A Small Water Chiller Has No Single Capacity — It Has a Curve
Two machines sit side by side in a comparison table. One says 1,000 W, the other says 800 W. The 1,000 W unit looks like the better buy, and at a setpoint of −20 °C it may deliver considerably less cooling than its rival.
The reason is that a small water chiller does not have a capacity. It has a capacity at a stated fluid temperature, and that figure falls as the setpoint drops. A single number on a listing is one point on a curve, and unless both machines are quoted at the same point, the comparison means nothing.
Nowhere does this bite harder than below freezing, where two things change at once: the capacity drops steeply, and the fluid has to change from water to a glycol mixture, which alters the flow calculation as well. This guide covers how to size across that boundary rather than either side of it.
Why Capacity Falls With the Setpoint
A vapor-compression cycle absorbs heat by boiling refrigerant at a temperature below whatever it is cooling. Ask for colder fluid and the refrigerant must evaporate colder still.
Three effects then stack up. The evaporating pressure falls, so the pressure ratio the compressor has to produce rises. The refrigerant vapor arriving at the compressor is less dense, so each revolution moves less mass. And the compressor’s own losses grow as a proportion of a shrinking useful output.
The result is a curve that falls away steeply rather than linearly. As a rough shape for a single-stage machine: if capacity is 100% at +20 °C, expect somewhere near 60% at 0 °C and around 30% at −20 °C. The exact numbers depend on the refrigerant and the machine, which is precisely why the manufacturer’s own curve matters more than any rule of thumb.
The practical rule: decide the setpoint before you compare capacities, and read every candidate at that setpoint.
Below Zero, the Fluid Changes Too
Water freezes, and a frozen evaporator is an expensive failure. So below about +5 °C the loop moves to a water/glycol mixture, and that brings three consequences most sizing calculations skip.
Lower heat capacity. Glycol carries less heat per degree than water. A 30% mixture holds roughly 10% less, so the same flow removes less heat for the same temperature rise.
Higher viscosity, and it worsens as it cools. This is the effect that surprises people. A mixture that pumps comfortably at +10 °C becomes noticeably thicker at −20 °C, and the same pump moves less of it through the same circuit.
Both effects push the same way. Lower heat capacity and lower flow compound, so a loop designed on water figures and then run cold on glycol can fall short on capacity and flow simultaneously — and the two symptoms look like separate faults.
The rule that follows: if the setpoint is below freezing, run the flow calculation with the glycol mixture’s properties at the cold end of its range, not with water’s at room temperature.
Which fluid, and at what concentration, is covered in the lab chiller coolant selection guide, and the water-versus-glycol ratio question specifically in the glycol ratio guide.
The Sizing Sequence
Sizing a small water chiller runs in four steps, in this order. Reversing any two of them produces the wrong machine.
- Establish the heat load. From the instrument’s documentation if it states one, from electrical input as an upper bound, or by measuring temperature rise across the load at a known flow. The one-line check for water is Q = flow (L/min) × 69.7 × ΔT (°C).
- Fix the setpoint. Not a range, not “as cold as possible” — the temperature the process actually requires.
- Read capacity at that setpoint. From the manufacturer’s curve or table. Add 20–30% margin here, at this condition.
- Check flow against your circuit. Required flow is Q ÷ (69.7 × allowable ΔT), and it must be delivered against your circuit’s real pressure drop — which means comparing against a pump curve, not a headline flow figure.
A Worked Example Across the Boundary
To show how much a small water chiller has to grow when the setpoint drops, here is the same instrument sized twice. The same instrument, sized at two setpoints, to show how much moves.
The load: 300 W, constant.
Case A — setpoint +15 °C. Fluid is distilled water — and the setpoint sits above MiniCool’s 0 °C ceiling, so this half of the problem belongs to the DV-series compact liquid chiller line (100–550 W across −20 to +20 °C). Derating at +15 °C is modest, so a unit in the upper half of that range carries 300 W with comfortable margin. Flow for a 3 °C rise is 300 ÷ (69.7 × 3) ≈ 1.4 L/min, easily delivered.
Case B — setpoint −20 °C. Same 300 W load. Now the fluid is a glycol mixture, and capacity at −20 °C may be roughly a third of the 20 °C figure. That same 450 W machine might deliver 150 W here — half of what the load requires. You need a machine rated closer to 900–1,000 W at 20 °C to have 300 W plus margin at −20 °C. And the flow calculation now uses glycol’s lower heat capacity and higher cold viscosity, so the pump has more work to do as well.
Same load, same instrument, and the honest answer is a machine roughly twice the nominal size. Nothing in the process changed except one number in the specification.
Reading a Capacity Table Properly
| What the listing says | What to ask | Why it matters |
|---|---|---|
| “1,000 W cooling capacity” | At what fluid temperature? | Without it, the number cannot be compared with anything |
| “Operating range 0 to −40 °C” | What is the capacity at the cold end? | Range describes what it can reach, not what it can remove there |
| “Flow rate 10 L/min” | At what pressure? | Open-pipe flow is not circuit flow |
| “±0.1 °C stability” | Measured where, and with what control? | Reservoir stability differs from stability at your instrument |
| “Ambient up to 35 °C” | What is the capacity at 35 °C? | Air-cooled capacity falls as the room warms |
The second row is the one that causes the most disappointment. A stated temperature range tells you the coldest the machine can reach with little or no load on it. It says nothing about how many watts it can remove down there, and those are different questions.
Where the heat is ultimately rejected changes the numbers again; the trade-offs are set out in air-cooled versus water-cooled lab chillers.
Stability, and Why It Is a Separate Specification
Capacity and stability are independent. A machine can have plenty of capacity and still hold temperature poorly, and the reason is how it controls.
An on/off controller switches the compressor around the setpoint, producing a sawtooth whose width is the stability you actually get. A modulating controller varies compressor speed to match the load continuously, holding much closer to a line.
Where a stability figure is quoted to a tenth of a degree — the MiniCool range, for instance, states ±0.1 °C across its 0 °C to −40 °C span — that describes modulating behavior at the sensor. Two follow-up questions are worth asking: where the sensor sits, and whether the figure holds across the whole temperature range or only at a convenient point.
Oversizing Does Not Buy Safety
Having established that sub-zero duty needs a larger small water chiller than it first appears, the opposite error is worth naming.
A chiller far larger than its load reaches setpoint quickly, stops, drifts, and restarts. On an on/off machine that produces a wide sawtooth — worse stability than a correctly sized unit, not better. It also puts more heat into the room, because an air-cooled machine rejects everything it removes plus its own electrical input, and it occupies bench space that laboratories rarely have spare.
Twenty to thirty percent above the calculated load, evaluated at the working setpoint, is the useful target. The reason sub-zero applications need large-sounding machines is not margin — it is that the capacity at the setpoint is genuinely much lower than the headline.
Line Losses: The Watts That Never Arrive
One term is missing from most sizing calculations and it grows exactly when the setpoint falls.
Fluid running below ambient picks up heat along every meter of tubing between the chiller and the instrument. At a +15 °C setpoint in a 22 °C room the difference driving that gain is 7 K and the loss is minor. At −20 °C it is 42 K, six times larger, and on an uninsulated run of a few meters it can amount to tens of watts — a meaningful fraction of a 300 W load.
Three consequences follow. Insulate the lines whenever the setpoint is below ambient, and insulate both of them rather than just the supply. Keep runs as short as the layout allows, since the loss is proportional to length. And measure at the instrument rather than at the chiller — a machine holding its setpoint perfectly can still be delivering warmer fluid than the process needs, and the display will never show it.
The failure mode this produces is distinctive: the chiller reads correct, the instrument runs warm, and adding capacity does not help because the heat is entering between the two. Insulation fixes it for a fraction of the price of a larger machine.
Where a Countertop Unit Fits
The small water chiller class serves loads of a few hundred watts to about a kilowatt, at setpoints running from ambient down to well below freezing, on instruments that sit on a bench and cannot be moved — that combination is what countertop recirculating chillers are built for.
The MiniCool™ countertop recirculating lab chiller covers this band in three sizes (MiniCool 450, 800, and 850), with a stated 450 W to over 1 kW of cooling capacity, a 0 °C to −40 °C range, ±0.1 °C stability, a choice of R134a, R290 or R1234yf refrigerant, and AC (120–240 V) or DC supply. The page’s own instrument list spans both temperature zones, but the published span is 0 °C to −40 °C — so MiniCool is the machine for Case B and its sub-zero relatives, while Case A and anything else above 0 °C belongs on the DV line described above.
If you are sizing across the freezing boundary and are unsure how much the derating costs you, send the three numbers rather than estimating. Contact the RIGID engineering team with your heat load in watts, your target setpoint, and the flow your instrument requires, and ask for capacity at that setpoint rather than the headline figure.
Frequently Asked Questions
Why is the capacity lower than the specification says?
Because the specification is quoted at a warmer fluid temperature than you are running, or a cooler ambient than your room, or both. Ask for the figure at your setpoint and your room temperature.
How much capacity do I lose going from 0 °C to −20 °C?
Often around half, though it depends on the machine and the refrigerant. This is why a single-stage chiller sized for chilled-water duty is not the same machine as one sized for sub-zero work.
Do I need glycol, and how much?
Below roughly +5 °C, yes. Use the minimum concentration that protects against your lowest setpoint — more than necessary reduces heat capacity and increases pumping effort without adding protection you need.
Can I just set a lower temperature if my load is too warm?
No. Lowering the setpoint reduces the machine’s capacity, so a chiller already struggling will struggle more. If the load is warm, the answer is capacity at the current setpoint, flow, or line losses — not a colder setting.
Does temperature range tell me anything about capacity?
Only indirectly. A range describes the coldest the machine can reach, typically at minimal load. The useful question is how many watts it can remove at the temperature you plan to run.
Summary
Sizing a small water chiller means reading a curve rather than a number. Capacity falls steeply as the setpoint drops — often to around a third of the headline figure by −20 °C — so a comparison between two machines quoted at different temperatures is not a comparison at all. Below freezing the fluid changes as well, and glycol’s lower heat capacity and higher cold viscosity reduce both the heat carried and the flow delivered, compounding the derating. Work in order: load, then setpoint, then capacity at that setpoint with 20–30% margin, then flow against your circuit’s actual pressure drop. And treat stability as a separate specification from capacity, because a machine with watts to spare can still hold temperature badly if it controls by cycling.