top of page

DC-BR packaged blowdown-recovery modules for AI data centres

Writer: Mark Cullens
Mark Cullens
Sep 16
4 min read


Blue-lit server-rack aisle inside a modern data centre

Contents

Direct answer

DC-BR is a packaged system that converts cooling-tower blowdown into reusable water while controlling concentrate and treatment waste. It is justified only after tower operation has been optimised and the site has a defined use for recovered product and an accepted residual route.


The module should be specified by net compliant product, sustainable recovery, availability, energy, chemical consumption and residual quality across an agreed feed envelope. A high headline recovery at one test condition is not an adequate basis for an operating data centre.


1. Position DC-BR within the water system

DC-BR begins at a segregated blowdown collection point and ends at defined product and residual battery limits. It may include equalisation, pretreatment, membranes, polishing, storage, chemical systems and transfer pumps. The module does not remove the evaporation required for heat rejection; it recovers a portion of the liquid that would otherwise leave as blowdown.


The published cooling-tower blowdown-recovery guide explains the overall opportunity. DC-BR focuses on how that opportunity becomes a repeatable packaged plant with measurable interfaces.


2. Confirm the baseline and opportunity

Measure tower make-up, blowdown, conductivity, cycles, heat load and weather. Reconcile flow-based and chemistry-based balances. Correct leaking bleed valves, overflow and poor control before sizing recovery; otherwise the module is designed to treat avoidable waste.


The investment case should value avoided make-up, discharge and capacity upgrades, then deduct energy, chemicals, consumables, maintenance and residual disposal. The business-case guide provides the calculation structure.


3. Characterise blowdown as a variable feed

Blowdown composition changes with source-water blend, cycles, tower chemistry, heat load and maintenance. Define normal, peak and upset ranges for flow, temperature, pH, conductivity, hardness, silica, alkalinity, chloride, sulphate, suspended solids, organics, oxidants and treatment residuals.


Equalisation can protect downstream treatment, but it must not hide an unsuitable feed or create biological growth. Include rules for diversion during shock dosing, cleaning or abnormal tower chemistry.


4. Select the recovery process

Treatment may combine filtration, activated carbon where appropriate, softening, cartridge filtration, reverse osmosis, high-recovery membrane stages and final polishing. Each step should remove a defined constraint. Complexity without a mass-balance purpose adds maintenance but not value.


Compare alternatives at the same feed cases, net product duty and residual boundary. The detailed filtration, softening and high-recovery RO guide describes how to arrange these processes around controlling chemistry.


5. Control scaling and fouling

Membrane recovery concentrates silica, hardness and salinity. Use saturation modelling, representative analysis and pilot work where uncertainty affects sustainable flux or cleaning interval. Pretreatment must also address suspended solids, organics and oxidant compatibility.


Set operating limits for normalised differential pressure, permeability, salt passage and cleaning trigger. The silica, hardness and salinity guide provides the chemistry framework.


6. Design the residual pathway

Recovery transfers contaminants into concentrate, backwash, sludge and cleaning waste. Define flow and mass for every residual at average, peak and upset conditions. Written acceptance from the receiving utility or disposal route should precede investment approval.


If discharge is constrained, compare partial recovery, brine minimisation and minimum- or zero-liquid-discharge options. The brine and discharge-control guide shows why the residual route often sets the economic optimum.


7. Configure capacity and redundancy

Size trains for phased campus demand, minimum stable flow and maintenance. An N+1 label is insufficient unless common pumps, tanks, analysers, chemical systems and controls are also assessed.


Provide a safe bypass to normal tower blowdown so the recovery plant cannot compromise cooling availability. Product storage may bridge short trips, but it should not be credited twice as both capacity and autonomy.


8. Integrate product reuse and controls

Recovered product may return to tower make-up, blend with another source or serve a different compatible duty. Model the recycle loop because membrane salt passage can accumulate when product returns to the same tower.


Controls should coordinate feed availability, product demand, tank inventory, off-spec diversion and residual capacity. Prevent product spill during low demand; production that cannot be reused is not a site saving.


9. Specify guarantees and verification

Guarantee net compliant product, sustainable recovery, availability, energy, chemicals, cleaning and residuals across agreed feed cases. Define sampling, corrections, exclusions and test duration.


Meter feed, product, concentrate and major waste streams. Separate process performance from site saving using the guarantees and measurement guide.


10. Plan retrofit and commissioning

For an operating site, survey tie-ins, access, temporary bypass, electrical capacity, controls and shutdown windows. Install isolation and prove the original blowdown route remains available during module maintenance.


Factory tests should challenge sequences and simulated failures. Site tests should prove hydraulics, product quality, residual routing, duty/standby changeover and recovery after trips. The retrofit guide covers live-site constraints in detail.


11. How Crescent Engineering can support the project

Crescent Engineering can audit the tower baseline, model water and salt balances, select treatment, package the DC-BR module, define guarantees and controls, and support retrofit, FAT, commissioning and performance verification.



12. References and further reading

Project recovery and discharge assumptions must be validated using representative site data.


Related Crescent articles

Continue exploring this topic:


Comments


bottom of page