Stratified Thermal Storage Tank: Design Rules For Usable Cooling Capacity
Created on Today
Stratified Thermal Storage Tank: Design Rules For Usable Cooling Capacity
Usable cooling capacity is the volume between the design full and discharge temperatures times the water's sensible storage density, not the nominal tank volume. It is set by the temperature spread, reduced by mixing and a chosen discharge endpoint, and it must be proven by a stratified-temperature test rather than assumed from the tank nameplate.
Introduction
Stratified thermal storage is one of the simpler ideas in cooling plant, and one of the easier ones to get wrong on paper. You store chilled water in a tall tank, let the cold water settle to the bottom and the warmer return rise to the top, and draw from the cold layer when the load peaks. The capacity you count in the sizing, however, is not the tank's nameplate volume; it is the smaller amount of cooling you can actually deliver between the design temperatures.
This article sets out the rules that connect nominal volume to deliverable cooling. It covers the difference between nominal and usable capacity, the temperature spread that defines the energy, the mixing that erodes it, the discharge endpoint that caps the draw, the linked set of rules a design must satisfy, how the usable capacity is tested, and how the result should be presented to the owner. The figures are derived from the sensible-heat relationship and from recognised test procedures, not from a single completed project.
The point throughout is that usable capacity is a calculated and verified quantity. A tank labelled by its geometric volume tells you almost nothing about how long it will carry the load, and the design work is in converting one to the other with numbers you can defend.
Nominal versus usable capacity
Nominal volume is the geometric interior volume of the tank, measured between the shell and the design high water level. It is the number that goes on the tank nameplate and the number that most preliminary estimates use. Usable cooling capacity is smaller, and it is the number that matters when you check the duty against the load profile. The difference is the volume that cannot be drawn at full temperature quality: the volume held above the design full temperature at the top, the volume left below the design discharge temperature at the bottom, any volume lost to inlet jet mixing, and the dead volume below the lowest draw-off nozzle.
You size the tank on the nameplate, but you prove the installation on the usable number. A specification that reports only the nominal volume leaves the reviewer unable to confirm the duty is met, and leaves you unable to defend the design if the plant falls short in the first summer. The usable figure is what the rest of the plant is sized on, so the gap between the two is the gap between a design that works and a design that merely looks complete on paper.
The temperature spread that defines capacity
Cooling capacity stored as sensible heat in water is the product of mass, specific heat, the usable volume and the temperature difference between the warm return and the cold supply: Q = rho c V T. With water at about 1000 kg/m3 and a specific heat of 4.186 kJ/kg.K, one cubic metre of water at a 10 K spread stores about 11.6 kWh of cooling, written as Q = 1.163 V T in kWh with V in m3 and T in K.
The spread therefore fixes the tank size for a given duty: a 6 K design spread needs roughly two-thirds more volume than a 10 K spread for the same energy. You set the spread from the chiller and coil capabilities, not from tank convenience, because narrowing the spread raises both tank volume and pump flow for the same duty. The Froude and Richardson numbers describe how inlet momentum disturbs the thermocline, and they are the reason a small spread is harder to protect than a large one, something the diffuser selection has to respect.
Mixing and the loss it causes
Stratification is maintained only if the warm water returns to the top and the cold water leaves from the bottom without the two streams cutting through each other. Inlet and outlet momentum, wall conduction, and heat gain through the envelope each erode the thermocline, the interface where temperature changes most sharply. The loss shows up as usable capacity below the value the pure temperature-spread calculation predicts, because a smeared interface carries warm water into the cold draw and forces an earlier shutdown.
You limit the loss with diffusers sized to keep the inlet Froude number low, with the inlet velocity below the threshold that the diffuser performance data supports, and with a tank aspect ratio that gives the thermocline room to stay thin. The penalty is not small: a poorly managed inlet can lose 10 to 20 percent of the usable capacity that the nominal volume suggests, which is why mixing is treated as a first-order design variable rather than a detail resolved at the pipe sizing stage.
Choosing the discharge endpoint
The discharge endpoint is the cold-supply temperature at which you stop drawing from the tank, and it is a design choice rather than a fixed property of the water. You set it from the cooling coil: the coil needs a supply temperature and a flow, and the lowest tank temperature that still satisfies the coil at its design flow becomes the endpoint.
Set the endpoint too cold and you may leave usable capacity in the tank but risk losing it to the coil's minimum approach; set it too warm and you throw away cold water the coil could have used. The endpoint also interacts with the chiller, because the chiller's minimum evaporator temperature and the coil approach set the practical floor. Stating the endpoint explicitly, with the coil approach temperature and the minimum chiller lift, lets you convert nominal volume to a single defensible usable number and keeps the selection from drifting during detailed design.
The rules that must be satisfied together
A stratified tank works only when several rules hold at once, and an installation that satisfies one and ignores the others will under-deliver. The tank must store the required energy within the stated spread, the thermocline must stay thin enough that mixing loss stays within the design allowance, the discharge endpoint must sit where the coil and chiller allow, the envelope heat gain must stay below the capacity the duty can absorb, and the inlet arrangement must keep the Froude number in the range the diffuser performance supports.
Each rule has a number: the energy balance, the diffuser inlet velocity, the envelope UA, the coil approach. You write them as a linked set, because raising the spread changes the volume, the volume changes the aspect ratio, and the aspect ratio changes the diffuser sizing. The design is the set of consistent numbers, not any one of them, and the handover should show that they were solved together rather than picked in isolation, which is the usual reason a reviewed design still fails the duty.
Verification and testing
Usable capacity is verified by a stratified-temperature test, not by reading the nameplate. You fill the tank through the cold inlet, establish the thermocline, then draw and measure: record the temperature profile from top to bottom as the tank discharges, and integrate the usable volume between the full and discharge temperatures against the measured flow and the measured temperatures.
The IAPMO Uniform Testing Procedures for Thermal Storage Tanks give a recognised method for this measurement, and ASHRAE guidance describes the instrumentation and the run length. A single top-and-bottom sensor is not enough; a vertical thermocline probe or a ladder of sensors is what shows whether the interface stayed thin or smeared. You report the measured usable capacity, the achieved spread, and the mixing loss against the design allowance, and you keep the run long enough that the result is repeatable rather than a single snapshot.
Common design failures
The failures repeat. A tank sized on nominal volume with no usable-capacity calculation arrives with 15 to 25 percent less deliverable cooling than the estimate implied. A diffuser chosen by pipe size rather than by Froude number destroys the thermocline on the first cycle. A discharge endpoint set by guess, not by coil approach, leaves cold water unused or starves the coil.
An envelope with no vapour sealing gains heat that the duty did not budget for, so the tank warms from the top even when idle. A single top-and-bottom sensor hides the mixing, so the owner blames the chiller. And a design reviewed section by section, never as a linked set, passes each check and still fails the duty. Each of these is a missing number, not a missing product, which is why the usable-capacity calculation and the test are what separate a working tank from a merely labelled one.
Presenting usable capacity to the owner
When you hand the tank over, present the usable number first and the nominal number second. Give the energy balance with the spread you assumed, the discharge endpoint you chose and the mixing loss you allowed, then state the resulting usable capacity in kWh and in equivalent run-time at the design load.
Show the measured value from the stratified test next to the design value, with the difference explained by the realised spread and endpoint rather than left as a surprise. The owner sizes the rest of the plant on the usable figure, and a duty that was counted in the building model on nominal volume will be found short when the plant runs. A one-page capacity statement, signed off against the test, is what keeps the storage from being relitigated every summer, and it is the document the operator actually uses when deciding whether the tank can carry the next peak.
Stratified Tank Design Rule Checklist
Design rule | Governing quantity | What you must state |
Nominal versus usable volume | Tank interior volume vs deliverable energy | Usable kWh and equivalent run-time, not nameplate volume only |
Temperature spread | Q = 1.163 V T (kWh, V in m3, T in K) | Return-to-supply difference taken from chiller and coil |
Mixing loss | Inlet Froude number, diffuser velocity | Allowed loss and realised thermocline thickness |
Discharge endpoint | Coil approach plus chiller minimum lift | Lowest usable temperature and the reason for it |
Envelope heat gain | Tank UA | Idle gain held within the duty budget |
Linked verification | Stratified-temperature test | Measured usable capacity versus the design value |
Capacity Loss From Nominal Volume To Usable
Loss source | Typical magnitude | Mechanism | Control |
Dead volume below nozzle | 1 to 3 percent of nominal | Water below the lowest draw-off is not recoverable | Lower nozzle placement |
Above-full-temperature hold | 2 to 5 percent | Top layer sits above the design warm limit | Level control to design high point |
Inlet jet mixing | 10 to 20 percent | Smeared thermocline, early shutdown | Diffuser sized to low Froude number |
Wall and roof conduction | 1 to 4 percent per idle day | Slow warm-up at the interface | Insulation with a vapour seal |
Discharge endpoint set high | up to 10 percent | Cold water left unused | Set by coil approach, not by guess |
About Center Enamel
Center Enamel (Shijiazhuang Zhengzhong Technology Co., Ltd.) designs and fabricates bolted steel storage tanks, welded steel tanks, fusion-bonded epoxy tanks, galvanized tanks, stainless steel vessels, pressure vessels, aluminum geodesic dome roofs and floating roofs. The company has more than 30 years of production experience and has supplied equipment to projects in more than 100 countries. Aluminum geodesic dome roofs are designed and fabricated to API 650 Appendix G, AWWA D108, the Aluminum Design Manual, ASCE 7 and IBC, and the company developed and exported a 72 m diameter aluminum dome roof in 2023. Engineering, fabrication, coating, quality control and site service are handled as one scope, which is why the drawings, the calculation basis and the delivered hardware can be reconciled against each other.
Frequently Asked Questions
How is usable cooling capacity defined for a storage tank?
Usable cooling capacity is the cooling energy you can actually draw from the tank between the design full temperature and the chosen discharge temperature. It equals the usable volume times the water's sensible storage density for the temperature spread, and it excludes the dead volume below the lowest nozzle, the volume held above the full temperature, and the capacity lost to mixing. You size the vessel on the nominal volume but you prove and report the duty on the usable figure, because only the usable number tells the operator how long the tank will carry the load.
What limits the usable capacity of a stratified tank?
Three things limit it: the temperature spread you assume, because capacity scales with the difference between the warm return and the cold supply; the discharge endpoint set by the coil approach and the chiller minimum lift, which fixes the lowest usable temperature; and mixing, which smears the thermocline and carries warm water into the cold draw. Envelope heat gain adds a slow loss during idle periods. Together these convert a nominal tank volume into a smaller, defensible usable capacity, and each one is a number you set or measure rather than a property of the tank shell.
How do you choose the discharge temperature?
You choose it from the cooling coil, not from the tank. The coil needs a supply temperature and a flow, and the lowest tank temperature that still meets the coil at its design flow is your endpoint. That floor sits above the chiller's minimum evaporator temperature plus the coil approach, so the chiller lift stays feasible. Set it too cold and you may strand usable capacity the coil cannot use; set it too warm and you discard cold water. State the endpoint with the coil approach and the chiller minimum, and the usable capacity follows as a single calculated number.
How much capacity does mixing cost?
Mixing can cost 10 to 20 percent of the usable capacity implied by the nominal volume, and in a badly run inlet the loss is larger. It appears as a smeared thermocline that lets warm return water reach the cold draw, forcing an early shutdown before the tank is empty. The cost is driven by inlet momentum, measured by the Froude number at the diffuser, and by heat gain through the envelope. A diffuser sized to keep the inlet velocity within its performance range, plus a tank aspect ratio that gives the thermocline room, is what recovers most of that capacity.
What rules must a stratified tank design satisfy?
A stratified tank design must satisfy a linked set: store the required energy within the stated spread, keep the thermocline thin enough that mixing loss stays within the design allowance, place the discharge endpoint where the coil and chiller allow, hold envelope heat gain below what the duty absorbs, and keep the inlet Froude number in the diffuser's supported range. Each rule has a number, and they interact: the spread sets the volume, the volume sets the aspect ratio, and the aspect ratio sets the diffuser sizing. The design is the consistent set, not any single value.
How is usable capacity verified?
You verify it with a stratified-temperature test. Fill through the cold inlet, establish the thermocline, then draw the tank while recording a vertical temperature profile, the flow and the supply and return temperatures. Integrate the usable volume between the full and discharge temperatures against the measured data. The IAPMO Uniform Testing Procedures for Thermal Storage Tanks describe a recognised method, and a vertical probe or sensor ladder is needed to see whether the interface stayed thin. Report the measured usable capacity, the realised spread and the mixing loss against the design allowance.
Why does a tank deliver less than planned?
Because the nominal volume is not the deliverable volume. The tank delivers the energy between the full and discharge temperatures after mixing, heat gain and dead volume are removed, and those losses are easy to ignore in a spreadsheet that uses only the geometric volume and a single spread. A design that never runs the usable-capacity calculation, never sizes the diffuser by Froude number, and never sets the endpoint by coil approach will report a capacity the tank cannot produce. The gap is a missing calculation, not a faulty tank.
Usable cooling capacity is the figure the plant is actually sized on, and it is a calculated, tested quantity rather than a nameplate number. Set the temperature spread from the chiller and coil, size the diffuser by Froude number so the thermocline stays thin, fix the discharge endpoint by the coil approach and the chiller minimum, and keep the envelope heat gain inside the duty budget. Those rules have to be solved together, because each one moves the others.
Hand the tank over with the usable capacity stated first and the nominal volume second, and back it with a stratified-temperature test that measures the realised spread and mixing loss. Send us the duty, the spread you can accept, the coil approach and the site constraints, and we will return a stratified tank specification with the usable capacity calculated and a test method defined, so the storage you count in the model is the storage the plant receives.
Talk To Our Engineering Team
Send the duty, the volume basis, the governing code and the site constraints, and our engineering team will come back with a specification that a supplier can price and a reviewer can approve. We will also tell you plainly when a different tank type or a different scope is the better answer.
WhatsApp