Sizing a chiller means matching its cooling capacity to the heat your process generates, at the temperature and flow it needs. Get it right and the system holds a stable temperature efficiently; get it wrong and you either can’t hold temperature or you waste energy and shorten equipment life. This guide walks through the inputs, the cooling-load calculation, a worked example, and the adjustments for glycol and ambient conditions.
The calculations below are a practical starting point based on standard heat-transfer principles. For a critical or complex process, confirm the result with a chiller specialist — small errors in heat load, flow, or fluid properties can change the required capacity significantly.
Why correct sizing matters
A chiller that is too small simply cannot remove heat fast enough at peak load, so the fluid temperature drifts above the set point and the process suffers. A chiller that is too large is just as problematic: it reaches temperature quickly and then shuts off, only to restart moments later. This “short-cycling” reduces temperature stability, wastes energy, and adds wear to the compressor. The goal of sizing is to land in the right range — enough capacity for the real peak load, with a sensible margin, and no more.
The inputs you need
Before calculating anything, gather the following:
- Heat load — how much heat the process produces, or enough data (flow and temperatures) to calculate it.
- Set-point temperature — the chilled-fluid temperature the process requires, and the precision needed.
- Temperature difference (ΔT) — the gap between the fluid leaving the chiller and the warmer fluid returning from the process.
- Flow rate — the volume of fluid circulated per minute or per hour.
- Coolant fluid — water or a water/glycol mix, since this affects heat transfer.
- Ambient conditions — the maximum environmental temperature, which affects heat rejection.
How to calculate the cooling load
Cooling load is driven by three things: how much fluid you move, how much you cool it, and the fluid’s ability to carry heat. The core relationship is:
Cooling load = flow rate × specific heat × temperature difference (ΔT)
In practice, two simplified forms are commonly used for water:
- Metric (kW): cooling load ≈ flow (litres/min) × ΔT (°C) × 0.07
- Imperial (BTU/hr): cooling load = flow (GPM) × 500 × ΔT (°F)
Capacity is often quoted in tons of refrigeration. One ton equals 12,000 BTU/hr, or about 3.517 kW — so to convert BTU/hr to tons, divide by 12,000, and to convert kW to tons, divide by about 3.5.
A worked example
Suppose a process circulates 20 litres per minute of water and needs the fluid cooled with a ΔT of 5°C between return and supply:
- Cooling load ≈ 20 × 5 × 0.07 = 7 kW
- Add a 20% safety factor: 7 × 1.2 ≈ 8.4 kW
So a chiller in the order of 8–9 kW would suit this load — assuming plain water and a reasonable ambient. If the fluid were a water/glycol mix, or the ambient unusually hot, the required capacity would rise (see below).
Add a safety factor
Calculated loads are estimates, and real systems face measurement error, gradual fouling that reduces heat transfer, and occasional peak conditions. For that reason, a margin of roughly 20% above the calculated load is a common rule of thumb. Tighten or widen that margin based on how confident you are in the heat-load figure and how critical the process is — a well-characterized, non-critical load needs less margin than an uncertain or mission-critical one.
Adjust for glycol and fluid type
If the loop uses a water/glycol mix rather than plain water, the fluid transfers heat less efficiently and is more viscous, so the same chiller carries less effective capacity. Sizing must therefore include a derate based on the glycol concentration and operating temperature, so the unit delivers full performance on the actual fluid — not on the water-only figure. The colder the process and the higher the glycol percentage, the larger this adjustment.
Account for ambient temperature
How a chiller rejects heat also affects sizing. An air-cooled chiller loses capacity as the surrounding air gets hotter, so it must be sized for the maximum expected ambient temperature, not an average day. A water-cooled chiller is less sensitive to ambient air but depends on the temperature of its condenser water. Either way, the published capacity of a chiller is tied to specific conditions, so it is important to confirm the rating at your conditions rather than the nameplate ideal.
Don’t forget flow and pressure
Capacity is only half the picture. The chiller’s pump must also deliver the flow rate and pressure the equipment needs at its coolant connection. Too little flow and the fluid heats up too much across the process; too much and you waste pumping energy. Flow, ΔT, and capacity are linked — a smaller ΔT demands higher flow to remove the same heat, and vice versa — so they should be specified together.
Common sizing mistakes
- Sizing on water when the loop runs glycol — leads to an undersized system in service.
- Using average instead of peak heat load — the chiller must cover the worst case, not the typical one.
- Ignoring maximum ambient — an air-cooled unit rated at 25°C ambient will fall short on a 40°C day.
- Oversizing “to be safe” — excessive margin causes short-cycling and instability, the opposite of the intent.
- Forgetting flow and pressure — a correctly sized capacity still fails if the pump can’t move the fluid as required.
Get an exact sizing
The method above gets you to a sound estimate, but a precise specification accounts for fluid properties, ambient conditions, flow and pressure, and the precision your process demands. Eldrotec sizes and builds chillers — standard or fully custom — to match your exact requirements across a capacity range up to 30 kW. If you have your flow, temperatures, and fluid in hand, the engineering team can confirm the right unit, or you can browse the full range on the recirculating chillers page.
Frequently Asked Questions
How do you calculate the size of a chiller?
Calculate the cooling load from the fluid flow rate and the temperature difference (ΔT) between supply and return. In metric, cooling load in kW for water is approximately flow in litres per minute × ΔT in °C × 0.07. In imperial, BTU/hr equals flow in GPM × 500 × ΔT in °F. Then add a safety factor and adjust for fluid type and ambient conditions.
What is a ton of cooling?
A ton of refrigeration is a unit of cooling capacity equal to 12,000 BTU per hour, or about 3.517 kW. Chiller capacity is often quoted in tons, kilowatts, or BTU/hr; one ton equals roughly 3.5 kW.
What temperature difference (ΔT) should I use?
ΔT is the difference between the chilled fluid leaving the chiller and the warmer fluid returning from the process. Many process loops are designed around a ΔT of about 5°C (roughly 10°F), but the correct value depends on the equipment. A smaller ΔT requires a higher flow rate to remove the same heat; a larger ΔT allows lower flow.
How much safety factor should I add when sizing a chiller?
A safety margin of around 20% above the calculated cooling load is common, to cover measurement uncertainty, fouling over time, and peak conditions. The exact margin depends on how well the heat load is known and how critical the process is.
Does glycol change chiller sizing?
Yes. A water/glycol mix transfers heat less efficiently than plain water, so a glycol system carries less cooling capacity at the same flow. Sizing must include a derate based on the glycol concentration and operating temperature, so the chiller delivers full performance on the actual fluid.
What happens if a chiller is oversized or undersized?
An undersized chiller cannot hold the set-point temperature under full load, causing the process to drift warm. An oversized chiller short-cycles — switching on and off frequently — which reduces temperature stability, wastes energy, and increases wear. Correct sizing keeps temperature stable and the system efficient.