Chilled Water Thermal Storage Tank: Chilled Water, Ice, and Phase Change Compared
Created on Today
Chilled Water Thermal Storage Tank: Chilled Water, Ice, and Phase Change Compared
For most data centers the chilled water thermal storage tank is the right choice. Sensible water holds about 6-8 kWh per cubic metre inside a 5-7 K window, needs no phase change and costs little more than the vessel itself; ice stores several times more per cubic metre but must be made with machine energy and then melted out, while PCM lifts density further and still needs a coil or jacket to charge. The medium is decided by floor area, the length of the peak window and the COP budget the plant is allowed, not by headline energy density. A chilled water thermal storage tank holds cooling capacity as cold water rather than as ice or phase change material, which keeps the hydraulic design of the plant unchanged and the shell at atmospheric pressure.
A storage question is really a floor area question wearing an energy costume. Two halls with the same cooling load can reach the same peak-shifting result with different shells, because the medium inside sets how many cubic metres are needed, how long the chiller runs to fill them and how the plant behaves the morning the tank is not ready.
Three media carry most of the installed base in critical facilities: chilled water held sensibly, ice formed on an internal coil, and phase change material in a bundle. Each buys higher density with a penalty that lands elsewhere in the mechanical design, usually in electrical power, in valve automation or in maintenance hours. This article sets them side by side on the quantities an owner's engineer must defend.
Sensible Water: The Baseline Every Other Medium Is Measured Against
Water is the reference case, because its storage mechanism is also the pipe fluid.
· Density comes from the temperature window. At cp ≈ 4.186 kJ/kg·K and ρ ≈ 1000 kg/m³, one cubic metre holds about 1.163 kWh per kelvin, so a 6 K window gives roughly 7 kWh per m³ before derating.
· Usable availability is understood. A chilled water thermal storage tank normally delivers 70-90% of theoretical, the shortfall being the thermocline and the reserves kept at each end.
· Failure is loud rather than silent. A stuck valve shows up as a rising supply temperature within minutes; the same fault in a frozen medium fades over hours.
The cost of that simplicity is volume, which is why the plot study should start before the medium choice is finalised.
Ice Storage: Density Bought With Machine Energy
Ice raises density by using latent heat rather than sensible heat, and the price is paid at the compressor.
· The melt path adds a surface. Discharge happens through an internal coil or a melt heat exchanger, and the rate is bounded by that surface rather than by the volume stored.
· Daily cycling is the operating norm. The charge window repeats almost every day, putting the coil, the defrost logic and the suction control under routine thermal cycling.
· Where space is capped, volume drops sharply. On an urban site with mechanical room area already committed to halls, the same duty needs a fraction of the water volume.
Ice leaves a project when the owner cannot justify the extra charge energy, or when the plant has no space for a melt coil that must stay clean.
Phase Change Material: Highest Density, Most Integration Work
PCM stores energy in a material that changes state near the supply water temperature.
· The heat transfer surface becomes the problem. Cold must enter and leave the capsule or pack through conduction, so discharge power follows pack geometry rather than pump flow.
· The charge medium still needs a coil either way. Water, glycol or refrigerant must move heat in and out, and that equipment occupies the room the PCM was meant to relieve.
· Supercooling and settling matter. A PCM that never nucleates, or that separates on standing, quietly turns an asset into a block with no useful duty.
PCM is a credible answer on a rich plant, and a poor answer for a facility that wants a tank a technician can walk into and verify with a thermometer.
Choosing by Consequence, Not by Density
The decision follows from what each medium forces the rest of the design to do.
1. Check the space against its alternative use. A shell that displaces hall space or forces a taller structure costs more than its price suggests.
2. Look at what the maintenance team can service. Water needs treatment; ice needs coil hygiene and defrost logic; PCM needs a supplier who will stand behind the pack after five years of cycling.
3. Keep one medium per storage train. Mixing a water tank and a PCM vessel on one header adds a mode-change logic that is hard to prove during commissioning.
Where the plant has room, the chilled water thermal storage tank remains the lowest-risk route: the storage fluid is the circulating fluid, the measurement is a thermometer and a flow meter, and the derating factors are the ones the industry has recorded for decades.
Technical Specification
Item | Chilled water (sensible) | Ice on coil | Phase change material |
Storage mechanism | Sensible cooling of liquid water | Latent fusion of water at an internal coil | Latent fusion of a designed alloy or hydrate |
Theoretical energy density | ≈ 1.163 kWh per m³ per K; about 7 kWh per m³ at 6 K | Latent heat of ice, several times the water volume | Set by the material's heat of fusion |
Usable withdrawal fraction | 70-90% of theoretical | Limited by melt coil surface and melting rate | Limited by pack conduction and contact area |
Extra charge energy | None beyond normal chiller operation | Latent heat of fusion plus a lower evaporating temperature | Charge medium energy plus the jacket approach |
Moving parts inside the shell | None | Coil, possibly agitators | None, but the pack must be supported |
Main maintenance item | Biocide, inhibitor, scale control, periodic cleaning | Coil hygiene, defrost and suction control | Capsule integrity, pack settling, nucleation |
Failure signature | Supply temperature drifts up within minutes | Capacity fades across the discharge window | Duty collapses when the pack stops cycling |
Typical fit | Campuses with yard space, 2-6 h shift | Urban sites where floor area is capped | Tight plots needing long hours in small volume |
Project Case
Our delivered reference work in the water and wastewater sector includes large multi-tank installations handed over to the same standards the data-center TES scope is engineered to, including the project below.
Project Case Summary
Project | Product | Capacity | Dimensions | Completion |
Namibia - potable water (P-A) | 01 GFS bolted tank | 44,900 m³ total | Four tanks at φ41,260 × 8,400 mm | September 2022 |
Four shells of this diameter are the proof of the water route's one weakness and one strength: the strength is that a shell this size holds temperature almost unchanged over years of standing, the weakness is that holding that cold takes volume. A chilled water thermal storage tank sized on the same logic keeps its thermocline wherever the diffuser leaves it, and the erection sequence for four matched tanks is the discipline used to bring a multi-unit storage farm online in stages.
Center Enamel Engineering Capability
Center Enamel (Shijiazhuang Zhengzhong Technology Co., Ltd) has designed and fabricated bolted storage tanks since 2008. As the first glass-fused-to-steel (GFS) tank manufacturer in China, the company holds close to 200 enamel-related patents, produces roughly 300,000 enamel-coated steel plates a year, has completed more than 30,000 installed projects and supplies its tanks to over 100 countries. The new 150,000 m² production base was added to raise output capacity, and single tanks are supplied up to 60,000 m³. Manufacturing runs under ISO 9001 and ISO 45001, with product certification including NSF/ANSI 61, WRAS, FDA, LFGB, CE (EN 1090), ISO 28765, FM, BSCI and EUROCODE, and design referenced to AWWA D103-09, AWWA C550 and NFPA where the application requires it.
Medium and shell choice
· Water as the storage medium keeps the loop at atmospheric pressure, which is the cheapest and lowest-risk way to hold several hours of cooling for a large hall; ice and PCM raise energy density but add coil, header or capsule geometry inside the shell.
· Enamel and fusion-bonded epoxy shells are inert against treated water, and the enamel route is verified with a 1500 V DC holiday test and a 0.25-0.45 mm fired layer.
· Our water-tank delivery record, from the Namibia 44,900 m³ scheme down to multi-tank industrial plants of a few hundred cubic metres, supports both utility-scale and hall-scale stores.
Frequently Asked Questions
Q1: Is chilled water really the lowest-risk medium for a data center?
A1: It is the lowest-risk where there is space and the peak window is short. Water is the same fluid as the circulating medium, has no internal melting surface and its derating factors are well recorded, which makes the commissioning test easier to prove.
Q2: How much more energy does ice storage consume per stored kilowatt-hour?
A2: More, and it belongs on the plant energy balance rather than in a claim. Making ice adds the latent heat of fusion plus the evaporator approach; compare that with the peak demand and tariff saving the shift produces.
Q3: Can a chilled water thermal storage tank share a plant with an ice system?
A3: They can share a common header, but the logic then has to decide which medium serves the loop, and that mode changing must be tested before it is relied on during an interruption.
Q4: What storage hours do water and ice each justify?
A4: Water is comfortable for two to six hours of shifted load and stretches further with volume; ice and PCM earn their extra equipment when the required hours push the water volume into an unavailable plot.
Q5: Which medium fails most quietly?
A5: The dense media. A water tank shows a rising supply temperature quickly, while a frozen or phase-changing pack gives up capacity gradually across the discharge window, so the test should be run against design duty and recorded.
Q6: Does the medium change the shell specification?
A6: The shell basics stay the same: atmospheric or within the system pressure boundary, insulated, with a controlled inlet and outlet and a lining fitted to the water. Ice and PCM change internal support, coil penetration and nozzle layout rather than the envelope.
Comparing chilled water, ice and phase change material comes down to which penalty the project can carry. Water carries volume, ice carries compressor energy and melt surface, PCM carries integration and long-term pack behaviour. For most critical facilities with a usable mechanical yard, the chilled water thermal storage tank delivers the shifted duty with the fewest new failure modes, which is why it is worth specifying first.
Talk to an Engineer
Send us the peak IT load, the hours you need to shift, the available mechanical space and the tariff structure. We will return a medium comparison on your numbers: energy density and gross volume for water, ice and PCM, the extra charge energy each demands, the usable fraction we would expect and the shell diameter and height that fit your plot. For a water scheme, we follow with a specification sheet covering lining, insulation, inlet geometry and nozzles.
WhatsApp