Cold Plunge Chiller HP vs Cooling Capacity: What Buyers Should Compare

HP and Cooling Capacity Answer Different Buyer Questions

A supplier data sheet may place HP, cooling capacity, input power and COP in one specification table. Their proximity does not make them interchangeable.

Horsepower is a descriptor, not delivered cooling

Mechanical horsepower is a unit of power, and one mechanical horsepower is approximately 746 watts. That conversion does not mean a 1 HP compressor produces 0.746 kW of cooling. Refrigeration equipment moves heat through a vapor-compression cycle. The useful cooling output depends on compressor design, refrigerant circuit, evaporator and condenser heat exchangers, fan performance, controls, water flow, water temperature and condenser-intake condition.

The buyer must therefore ask what the HP field names. It may refer to a compressor motor class, a nominal compressor designation, a pump motor, or a product-series label. If the supplier cannot define it, retain HP only as an identification field. Do not use it as the contractual output.

Cooling capacity is useful output at a stated condition

Cooling capacity is the rate at which the defined system removes heat. It may be reported as watts, kilowatts, BTU/h or refrigeration tons. Unit conversion is straightforward: 1 kW is approximately 3,412 BTU/h. Comparability is not straightforward. A 2.5 kW figure at warm entering water and moderate ambient air may not remain 2.5 kW as water approaches a low cold-plunge setpoint or as condenser intake rises.

The value is incomplete unless the record identifies at least:

• water temperature or entering/leaving water conditions;
• condenser-intake ambient temperature;
• water flow through the heat exchanger;
• electrical supply and control state;
• whether the pump, fan and auxiliaries are inside the test boundary;
• whether capacity is measured, calculated, simulated or nominal.

Input power describes demand, not output

Total electrical input is needed for site coordination and efficiency comparison. A compressor-only input and a complete-system input are different boundaries. If one quotation includes condenser fan, circulation pump and controls while another lists compressor input only, their input figures and COP values cannot be compared directly.

Nameplate maximum current also does not equal average operating power. It supports electrical protection and conductor selection by a qualified professional, but it does not by itself predict energy use over a changing pull-down or holding cycle.

COP compares output with input only inside the same boundary

For cooling, a simplified coefficient of performance is:

COP = useful cooling output / electrical input

If useful output is 2.4 kW and complete measured input is 1.0 kW at the same stable condition, the calculated COP is 2.4. If the 1.0 kW excludes a 0.2 kW pump that the other supplier includes, the comparison is biased. A single COP point also does not describe performance across a complete pull-down, low-water-temperature operation or hot-ambient condition.

Define Three Duties Before Comparing a Chiller

The project does not have one cooling requirement. It has at least three. A unit can pass one and fail another.

Pull-down duty

Pull-down is the planned reduction from an initial water temperature to a target temperature within an agreed time. The basis must identify actual operating water volume, not only the vessel's external dimensions or gross fill volume. It must also state whether the cover is on, whether the system starts after a fresh fill, and the design ambient condition.

Pull-down is often a commissioning or start-of-day requirement. A commercial venue that fills or warms the vessel between operating periods may care about it strongly. Another venue that holds temperature continuously may be governed instead by recovery and holding duty.

Holding duty

Holding duty is the heat that enters while the system maintains a temperature band. Sources can include the open water surface, vessel and pipe heat gain, solar exposure, warm air, pump heat, nearby equipment and repeated cover removal. The load changes through the day. A covered overnight test is not evidence for an uncovered afternoon operation unless the buyer accepts that boundary.

Holding also depends on control behavior. The system may cycle inside a setpoint band instead of producing constant output. Record the controller's measured variable, sensor location, start/stop logic and any minimum run or anti-short-cycle behavior that materially affects the test.

Post-use recovery duty

Recovery is the system's ability to return to an agreed temperature band after user sessions or another defined heat event. "Users per day" is too weak for this decision. Ten users spread across ten hours create a different recovery problem from ten users in one hour.

Define the session block, number of entries, approximate exposure pattern where operationally relevant, cover state, mixing period and required recovery endpoint. Do not convert a user's body heat into a universal load assumption. Where project data are unavailable, measure the bulk-water temperature rise during a representative, controlled operating period and retain the record as a project input.

Before a supplier recommendation, the buyer should complete a duty-basis record.

Calculate the Water Load, Then Expose the Assumptions

A transparent calculation is not a promise of performance. Its purpose is to show which inputs control the decision and to reveal where supplier data must replace an assumption.

Step 1: calculate the thermal energy removed from the water

For a first-pass calculation using water without a significant additive concentration:

Q = m x cp x Delta T

where:

• Q is thermal energy, in kJ;
• m is water mass, in kg;
• cp is the assumed specific heat capacity, in kJ/(kg K);
• Delta T is the water temperature reduction, in K or degrees C.

For early screening, water mass in kilograms is often approximated from volume in liters using approximately 1 kg/L, and cp is often approximated as 4.186 kJ/(kg K) near the relevant liquid-water range. The final calculation should use an appropriate property basis if additives, unusual temperatures or higher precision matter.

Illustrative input set:

• operating volume: 300 L, approximated as 300 kg;
• initial temperature: 25 C;
• target temperature: 10 C;
• temperature reduction: 15 K.

The water-only energy is:

Q = 300 kg x 4.186 kJ/(kg K) x 15 K = 18,837 kJ

To convert kJ to kWh, divide by 3,600:

18,837 kJ / 3,600 = 5.23 kWh of thermal energy

This is not the electrical energy use. It is the heat that must be removed from the water under the stated approximation. It excludes heat entering during the test and thermal energy in the vessel, pipework and connected components.

Step 2: separate gross chiller capacity from net capacity available to the water

If a condition-linked chiller output averages 2.4 kW during the relevant part of the pull-down and ongoing heat entering the defined water system averages 0.4 kW, the simplified net capacity is:

Net capacity = 2.4 kW - 0.4 kW = 2.0 kW

The idealized time based on the water-only energy is:

Time = 5.23 kWh / 2.0 kW = 2.62 hours

That result is an illustration, not a forecast for a product. The 2.4 kW output and 0.4 kW heat-gain assumptions must be replaced with condition-linked data or a measured curve. If chiller capacity falls as water temperature falls, using one average value without a performance map can understate the elapsed time.

Step 3: treat capacity as a curve when conditions change

During pull-down, water temperature changes continuously. Ambient air and solar exposure may also change. A stronger model divides the pull-down into intervals and uses capacity appropriate to each interval:

Delta time_i = thermal energy removed in interval_i / net capacity_i

The total time is the sum of all intervals. A measured time-series test can be more reliable than an elaborate calculation when the exact configuration exists, provided the test boundary matches the project and the instruments and data treatment are controlled.

Step 4: quantify recovery from a measured bulk-water rise

Assume the same illustrative 300 L system rises by 1.2 K during a defined session block after the water is mixed and the measurement is stable. The added water energy is approximately:

Q_recovery = 300 kg x 4.186 kJ/(kg K) x 1.2 K = 1,507 kJ

1,507 kJ / 3,600 = 0.419 kWh thermal

At an illustrative net recovery capacity of 2.0 kW, the water-only ideal recovery time is:

0.419 kWh / 2.0 kW = 0.210 hours, or about 12.6 minutes

Actual recovery can be longer because the capacity and heat gain vary, the controller may cycle, and the temperature reading may not represent fully mixed water. The value is useful only when the session definition, measurement point and endpoint are recorded.

For the full equipment-selection model, continue to the cold plunge thermal engineering and chiller sizing guide.

Measure Useful Water-Side Capacity Instead of Inferring It from HP

Two measurement routes are practical for procurement and commissioning: a steady water-side heat-balance test and a complete-system pull-down test. They answer related but different questions.

Steady water-side method

Where stable flow and a measurable chiller inlet-to-outlet temperature difference are available, useful water-side capacity can be estimated by:

Cooling rate = mass flow x cp x (T_in - T_out)

For an illustrative pure-water approximation:

• measured flow: 30 L/min, approximately 0.50 kg/s;
• chiller inlet-to-outlet temperature difference: 1.0 K;
• assumed cp: 4.186 kJ/(kg K).

Then:

Cooling rate = 0.50 kg/s x 4.186 kJ/(kg K) x 1.0 K = 2.09 kW

This method can fail quietly when the temperature difference is small. If two sensors each have material error relative to a 1.0 K difference, the calculated capacity can shift substantially. Use paired or characterized sensors, consistent immersion and placement, stable flow, synchronized timestamps and an uncertainty review appropriate to the decision. A display that shows temperatures to 0.1 C does not prove 0.1 C accuracy.

Mass flow should be measured or defensibly derived at the actual operating point. A pump label or unrestricted nominal flow does not establish flow through a circuit containing hoses, elbows, valves, filters and a heat exchanger.

Complete-system pull-down method

A pull-down test measures the result of the installed system, including control behavior and ongoing site loads. Record:

• verified operating water volume;
• water temperature at a representative mixed location;
• chiller inlet and outlet temperatures where available;
• condenser-intake air temperature, not only room or weather data;
• water flow or a validated flow indicator;
• total electrical input where efficiency is being evaluated;
• cover, solar and door/enclosure state;
• filter condition and clean commissioning baseline;
• start time, endpoint and data exclusions;
• controller setpoint, mode and alarm history.

Use a sampling interval short enough to show cycling and changing slope. State the interval in the protocol rather than applying one universal interval to every instrument and controller.

Mixing and sensor location control the meaning of the result

A sensor near the cold return can report a lower temperature than the bulk vessel. A sensor near a warm surface or user zone can report higher. Agree whether the acceptance endpoint is a mixed bulk-water temperature, a controller reading or another defined measurement. For a mixed endpoint, document how circulation and any manual mixing are controlled before the reading is accepted.

Data traceability must match the decision risk

NIST describes metrological traceability as a documented unbroken chain of calibrations, each contributing to measurement uncertainty. A field acceptance test does not automatically require a national laboratory to be present, but the buyer should know instrument identity, calibration or verification status, range, resolution, placement and time synchronization. The required rigor should scale with contract value and dispute risk.

Data-quality hold: Do not use the result to reject or approve equipment when operating volume, temperature location, flow status or condenser-intake condition is unknown. Close the data gap or repeat the test under a controlled boundary.

Normalize Supplier Quotations into One Evidence Record

The buyer's comparison sheet should preserve what each supplier actually stated. Do not silently convert unmatched data into an apparently uniform table.

Required capacity fields

For every capacity point, request:

1. useful cooling output and unit;
2. water entering/leaving or operating temperature;
3. condenser-intake air condition;
4. measured or required water flow;
5. electrical supply;
6. total-input boundary;
7. control state and test duration or stabilization rule;
8. source: measured, calculated, simulated or nominal;
9. model and configuration revision;
10. applicable operating limits and protection conditions.

If a supplier provides only HP and a maximum tub volume, mark cooling capacity as not evidenced. If a supplier provides capacity without test conditions, mark the conditions as open. Neither omission proves poor equipment, but neither supports a condition-specific project decision.

Convert units, not conditions

BTU/h can be converted to kW and vice versa. Fahrenheit temperature differences can be converted to Kelvin or Celsius differences. Those conversions do not normalize different entering-water temperatures, ambient conditions, flow or included auxiliaries.

When conditions differ, use one of three routes:

• request capacity at the buyer's common comparison point;
• use a supplier performance map or approved interpolation method;
• retain the data as unmatched and make the uncertainty visible.

Do not create a derating factor without a technical basis.

Define the COP boundary

Request both useful output and total input at the same condition. List compressor, fan, pump, controls and any heater separately where possible, then state which are included in the COP. If a pump is site-supplied, the product COP and installed-system COP are different decisions.

Before requesting a model recommendation, assemble the operating volume, temperatures, time, ambient, flow and electrical data. A supplier can then respond to a defined duty instead of guessing from HP.

Grade Evidence Before Relying on a Performance Claim

Evidence quality has two dimensions: strength and scope. A strong record for one model at one condition does not automatically apply to every model or site.

Evidence must match the ordered configuration. A capacity report for a different heat exchanger, compressor, pump, power supply or control revision should be treated as a gap until the supplier explains applicability in writing.

An official standard can define a rating method or equipment scope; it does not prove that a product has been tested or certified to that standard. AHRI maintains performance-rating standards intended to make equipment specifications and evaluations more consistent. The applicable standard, its scope and any related certification must be verified for the exact cold-plunge chiller model and destination market. A standards-directory reference must never be used as decoration or implied certification.

Account for Site, Hydraulic and Control Losses

A valid factory rating can coexist with weak field performance. The next task is to determine whether the field boundary matches the rating boundary.

Air-side conditions

An air-cooled condenser must reject the heat removed from the water plus compressor work. High intake air, blocked coil area or discharge recirculation can reduce available capacity or trigger protection. Measure air where the condenser takes it in. A room thermostat several meters away may not capture the microclimate inside a cabinet or behind a screen.

Typical installation planning example

• Problem: A chiller with condition-linked capacity data misses the pull-down endpoint after decorative screening is installed.
• Likely cause: Hot condenser discharge air returns to the intake, so the equipment operates above the specified ambient condition.
• Evidence check: Log condenser-intake and discharge temperature, inspect airflow direction and compare clearances with the approved installation instructions.
• Prevention/decision: Restore the approved air path and repeat the controlled test before changing capacity or rejecting the unit.

Water-side conditions

The heat exchanger needs flow within its approved range. Pipe diameter, total length, elbows, valves, filter loading, elevation effects where applicable and air in the circuit change the operating point. More pump nameplate flow does not guarantee more circuit flow, and more flow is not always better if it exceeds component limits.

Establish a clean-filter commissioning baseline. Later performance can then be compared with pressure, flow or other approved indicators. If the filter is blocked or the circuit contains air, restore the hydraulic condition before interpreting a refrigeration-capacity shortfall.

Controls and electrical boundary

Controller sensor position, deadband, anti-short-cycle logic, minimum run time and alarms affect the observed curve. Voltage and frequency must match the equipment and destination. Fixed electrical work, protective devices, grounding, bonding and local code compliance require a qualified local professional. A performance test must not bypass a safety control to force continuous operation.

Use the dedicated guides to evaluate ambient-temperature effects on pull-down and duty-cycle and oversizing trade-offs.

Assign Responsibility and Hold Points Before Purchase-Order Release

Performance disputes often begin as ownership gaps. The supplier assumes the site will provide flow and airflow; the installer assumes the equipment selection already accounts for the circuit; the buyer assumes the headline capacity covers the operating schedule. A responsibility matrix makes those assumptions visible.

Use four hold points:

1. Design-input hold: operating volume, temperature duty, ambient, use schedule, utilities and responsibility are approved.
2. Technical-submittal hold: model-specific capacity, flow, input boundary, controls, drawings and evidence scope are accepted.
3. Shipment hold: ordered configuration, revision, labels, manuals, test records and packing scope match the approved submittal.
4. Commissioning hold: installation, flow, airflow, electrical records, sensor setup and acceptance protocol are ready before performance is judged.

No party should release a hold point by replacing missing technical evidence with an unsupported HP assumption.

Write an Acceptance Test That Can Survive a Dispute

"Chiller reaches 10 C" is not an acceptance test. It lacks initial condition, water volume, time, ambient, cover, flow and measurement definition. A useful protocol must allow another competent person to understand why the test passed or failed.

Minimum acceptance protocol

Record and agree:

• exact model, serial or order identity and configuration revision;
• operating water volume and how it was established;
• initial, target and allowable temperature band;
• measurement locations and instrument IDs;
• condenser-intake ambient range;
• water flow range and evidence method;
• cover, solar, enclosure and room state;
• clean-filter and air-purged baseline;
• electrical supply and total-input boundary where measured;
• controller mode, setpoint and alarm treatment;
• start rule, endpoint rule and stabilization/mixing rule;
• data sampling interval and permitted exclusions;
• witness and sign-off roles;
• deviation, retest and change-control process.

Do not erase deviations

If ambient exceeds the agreed band or flow falls outside the approved range, retain the run as diagnostic evidence but do not relabel it as the contract acceptance test. Record the deviation, assign an owner and decide whether to correct and retest or formally accept a revised boundary.

Split product, site and data causes before assigning liability

An acceptance failure can originate in:

• product output below condition-linked evidence;
• site airflow, hydraulics, electrical supply or heat gain outside the approved basis;
• incorrect operating volume or test setup;
• sensor, flow or power data with inadequate quality;
• a change in configuration, controls or operating schedule.

The first decision is not who pays. It is which evidence can separate these causes.

Illustrative acceptance dispute

• Problem: A supplier reports 2.5 kW capacity, but the buyer's field calculation shows 1.9 kW.
• Cause: The supplier point used warmer entering water, lower condenser-intake temperature and a declared heat-exchanger flow. The field estimate used a small inlet/outlet Delta T, an unverified pump flow and included site pump heat.
• Evidence check: Align water and ambient conditions, measure actual flow, review sensor uncertainty, define auxiliary boundaries and compare the exact configuration.
• Prevention/decision: Put the common test point and evidence method in the technical submittal before purchase-order release. Until boundaries match, label the figures unmatched rather than declaring either party wrong.

Diagnose a Failed Test Without Unsafe Service Work

Operators and buyers can review records and visible installation conditions. Refrigerant-circuit work, fixed wiring, protection changes and internal electrical diagnosis belong to qualified personnel.

Stop the test and use qualified service when there is a refrigerant leak indication, repeated high-pressure or electrical protection trip, abnormal sound with loss of flow, water near unsafe electrical interfaces, damaged wiring, overheating, smoke or any condition prohibited by the model manual. Do not bypass a switch or protection control to complete an acceptance run.