.
Lowest-barrier use and a sensible first phase while a larger scheme is evaluated.
Widely permitted and well understood; needs a marked second network and backflow prevention.
Naturally soft, so it can cut detergent dosing — but supply reliability must be proven before the laundry depends on it.
Acceptable where public contact is incidental; stagnation and aerosol control still apply.
Source: World Health Organization 2006
Greywater recycling
GreywaterLightly-soiled water from showers, baths and washbasins, treated and reused for non-potable duties.
Why it is overlooked: It is by far the largest recoverable stream in a hotel, but in an existing building the drainage is already combined with WC waste — separating it is the real project.
- Retrofit complexity
- Very high — structural or full second-pipe network
- Capital category
- Major capital project
- Minimum treatment
- Biological treatment and disinfection
Available whenever the property is occupied; closely tracks occupancy.
Health-critical. Requires treatment to a defined quality class, continuous monitoring, marked separate pipework, backflow prevention and — in the EU — a documented risk-management plan. Never proceed without a qualified engineer and the local regulator.
Permitted end uses, best match first
Largest single non-potable substitution available in a hotel, and demand matches supply almost hour for hour.
Source: World Health Organization 2006, US Environmental Protection Agency 2012
In the EU, reclaimed water for irrigation falls under Regulation 2020/741 quality classes and a mandatory risk-management plan. Sub-surface drip lowers exposure.
Possible, but nutrient and organic load raise biofouling and Legionella risk in the tower. Specialist water-treatment input is mandatory.
Rarely accepted for linen in hospitality because of hygiene expectations and machine warranties.
Source: World Health Organization 2006
Source: World Health Organization 2006, European Union 2020, US Environmental Protection Agency 2012
Pool filter backwash recovery
Pool backwashWater discharged when pool sand filters are backwashed, plus pool drain-down at season change.
Why it is overlooked: It leaves in short, large slugs on a maintenance schedule nobody logs, so it never appears in a water balance.
- Retrofit complexity
- Low — largely surface work
- Capital category
- Small capital project
- Minimum treatment
- Filtration and disinfection
Weekly through the operating season, in discrete batches.
Carries chlorine or bromine residual, high solids and sometimes cyanuric acid or salt. Dechlorination and salinity checks are essential before any irrigation use — salt-chlorinated pools are frequently unsuitable.
Permitted end uses, best match first
Simplest defensible use, and often the fastest way to demonstrate a working loop on site.
Viable only after dechlorination and a salinity check — salt-chlorinated pools usually rule this out.
Backwash-to-pool recovery units exist but are pool-plant engineering decisions governed by bathing-water standards.
Laundry final-rinse recovery
Laundry rinseFinal-rinse water from an on-site laundry, captured and reused as wash water for the next cycle.
Why it is overlooked: Modern tunnel washers can do this natively, but retrofit on older machines is rarely evaluated.
- Retrofit complexity
- Moderate — some plant and pipework
- Capital category
- Medium capital project
- Minimum treatment
- Filtration
Available on every laundry shift; strongest where laundry runs seven days a week.
Surfactant and temperature carry-over affect wash quality and machine warranty. Confirm the approach with the machine manufacturer and the chemical supplier before committing.
Permitted end uses, best match first
Closed-loop within one plant room — no second building network, which is why it is often the cheapest real saving on site.
Secondary option where rinse volumes exceed what the wash cycle can absorb.
Desalination / RO reject recovery
RO rejectConcentrate rejected by a reverse-osmosis or desalination plant, which is typically discharged.
Why it is overlooked: It is treated purely as an effluent disposal problem rather than a partially-usable stream.
- Retrofit complexity
- High — significant pipework or plant room changes
- Capital category
- Medium capital project
- Minimum treatment
- Advanced treatment (membrane or equivalent)
Continuous while the plant runs.
High salinity makes most reuse unsuitable and discharge is usually permit-controlled. Realistic options are limited to salt-tolerant duties or recovery-rate improvement rather than direct reuse.
Permitted end uses, best match first
Salinity limits this to non-corrosion-sensitive surfaces; recovery-rate improvement on the plant is usually the better project.
Only credible with salt-tolerant planting and soil salinity monitoring. Usually not recommended.
The end uses recovered water can serve
Share of total property demand each duty typically represents. These shares cap how much of any stream can usefully be absorbed — a large stream matched to a small duty is a small opportunity.
Landscape irrigation
5-35% of demandGardens, lawns, planters and green screening, ideally sub-surface drip.
Toilet flushing
8-16% of demandWC cisterns in guest rooms and public areas, via a marked second network.
Cooling tower make-up
3-15% of demandReplacing evaporative and blowdown losses on cooling towers.
Pool top-up
2-12% of demandReplacing pool evaporation and backwash losses.
Laundry pre-wash
5-20% of demandFirst wash stages in an on-site laundry, where quality tolerance is highest.
External washdown
1-4% of demandTerraces, hard standing, back-of-house areas and vehicle washing.
Ornamental water features
1-3% of demandFountains and ponds where public contact is incidental.
Open working
The coefficients we use
The actual bands the engine applies, with the source each came from. This is the section that turns trust us into check us.
Whole-property intensity by type
Litres per occupied room night, measured at the property boundary so a hotel already reporting HWMI can compare like for like.
City hotel
Urban, year-round, limited outdoor space. In-room share of demand: 60%.
200 - 450 L / occupied room night (typically 320)
Boutique hotel
Under ~60 rooms, high service ratio. In-room share of demand: 58%.
250 - 600 L / occupied room night (typically 400)
Resort
Pool, gardens, leisure facilities. In-room share of demand: 42%.
450 - 1,200 L / occupied room night (typically 750)
All-inclusive resort
Full board, multiple outlets, extensive grounds. In-room share of demand: 36%.
600 - 1,500 L / occupied room night (typically 950)
Aparthotel / serviced apartments
Self-catering units, in-unit laundry common. In-room share of demand: 68%.
220 - 520 L / occupied room night (typically 360)
Villa / short-let
Private villa, holiday home or short-let apartment. In-room share of demand: 72%.
150 - 480 L / occupied room night (typically 280)
Demand-side coefficients
On-site laundry
Water per kilogram of linen through a commercial laundry.
8 - 20 L / kg linen (typically 13)
Linen generated
Kilograms of linen per occupied room night.
3.0 - 6.5 kg / occupied room night (typically 4.5)
Production kitchen
Per cover served, including warewashing.
12 - 32 L / cover (typically 20)
Spa and wet areas
Per guest night where a spa or thermal circuit operates.
15 - 70 L / guest night (typically 35)
Pool filter backwash
Share of pool volume discharged per week of operation.
1% - 5% (typically 3%)
Recovery coefficients
Condensate yield per unit of cooling
Litres of condensate per kWh of cooling delivered, by humidity band.
arid: 0.05-0.2 L/kWh · moderate: 0.15-0.5 L/kWh · humid: 0.3-0.85 L/kWh · very-humid: 0.5-1.3 L/kWh
Cooling delivered per occupied room
kWh of cooling per occupied room night, by humidity band.
arid: 8-22 kWh · moderate: 5-16 kWh · humid: 10-26 kWh · very-humid: 14-34 kWh
Condensate capture efficiency
Share of theoretical condensate realistically collected, by system type.
central-chilled: 55-90% · mixed: 30-70% · split-units: 15-50% · none: 0-0%
Roof runoff coefficient
Rainfall that reaches storage after first-flush diversion and losses.
65% - 90% (typically 80%)
Greywater share of in-room demand
Shower and basin flows only. WC waste is excluded by definition.
50% - 75% (typically 62%)
Greywater treatment yield
Share of collected greywater that survives treatment as usable product water.
70% - 90% (typically 82%)
Laundry final-rinse recovery
Share of laundry water recoverable from the final rinse.
15% - 40% (typically 28%)
Regional profiles
Deliberately coarse climate bands, not site meteorology. A specialist replaces them with local station data.
| Region | Rainfall (mm/yr) | Off-season rain | Pool evap. (mm/day) | Irrigation (L/m²/yr) | Water stress |
|---|---|---|---|---|---|
| Mediterranean island | 350-750 | 80% | 4-8 | 400-900 | high |
| Mediterranean mainland | 400-900 | 75% | 3.5-7.5 | 350-850 | high |
| Gulf / arid | 40-160 | 50% | 6-12 | 900-2000 | extreme |
| Tropical island | 1100-2600 | 35% | 3-6 | 150-600 | high |
| Temperate northern Europe | 600-1200 | 55% | 1.5-3.5 | 50-300 | medium |
| Somewhere else | 300-1400 | 50% | 2-9 | 100-1500 | medium |
Water-stress classification follows WRI Aqueduct.
Standing assumptions
What the engine assumes when you do not tell it
Each assumption below is stated with the direction it could be wrong in, so you know which way to lean when you read your own result.
Occupancy applies to operating days, not calendar days
Volumes are built from occupied room nights: keys x occupancy x operating days. Year-round properties are modelled at 350-365 days, seasonal properties at 150-240 days.
If it does not hold: A property with a long shoulder season sits above the seasonal band, so its recoverable volumes are understated.
Stated occupancy is an annual average
A single occupancy figure is applied evenly across the operating period. No monthly profile is modelled, because a pre-feasibility screen does not need one.
If it does not hold: Sharp peaks are flattened, so peak-day sizing will differ from anything implied here. Sizing is the specialist's job.
Drainage is assumed separable
Greywater and laundry figures assume the relevant drainage can be separated from WC waste at reasonable cost. In an existing building this is the single largest unknown.
If it does not hold: Optimistic. If risers are combined and inaccessible, the greywater opportunity can collapse from large to unviable.
Roof area is treated as effective catchment
Stated roof area is taken as usable collection area, reduced by the published runoff coefficient for first flush and losses.
If it does not hold: Optimistic where roofs are terraced, planted, shaded by plant or drain to inaccessible points.
Condensate is collectable from existing drain runs
Condensate yield is reduced by a capture efficiency band reflecting the cooling system type, which stands in for how gatherable the drains are.
If it does not hold: Split-unit properties sit at the bottom of the band precisely because per-unit collection is rarely worth it.
No leakage allowance beyond the published bands
Distribution losses are not modelled separately. Benchmark intensities already include the typical losses of the properties they were measured on.
If it does not hold: A property with a known leak reads as high demand rather than as a leak. Leak detection is almost always the cheaper first move.
No behavioural or fixture change is credited
The snapshot models the property as it operates today. Aerators, linen-reuse programmes and guest campaigns are not counted as recovery.
If it does not hold: Conservative. Efficiency measures usually reduce demand before any reuse scheme is built.
No price, tariff or payback is assumed
Water and wastewater tariffs vary too widely to model responsibly, so the engine outputs volumes, complexity and capital categories only.
If it does not hold: Neutral. Any business case must be built on your actual tariff and discharge charges.
Regulatory feasibility is assumed, never verified
Reuse pairings are shown at the treatment class the literature requires. Local consent, risk-management plans and permits are not checked.
If it does not hold: Optimistic. In some jurisdictions a technically sound pairing is simply not permitted.
Comparison
How we compare you to peers
Published per-type bands, adjusted for climate, pool provision and landscape load so you are compared with properties that look like yours.
Climate adjustment
Multiplier applied to the published per-type band.
- Mediterranean island×1.05
- Mediterranean mainland×1.00
- Gulf / arid×1.30
- Tropical island×1.15
- Temperate northern Europe×0.85
- Somewhere else×1.00
Leisure load adjustment
Pool provision, plus irrigated landscape per key.
- none pool×0.92
- small pool×1.00
- medium pool×1.08
- large pool×1.18
- Irrigated landscape, per m² per key+0.0035 (capped at ×1.25)
Recovery headroom by property type
Share of total demand a well-executed peer of this type typically recycles. Your snapshot is placed against this band, not against a single target.
- City hotel5% - 18% (typically 10%)
- Boutique hotel5% - 20% (typically 11%)
- Resort10% - 32% (typically 19%)
- All-inclusive resort12% - 36% (typically 22%)
- Aparthotel / serviced apartments6% - 21% (typically 12%)
- Villa / short-let5% - 24% (typically 13%)
No other property's private data is ever used in your comparison. Peer bands are built from published datasets and adjusted deterministically by the factors above, so two properties with identical answers always receive an identical band.
Science base
Where the science comes from
The full reference library behind the knowledge base. Adding a source to the engine adds it here automatically.
Peer-reviewed research
Sustainability (MDPI) — Magrini, Cattani, Cartesegna, Magnani · 2017
What we take from it: Condensate production per unit of cooling delivered, and its dependence on ambient humidity.
Journal of Physics: Conference Series — Khan et al. · 2026
What we take from it: Recoverable fraction of theoretical condensate once collection losses and drainage geometry are accounted for.
Water Resources Management (Springer) — open access · 2025
What we take from it: Condensate quality characteristics and appropriate end uses in large buildings.
Environmental Science and Pollution Research - Ghaitidak, Yadav · 2013
What we take from it: Greywater strength and variability from shower and basin sources, and the treatment classes needed before non-potable reuse.
Water, Air, & Soil Pollution - Oteng-Peprah, Acheampong, deVries · 2018
What we take from it: Product-water yield across greywater treatment trains, and the operational monitoring burden that follows a reuse scheme.
Journal of Cleaner Production - Domenech, Sauri · 2011
What we take from it: Realised rainwater savings in a Mediterranean climate, and evidence that storage volume rather than roof area limits yield.
Tourism Management - Gossling, Peeters, Hall et al. · 2012
What we take from it: Spread of per-guest water use across accommodation types and destinations, used to sanity-check the property-type bands.
Journal of Environmental Management - Deya Tortella, Tirado · 2011
What we take from it: Effect of pools, gardens, star rating and seasonality on hotel water intensity on a water-stressed island.
Standards and frameworks
Sustainable Hospitality Alliance · 2020
What we take from it: Boundary definitions and the per-occupied-room and per-m² meeting space normalisation Aquiterum reports against.
International Organization for Standardization · 2019
What we take from it: Structure of a defensible water-efficiency review: baseline, significant water uses, targets and verification.
Alliance for Water Stewardship · 2019
What we take from it: Catchment-context framing and the expectation that on-site action is judged against local water stress.
Benchmark datasets
Cornell Center for Hospitality Research · 2024
What we take from it: Litres per occupied room bands by property class and region, drawn from tens of thousands of reporting hotels.
Cornell Center for Hospitality Research · 2021
What we take from it: Separate full-service resort banding, which sits materially above the all-hotel average.
World Resources Institute · 2023
What we take from it: Baseline water-stress classification used to weight regional urgency.
Regulation
European Union · 2020
What we take from it: Reclaimed water quality classes A–D for agricultural irrigation, and the risk-management plan requirement that applies to reuse schemes in EU member states.
Public-health and regulator guidance
World Health Organization · 2006
What we take from it: Health-protection framework for greywater reuse and the exposure controls that govern which end uses are acceptable.
US Environmental Protection Agency · 2012
What we take from it: Treatment-class expectations by end use (irrigation, cooling make-up, toilet flushing) and cross-connection control requirements.
Sustainable Hospitality Alliance · 2024
What we take from it: Destination-level water-stress framing and the overlap between tourism demand and water-scarce basins.
Honesty
Limits of the method
A pre-feasibility screen tells you where to look and what to ask. It does not tell you what to build.
- No hydraulic modelling. We do not size pipework, pumps, tanks or plant.
- No site survey. Drainage separability, plant room space and routing are assumed, never verified.
- No water-quality testing. Stream quality is inferred from literature, not sampled.
- No pricing. We give capital and complexity categories, never a cost or a payback figure.
- No regulatory approval. Reuse consents, risk-management plans and permits sit with your specialist and your regulator.
- No vendor selection. We name intervention categories only, and receive nothing from anyone who sells equipment.
No Aquiterum output is a design, a specification or a permit. Every opportunity must be verified on site by a qualified water engineer and cleared with your local regulator before any commitment is made. Greywater reuse in particular is health-critical and is governed in the EU by Regulation (EU) 2020/741.
Going deeper
Research and reference reading
If you want to interrogate the underlying literature rather than take our summary of it, start here.
Condensate harvesting
Measured condensate yields per unit of cooling, quality characterisation and collection losses in hot and humid climates (Magrini 2017; Khan 2026).
Greywater reuse and health
Characterisation, treatment-train performance and risk management (Ghaitidak & Yadav 2013; Oteng-Peprah 2018; WHO 2006; Regulation (EU) 2020/741).
Rainwater in dry-summer climates
Realised savings and the storage constraint in Mediterranean settings (Domenech & Sauri 2011), alongside US EPA reuse guidelines.
Hospitality water benchmarking
Per-guest intensity across destinations and property classes (Gossling 2012; Deya Tortella & Tirado 2011; Cornell CHSB; HWMI), plus ISO 46001 and the AWS Standard for management-system framing.
Independent by design
Aquiterum does not manufacture water systems, sell equipment, or take a percentage of installation value. Recommendations are driven by property data, evidence quality and opportunity logic - not commercial relationships.
- No paid ranking
- No equipment commissions
- No sponsored opportunity results
Found something you disagree with?
Challenges to our coefficients are welcome and are how the knowledge base improves. Tell us which band looks wrong and what evidence you have, and we will publish the change with a version bump.