Should a data centre use ice or chilled water in its Cold Storage Tank? The answer depends on more than storage capacity. Temperature requirements, available space, charging efficiency, peak-load demand and system integration all influence the choice. A compact ice system may look attractive on a site plan, for example, but its lower charging temperature can change the economics of the entire cooling plant.
For a data centre, the better question is not simply “Which medium stores more cooling?” It is “Which arrangement can deliver the required cooling temperature and flow, for the required duration, under the conditions the facility actually needs to survive?”
A chilled-water tank stores cooling primarily through the temperature difference between warmer return water and colder supply water. Ice storage uses the energy absorbed as ice melts, in addition to any sensible temperature change in the storage medium. That phase change allows ice systems to store more cooling in a given volume, although the useful capacity of either system depends on its operating temperatures and design.
Before comparing equipment, separate the duties the storage system is expected to perform. A tank intended to shift cooling production away from a high-tariff period has a different job from one intended to bridge a chiller restart. The first may discharge for hours; the second may need to respond immediately after a power or equipment disturbance.
Define the load profile, required supply temperature, return temperature, discharge duration and acceptable temperature rise at the load. Then test the design against the most demanding operating state—not just a convenient average. A storage system can have enough theoretical energy capacity and still fail if it cannot deliver the required cooling rate near the end of discharge.
Chilled-water storage often fits naturally into a plant that already supplies water at suitable temperatures. It can avoid the need to make ice and may integrate with existing chillers, pumps and heat exchangers more directly. For a facility with adequate space and a relatively short bridging duty, that simplicity can be worth more than a smaller tank footprint.
The practical constraint is the usable temperature difference. A narrow difference between stored supply water and return water means more water volume is needed for the same cooling duty. Designers should use the temperatures expected during the actual discharge event, rather than relying only on nominal plant conditions.
Stratification also deserves attention. In a stratified tank, warmer and colder water are kept in separate layers as far as practical. Poor inlet distribution, unsuitable flow rates or repeated cycling can increase mixing and reduce useful capacity. Ask for an explanation of diffuser design, operating flow range and how the control system protects the temperature layers. Tank volume alone does not demonstrate performance.
Ice storage is worth investigating where available space is tight or where a substantial cooling load must be shifted into a defined charging window. Its higher storage density can reduce the space needed for storage, but the system usually requires colder charging conditions than a comparable chilled-water arrangement.
That lower temperature can reduce chiller efficiency during charging. It may also require a different chiller selection, a secondary fluid, additional heat exchange or controls suited to ice formation and melting. The right comparison therefore includes electricity used to charge storage, pump energy, heat-exchanger losses and the value of cooling delivered—not just tank size or installed storage capacity.
Discharge performance needs particular scrutiny. The rate at which an ice system can provide cooling depends on its design and state of charge; it should not be assumed constant throughout a cycle. Request a discharge curve or operating schedule showing supply temperature and available cooling rate from the start of discharge to the minimum permitted state of charge. Also check whether the plant can recharge fully before the next expected event.
Storage does not operate in isolation. Trace the complete path from tank to heat exchanger, cooling distribution unit, distribution pipework and IT equipment. Each interface introduces temperature differences, pressure losses and control requirements. A tank that meets a plant-side temperature target may still leave too little temperature margin at the CDU or rack.
This matters especially for liquid-cooled data centres, where the facility-water circuit and the technology-cooling circuit may be separated by a heat exchanger. The colder output associated with an ice system does not automatically mean the server circuit should receive colder fluid. Supply temperatures must remain compatible with equipment requirements and condensation controls, including the relevant dew-point conditions.
Distribution hardware belongs in the review too. Shandong Liangdi Energy Saving Technology Co., Ltd., based in Jinan, works across data-centre storage tanks, CDUs, heat-exchanger units and distribution equipment. Its Liquid-Cooled Manifold is offered in single-row and double-row configurations, with SUS304 or SUS316L material options and dimensions listed as 30x30, 40x40 and 50x50. The manifold can be customised for cabinet arrangements; the dimensions, connection details and fluid compatibility should be confirmed for the specific design rather than assumed from the listing.
If a glycol-based fluid is proposed for an ice circuit, verify its concentration, material compatibility, viscosity and heat-transfer characteristics at the expected temperatures. Do not assume that the same fluid is suitable on the server side. The chosen heat-exchanger arrangement and equipment manufacturer requirements determine where each medium can be used.
Thermal storage can provide a cooling buffer, but it is not automatically a complete backup system. During a power interruption, identify which pumps, valves, controls, sensors and CDU components remain energised. Check the time required to detect the event, change valve positions and establish flow. If those functions depend on unavailable power, stored cooling may not reach the load when it is needed.
Review failure modes as well as normal operation. Can one pump or control component isolate the whole storage path? What happens if the tank is only partly charged when an outage occurs? How will operators know the usable state of charge? A commissioning test should demonstrate the intended transition and measure temperatures and flows at relevant points, within the limits of the agreed test procedure.
Ask suppliers to model both options against the same load profile, charging window, discharge duration and delivery temperature. Include the storage vessel, chillers or charging equipment, pumps, heat exchangers, controls and required building space. Compare maintenance access, water treatment or fluid management, and the consequences of a component failure alongside capital and operating costs.
Where space is available and plant temperatures are compatible, chilled water is often the more straightforward option to investigate first. Where storage volume is the limiting constraint, ice may justify its additional charging and integration demands. Either way, do not approve the selection until the proposed system can show the required cooling rate and temperature at the load throughout the defined discharge event.
Leave A Message
If you are interested in our products and want to know more details, please leave a message here, we will reply you as soon as we can.