Start here
What does low chilled-water delta T mean?
Delta T is return-water temperature minus supply-water temperature at a defined measurement boundary. A lower-than-expected difference may reflect load, flow, mixing, airside performance, controls or measurement problems. Compare synchronized readings with the intended operating condition and relevant design information before calling it low-delta-T syndrome or approving a corrective change.
First identify what the temperatures actually describe
A chiller evaporator, distribution header and individual cooling coil are different measurement boundaries. Mark each sensor location on the available schematic. Do not combine a chiller leaving-water reading with a return reading from an unrelated branch and call the result the plant delta T.
For a cooling circuit, 44°F supply and 56°F return give a 12°F difference. Those numbers are an arithmetic example, not target temperatures. The reference for your system comes from its design, equipment selections and operating sequence. Keep the units explicit: a temperature difference of 12°F is approximately 6.7°C, not 12°C.
Why the same temperature difference can mean different things
Trane's chilled-water training identifies several possible contributors to low delta T, including water-flow imbalance, dirty filters or coils, air-handler controls and coil-control valves. It also explains that the plant consequence depends on system arrangement. That supports a system investigation rather than a diagnosis based only on the chiller display.
A comparison with design delta T needs load context. Record whether the building was starting up, lightly occupied, responding to a sudden load change or running steadily. Ask the engineer what delta T is expected under that condition. A design-point figure is not automatically the required value at every hour.
References: [1]
Annotated example: similar calculated load, different water flow
The following invented records illustrate how to read a trend. They describe the same hypothetical water-only distribution boundary, using synchronized, validated measurements. They are not an Airplex project, a design recommendation or evidence that a real pump should be adjusted.
| Record | Supply / return | Measured flow | Delta T and approximate load | What to investigate |
|---|---|---|---|---|
| A · 10:00 | 44°F / 56°F | 200 GPM | 12°F; 1,200,000 Btu/h | Establish the contemporaneous load, valve states and sensor locations. |
| B · 14:00 | 44°F / 50°F | 400 GPM | 6°F; 1,200,000 Btu/h | Ask why twice the water flow carried approximately the same heat. |
| C · 14:15 | 44°F / 50°F | Not recorded | 6°F; load unknown | Do not assume record B's load without the missing flow measurement. |
References: [1]
Ask for evidence that separates the possible explanations
Use the trend to frame questions for the responsible technician or engineer. If one branch is suspected, request evidence from that branch as well as the plant. A system average can conceal differences among coils or zones.
Record a control command separately from a measured result. A valve commanded closed is not proof of zero water flow; a pump-speed percentage is not a flow measurement. If the distinction matters to the diagnosis, the investigation should state how the physical condition will be confirmed.
| Question | Useful evidence | Conclusion to avoid |
|---|---|---|
| Are the temperatures trustworthy? | Sensor IDs, locations, timestamps and validation method | Treating an unexplained sensor difference as a system fault |
| Is heat reaching the coil? | Airside operating condition and relevant coil/air-handler findings | Assuming every low-delta-T event is a pump problem |
| Where is the water going? | Flow measurements and the documented piping arrangement | Equating pump speed with delivered GPM |
| What changed during the event? | Schedules, alarms, load changes and control trends | Attributing a transient event to a permanent capacity shortage |
Specify the finding you need before approving a remedy
A useful investigation deliverable identifies the affected boundary, confirms the measurements, explains the operating condition and connects a proposed cause to supporting evidence. It should state the proposed correction and how the result will be checked under a comparable condition.
Collect existing trends and records without changing live pump speeds, valve positions or plant sequences for this worksheet. Where operational testing is needed, it belongs in the site's authorized technical procedure. The flow calculator can check arithmetic for a stated load and temperature difference; it cannot establish why your system is behaving that way.
Put this guide to work
Low-delta-T measurement-context record
Use one row per existing reading or evidence item. Preserve the original timestamps and units; identify missing or unvalidated values before drawing conclusions.
Open each item to add your notes, then download the CSV to keep a copy. This tool does not save or submit your entries; do not rely on your browser restoring them.
Supply-water temperature
Paired return-water temperature
Measured circuit flow
Expected delta T and reference condition
Building load / occupancy / start-up condition
Air-handler or coil operating evidence
Relevant valve commands and verified positions
Alarms, overrides or recent system changes
Continue with a related decision
Questions & answers
Is a 6°F chilled-water delta T always bad?
No universal number answers that. Compare it with the system's expected behavior at the measured load, flow and operating mode. The same temperature difference can occur under different conditions.
Does low delta T prove the chiller is undersized?
No. Investigate the measurement boundary, load, water distribution, coils and controls before reaching a capacity conclusion. The example shows why a temperature difference alone cannot establish delivered load.
Can I use pump speed instead of measured flow?
Record speed as operating context, but do not label it GPM. A flow estimate needs a defensible method and system information; identify whether a value is measured, calculated or unavailable.
Sources & technical references
- Trane: Chilled-Water Systems, January 2021
Accessed November 16, 2026. Printed page 24 gives the water heat-transfer relationship and fluid-property limitation; pages 91–93 discuss low-delta-T causes and system effects. Used for physical principles, not current code requirements. Example values are original and hypothetical.
NYC & Northern New Jersey
Bring the readings and their context
Share the system arrangement, paired trends and known operating conditions with Airplex Mechanical to discuss the chiller, pump and related system scope that needs investigation.
Discuss a chilled-water investigationCall (347) 627-7146