“Free cooling” means using suitable outdoor conditions to reduce mechanical refrigeration, not eliminating all cooling-energy consumption. Fans, pumps, filtration and heat rejection equipment still use power.
The actual value of an economizer depends on the location, supply temperature requirements, IT-load profile and the safe transition into and out of economizer operation.
What to include in the decision
- Distinguish airside and waterside systems
- Count partial as well as full economizer hours
- Calculate net electrical savings
- Commission changeover behavior for reliability
Distinguish airside and waterside systems
Direct airside economization introduces filtered outdoor air when conditions are acceptable. Indirect airside systems use a heat exchanger to separate outside air from data hall air. Waterside economizers move heat through a liquid loop and may reduce compressor usage when ambient temperatures are low enough.
Each design has separate air quality, water, humidity, freeze-protection and maintenance requirements. A facility cannot select the best option from outside-air temperature alone.
Count partial as well as full economizer hours
Avoid the binary claim that a site either operates in “free cooling” or does not. Integrated economizer operation can reduce compressor work during moderate weather without fully switching it off.
Build an hourly or temperature-bin model using local dry bulb or wet bulb as appropriate, supply-water temperatures, heat-exchanger approach and minimum controllable plant load.
Calculate net electrical savings
The gross reduction in compressor kWh must be adjusted for added pump and fan electricity, water treatment, humidification and controls. Energy tariffs can also vary by time of use, which changes the value of each saved kWh.
For a water-limited location, a dry heat-rejection option may be preferable even if a wet option has lower electrical demand. Measure both PUE and water use rather than optimizing one metric in isolation.
Commission changeover behavior for reliability
Economizer valves, dampers, sensors and control logic create failure modes. Test return to mechanical cooling at changing ambient conditions, low-flow alarms, stuck valves and loss of a cooling unit.
ASHRAE’s data center guidance emphasizes that ambient-based savings must be achieved within appropriate IT environmental classes and with reliable control transitions.
Illustrative avoided-compressor estimate
If modeling identifies 2,500 eligible hours per year when an average of 600 kW of compressor power can be avoided, the gross reduction is 1,500,000 kWh. At an illustrative $0.09/kWh, gross energy value is $135,000/year. Subtract extra pump/fan energy, maintenance, water and capital recovery before calling that ROI. These figures are a worked example, not a site estimate.
A baseline-to-project measurement plan
Before committing capital, log chiller compressor kW, cooling supply/return temperatures, outside dry- and wet-bulb conditions, pump and fan power, water use and IT load. Fit the existing operating behavior to weather and demand so the proposed savings are not merely the effect of a cooler season.
Estimate full and partial economizer operation by weather bin. After commissioning, compare actual metered performance with the model at similar loads and temperatures. This avoids claiming savings from differences in IT utilization or a changed energy tariff.
For ROI, annualize net avoided cost and compare it against installed CAPEX, added service requirements and lifecycle replacements. If a design raises water consumption to save electricity, display both effects explicitly.
Net savings: include the energy used by the economizer itself
Suppose an economizer avoids an estimated 400 kW of mechanical compressor power during 2,000 qualifying hours. Gross avoided energy is 800,000 kWh. If the economizer mode adds 70 kW of pump and fan demand throughout those hours, incremental energy is 140,000 kWh. Net energy reduction is 660,000 kWh; at a hypothetical $0.10/kWh it is $66,000 annually before water, treatment, additional service and capital costs.
Qualifying hours must come from actual location weather and the required supply temperature. Water-side economizers are sensitive to outdoor wet-bulb conditions and heat-exchanger approach temperature. Air-side options must also satisfy humidity, particulate and equipment limits. A nominal 2,000-hour estimate without weather and controls assumptions is not a reliable project forecast.
Partial economizer hours can still reduce chiller load; however, they require an integrated plant model. Avoid classifying every cool outdoor hour as a full compressor-off hour.
A measurable economizer investment case
| Input | Evidence needed | Error to avoid |
|---|---|---|
| Climate / hours | Weather bins: dry-bulb and wet-bulb as applicable | Using annual average temperature only |
| Plant baseline | Compressor, pump and fan meters | Ignoring ancillary energy |
| Set points / envelope | IT inlet and chilled-water temperatures | Applying unconstrained higher temperatures |
| Water / maintenance | Water consumption, treatment and service | Assuming electricity savings are free |
| Reliability | Bypass, sensor faults and failover tests | Removing backup cooling capability |
Measure actual performance after commissioning
Set an acceptance period covering mild-weather conditions and representative facility load. Record actual economizer mode hours, chiller starts, chiller kWh, pump and fan kWh and temperature alarms. Report avoided energy as a modeled counterfactual with stated assumptions, not as a direct metered quantity.
DOE FEMP’s data center cooling water guidance discusses both air-side and water-side economizing and the importance of humidity, contaminants and plant integration. Use that as a technical starting point; final ROI comes from site-specific measured and modeled values.
Questions to ask before signing or building
- Are savings calculated from hourly climate and real design supply temperatures?
- Are added pump, fan, treatment and water costs included?
- Does the analysis use actual electric tariff periods and plant-load variation?
- Have mechanical/economizer changeovers been tested under failures?
Methodology: illustrative figures are deliberately separated from sourced claims. Utility rates, commercial charges and engineering performance must be validated against site documents. Publication date: October 19, 2026.
Research and technical references
- US DOE FEMP: air- and water-side economizer considerations — technical or provider documentation, not a project quotation
- ASHRAE Handbook: Data Center Economizers — background reference, not a project quotation
- ASHRAE Handbook: Waterside Free Cooling — background reference, not a project quotation