Water-risk assessment during site selection: a Middle East & GCC strategy guide
Direct answer
A water-risk assessment for AI data-centre site selection should determine whether the proposed cooling and support loads can be supplied, treated and discharged reliably throughout the project life. It must examine physical scarcity, utility infrastructure, water quality, competing demand, regulation, climate exposure and community context—not simply the presence of a pipe at the plot boundary.
For a Middle East or GCC project, the assessment should compare sites using the same load and cooling assumptions, distinguish committed utility capacity from unverified availability, and test normal, peak, phased and interrupted conditions. The result should identify fatal flaws, mitigations, cost and schedule exposure before land acquisition or a cooling architecture becomes difficult to change.
The strongest output is a decision record: evidence for every material assumption, an accountable owner, a date for confirmation and a clear trigger for reassessment.
Define the decision before collecting data
Site-selection studies often accumulate data without first defining what the decision requires. The project should state the proposed IT capacity, development phases, availability target, candidate cooling modes and evaluation period. It should also identify which conditions would make a site unacceptable rather than merely less attractive.
The water scope includes all significant uses: heat-rejection make-up, treatment losses, closed-loop filling, humidification where applicable, domestic services, cleaning, construction and landscaping. It also includes every liquid residual leaving the site. Ignoring a modest but regulated waste stream can delay a project whose main supply appears secure.
Assessment boundaries should be consistent across candidates. One site should not be credited with a future reclaimed-water scheme while another is judged only on existing infrastructure unless the delivery probability and schedule are evaluated equally.
The team should also agree the period of control. A hyperscale owner may require evidence through the full asset life, whereas a tenant may control only a lease term. Either way, future campus phases and surrounding development must be visible because they can consume headroom that appears abundant during the first connection application.
Establish a comparable demand basis
Water demand cannot be compared until the thermal assumptions are aligned. The model should translate IT load and facility overhead into heat-rejection duty, then calculate make-up and blowdown from weather, operating mode and source chemistry. Average annual volume, peak-day flow and instantaneous connection capacity answer different questions.
The demand profile should follow the construction and IT ramp. Early operation may have low water use but poor turndown in oversized treatment equipment. Later phases may exceed the original connection or discharge allocation. Maintenance, cleaning and simultaneous failure conditions need separate allowances.
Sensitivity cases should include alternative rack densities, liquid-cooling capture, dry or hybrid operation, higher source salinity and restricted cooling-tower cycles. A site that is attractive under one optimistic concept may remain viable only if a more expensive cooling architecture is adopted.
Test physical and seasonal water availability
Regional supply statistics do not prove deliverability at a particular parcel. The study needs evidence for the exact pressure zone, utility connection and time horizon. Minimum flow, peak allocation, pressure, interruption history and planned network upgrades should be obtained from the responsible provider.
Potential sources can include desalinated municipal water, treated sewage effluent, industrial reclaimed water, permitted groundwater, district networks or emergency tankers. Each has different constraints. Reclaimed supply may vary with upstream wastewater production; groundwater may face licensing or salinity limits; tanker capacity may be constrained during widespread emergencies.
Availability should be expressed in confidence bands. Existing measured capacity, contractually reserved capacity, approved future capacity and speculative capacity are not equivalent. The site record should show the evidence supporting each category.
Investigate utility infrastructure and common failures
A utility connection is a chain of assets: production, treatment, pumping, storage, transmission and local distribution. The project should trace critical dependencies, including electrical feeders and remote-control systems. Two incoming mains can share a vulnerability far upstream.
Route surveys should consider distance, elevation, wayleaves, road crossings, third-party land and construction interfaces. Connection cost and schedule can outweigh the nominal water tariff. A future pipeline shown in a masterplan should not be treated as available until its funding, approvals and completion date are credible.
On-site resilience options—dual connections, storage, alternative sources and reduced-water cooling modes—should be assessed against the failure mechanisms discovered. Adding tanks cannot solve an indefinitely constrained allocation, while a second connection provides limited value if it follows the same corridor.
Characterise source-water quality
Water quality affects cooling chemistry, treatment recovery, materials, discharge and operating cost. A short utility certificate may omit the constituents that control the design. Representative analysis should include dissolved salts, hardness, alkalinity, silica, chloride, sulphate, suspended solids, organic indicators and microbiological parameters appropriate to the source.
Historical data should cover seasonal change and unusual operating events. Desalinated supplies may be blended with other water; reclaimed-water quality can respond to upstream industrial discharges or treatment outages. Temperature and disinfectant residual can change during long conveyance or storage.
The assessment should convert analysis into consequences. Higher salinity may lower tower cycles or require membrane treatment; organics may increase biological-control demand; silica or hardness may set recovery limits. Comparing raw numbers without connecting them to the intended system understates the risk.
Sampling governance matters. Laboratory methods, detection limits, units and chain of custody should be consistent across sites. A single favourable sample should not replace the utility's historical record, while old regional data should not override recent samples at the actual connection. Conflicting evidence should be recorded and resolved through further testing.
Confirm discharge and residual routes
Every source and treatment option creates a residual. Candidate streams include cooling-tower blowdown, reverse-osmosis concentrate, filter backwash, softener regenerant, cleaning waste, tank washout and domestic sewage. Their combined flow and chemistry must have an approved destination.
The team should confirm sewer capacity, acceptance limits, connection level, tariff, monitoring and permitting. An evaporation pond or off-site disposal route needs land, environmental approval, vehicle access and a credible balance under worst-case weather. Zero-liquid-discharge proposals require substantial energy, chemicals, solids handling and operator capability.
Discharge constraints can change the preferred source. A high-salinity water supply may be physically available but produce an unacceptable concentrate. This issue should be resolved during site comparison rather than passed to detailed design as a generic mitigation.
Evaluate drought, climate and competing demand
Physical water risk includes both long-term scarcity and short disruption. Historic drought, projected heat, flood exposure, coastal hazards and extreme rainfall can affect supply and discharge assets. High ambient temperature also increases cooling demand and can reduce the capacity of dry equipment.
Competing demand matters because restrictions are policy decisions as well as hydrological events. Municipal growth, industry, agriculture and tourism may rely on the same network. The assessment should consider allocation priority and the reputational effect of using potable water where suitable alternatives may exist.
Climate scenarios should be applied to the project lifetime and capacity ramp. Precision is limited, so the purpose is to test whether the site remains workable across plausible conditions and which design choices preserve flexibility.
Assess regulatory and community context
Permits may govern abstraction, non-potable reuse, chemical storage, cooling-tower operation, aerosol control, discharge and environmental monitoring. Requirements differ across GCC jurisdictions and can change by development zone or utility. Written authority feedback is stronger evidence than assumptions transferred from another project.
Community context extends beyond legal compliance. A facility can face opposition if its water demand appears to compete with local needs, even when the supply is technically approved. Transparent reporting, reclaimed-water use and verified efficiency can strengthen the project’s position, provided claims use clearly defined boundaries.
Public-health obligations require particular care where reclaimed water or aerosol-generating cooling systems are proposed. Separation, labelling, access control, drift management and a competent Legionella programme should be included in the feasibility basis.
Stakeholder mapping can reveal constraints that engineering drawings miss. The development authority, water and sewer utilities, environmental regulator, public-health body, neighbouring landowners and emergency services may each control part of the solution. Early coordinated meetings reduce the risk of receiving individually positive responses that cannot be combined into one approvable scheme.
Compare cooling architectures at each site
The site and cooling concept should be selected together. Dry heat rejection reduces routine process-water demand but may require more land and electrical capacity in hot conditions. Evaporative systems can improve thermal efficiency while creating make-up and blowdown obligations. Hybrid plant can move between modes if the control strategy and equipment sizing support it.
Liquid cooling at the rack changes heat capture and facility-water temperatures but does not alone determine external water consumption. The final heat-rejection method remains decisive. Each site comparison should therefore model the whole path from IT equipment to atmosphere.
An option that works at one location may not transfer to another. Wet-bulb temperature, utility tariffs, water quality, plot geometry and discharge acceptance can reverse the ranking. Standard design modules should retain site-specific interfaces and operating envelopes.
Convert uncertainty into cost and programme risk
Risk should be visible in the commercial model. Connection charges, off-site pipework, treatment, storage, additional power capacity, residual disposal and permitting studies belong in the site comparison. Operating costs should reflect energy, water, chemicals, maintenance, membranes and laboratory testing.
Programme exposure can be more important than discounted cost. A utility upgrade, environmental permit or land crossing on the critical path may delay revenue. The schedule should show decision dates, information dependencies and contingency time rather than hiding them in a general risk allowance.
Where evidence is incomplete, use an explicit range and mitigation. A provisional reclaimed-water source might carry a fallback cost for dry cooling or potable supply. This keeps uncertainty from being mistaken for a favourable base case.
Lifecycle evaluation should include the value of optionality. Space for another treatment module, a second utility route or the ability to change heat-rejection mode may have little first-phase utilisation but substantial value if allocation rules or technology change. Conversely, an expensive contingency that cannot be commissioned within the required response time should not receive resilience credit.
Use a transparent site-scoring method
A weighted score can support comparison, but it must not average away fatal flaws. First apply pass/fail gates for legal availability, peak capacity, deliverable schedule, discharge route and minimum resilience. Only viable sites should proceed to weighted evaluation.
Scored criteria can include source diversity, water quality, treatment complexity, water and energy performance, expansion capacity, community sensitivity, cost and schedule confidence. Weights should reflect the owner’s priorities and be tested for sensitivity.
Every score should link to evidence and an owner. A review panel can then challenge assumptions, understand why two sites differ and update the decision when information changes. The method is valuable because it exposes judgement, not because it produces a mathematically precise winner.
Complete due diligence before commitment
Before land commitment, obtain utility correspondence, representative analyses, preliminary connection and discharge concepts, route constraints, design-weather data and an agreed cooling basis. Confirm which capacities are reserved and what approvals remain outstanding.
The project should also define the investigations needed immediately after selection: extended sampling, pilot testing, hydraulic confirmation, environmental studies or vendor performance modelling. Conditions precedent can protect the owner where a critical utility commitment remains incomplete.
The final site-selection report should record the water balance, source and discharge evidence, option comparison, resilience cases, cost range, schedule and open actions. This creates a traceable starting point for concept and detailed design.
How Crescent Engineering can support the project
Crescent Engineering can prepare comparative water balances, review source analyses, establish treatment concepts, assess reuse and discharge options, and translate uncertainty into site-specific infrastructure requirements. Early engineering helps identify where a source is genuinely usable rather than merely present.
Useful inputs include candidate-site coordinates, proposed IT phases, cooling options, utility information, available water analyses, plot plans and project availability requirements.
Crescent Engineering — Water Treatment
References and further reading
The final assessment must use current utility commitments, jurisdiction-specific requirements and project-approved cooling and availability assumptions.
Comments