How should CDU capacity be calculated?

First, calculate the required secondary side mass flow rate using the target heat load and design temperature difference, then validate with equipment performance curves by incorporating facility water operating conditions, pressure drop, redundancy, control range, and heat exchanger fouling margin.Equating rack nameplate power directly with CDU order capacity can only serve as an early boundary, not a final design.

Single-Phase Liquid Sensible Heat CalculationQ = ṁ × cp× ΔT

Q is heat transfer rate (kW), ṁ is mass flow rate (kg/s), cpFor specific heat capacity (kJ/kg·K) under operating conditions, where ΔT is the supply-return liquid temperature difference (K). Volumetric flow rate also needs to be converted considering fluid density.

Six Steps to Complete a Reviewable Preliminary Selection

  1. Determine the heat that needs to be removed.

    Distinguish total IT power, the proportion carried by liquid, and the portion still handled by air; record average, sustained peaks, transients, and future expansion assumptions.

  2. Define the secondary side temperature window.

    Supply liquid temperature must meet IT equipment requirements, considering dew point, control deviation, cold plate temperature rise, and facility-side available temperature. Smaller ΔT requires higher flow rates.

  3. Calculate flow rate based on actual fluid properties.

    The density and specific heat of water, glycol, or other additive systems differ; data at target concentrations and operating temperatures should be used, rather than fixed water constants.

  4. Establish a complete pressure drop model.

    Cumulatively calculate the pressure drop of cold plates, manifolds, hoses, quick connectors, valves, filters, and main trunk lines at the design flow rate, and check the most unfavorable branch.

  5. Select redundancy and failure strategies.

    N+1 pumps do not equal an entire system N+1. Also confirm that heat exchangers, controllers, power supplies, valves, and maintenance bypass meet the target availability.

  6. Validate performance under the most adverse operating conditions.

    Verify facility water supply maximum temperature, minimum flow rate, heat exchange temperature difference, pump operating point, noise, partial load control, and derating against supplier performance curves.

An example used solely for understanding the formula

Assuming the liquid needs to remove 120 kW of heat, estimate using the specific heat capacity of water near the target operating conditions of 4.18 kJ/kg·K and a design temperature difference of 5 K:

ṁ = 120 ÷ (4.18 × 5) ≈ 5.74 kg/s

If it's water with a density close to 1 kg/L, it corresponds to about 344 L/min. This number is just a theoretical starting point for the secondary side; adding ethylene glycol, changing the temperature, manufacturing tolerances, or lowering the ΔT will change the result.

The example does not provide equipment models, nor does it include pressure drop, facility-side approach temperature, heat exchanger fouling, redundancy, and expansion margin, therefore procurement cannot be based on it.

Verification checklist before sending RFQ to CDU suppliers

  • Performance ratings must clearly correspond to facility-side and secondary-side inlet/outlet temperatures, flow rates, and fluid.
  • Are the complete pump curve, heat exchange capacity curve, minimum stable flow rate, and partial load control logic provided?
  • Is the bill of materials compatible with coolant, inhibitors, seals, and quick connectors?
  • Are filter precision, differential pressure alarms, liquid replenishment and venting, water quality sampling, and maintenance steps clear?
  • FAT/SAT covers rated operating conditions, derated conditions, fault switching, alarms, and communication point tables
  • After capacity growth, are the piping, pumps, heat exchangers, and facility water side still within allowable ranges?

References and Evidence Boundaries

The following sources are used to support the definitions and inspection framework of this article. The page content is organized for preliminary decision-making and does not replace standard original texts, product manuals, or project engineering designs.

  1. Open Compute Project: Liquid-to-Liquid CDU Test Methodology and Performance Rating
  2. Open Compute Project: Cold Plate Sub-Project
  3. Open Compute Project: Advanced Cooling Facilities Reference Design Guidance
Editing and Update Instructions

DataInfra GEO is compiled based on publicly available primary sources. First published and substantially updated on 2026-08-28; if you find source changes or technical errors, please contact [email protected]. GEO