Low delta-T in a hospital chilled-water retrofit
A 1,200 TR chilled-water plant serves a 10-storey private hospital. The design is primary-secondary: three equal chillers, constant-speed primary pumps, and variable-speed secondary pumps. The differential-pressure sensor is installed across the secondary pump discharge header, not at the index coil. At 85% load the building operator reports all AHU control valves fully open, supply water 7.0°C, return 10.8°C (ΔT 3.8 K against a 6.7 K design), one chiller at 60% while another cycles, and a warm bypass line. The AHU coils are original three-way valve type. The hospital cannot isolate clinical areas, and a retrofit contractor is pressing to replace all AHU coils at a cost of MYR 1.8 million. You are the mechanical engineer asked to review the diagnosis before the board approves the expenditure.
1. Separate the plausible causes of the low plant delta-T into hydraulic/control causes and coil/heat-transfer causes. What evidence would confirm or exclude each, and why is wholesale coil replacement premature?
Read the worked answer
Hydraulic and control causes
- Bypass flow: a warm bypass line at part load suggests the decoupler is passing. Measure bypass flow; more than about 5% of plant flow at 85% load points to over-pumping or faulty control, not coil failure.
- Three-way valves: these keep coil flow constant and bypass hot return water around the plant. Check whether the bypass port is passing and whether valves are commanded open unnecessarily.
- DP sensor location: sensing at the pump header makes the pumps maintain header DP, so at low load they over-pressurise the distribution and force flow through the bypass. Compare measured DP at the index coil with the setpoint; a large excess is diagnostic.
- Sensor and sequencing faults: compare calibrated temperatures and check chiller staging; one chiller at 60% while another cycles indicates unstable control.
Coil/heat-transfer causes
- Fouled or air-bound coils show a high coil approach (air-off minus chilled-water supply) and often reduced water-side ΔT with the valve wide open. A clean coil at design flow should recover the design approach.
Why replacement is premature The plant ΔT is low while valves are fully open, which is the classic signature of excess flow and bypass rather than uniformly fouled coils. Replacing every coil without isolating the hydraulic cause would leave the bypass and DP problem in place. Test a representative sample of coils first; only coils whose approach exceeds design after flow correction justify cleaning or replacement.
2. Propose a measurement and verification plan, including instrumentation and acceptance criteria, before any coil replacement is approved.
Read the worked answer
Instrumentation and data
- Calibrated immersion temperature sensors in the supply, return, and each coil branch, with an accuracy target of ±0.2 K; clamp-on ultrasonic flow meters on the plant, bypass, and selected coils.
- Pressure gauges or transducers at the pump header and index coil; valve-position feedback; chiller power and staging status.
- Trend at 15-minute intervals for at least one stable 85% load period, plus a controlled test with the bypass isolated and the DP sensor temporarily referenced to the index coil.
Tests
- Confirm sensor calibration by comparison with a certified reference.
- Measure bypass flow with all AHUs in normal control.
- Record plant ΔT, chiller load, pump speed, and index DP simultaneously.
- For 5–10 coils across the building, record air-on/air-off and water in/out temperatures and water flow; calculate coil approach and capacity.
Acceptance criteria before capital spend
- Plant ΔT at 80–85% load should recover to at least 6 K, with bypass flow no more than 5% of plant flow.
- Index DP should be maintained at setpoint without excessive pump speed.
- Coil approach should be within about 1 K of design for at least 80% of sampled coils; only outliers justify cleaning or replacement.
- If these criteria are not met after control and bypass corrections, then revisit coil condition with a targeted replacement list.
3. Recommend a staged remediation sequence with risk controls suitable for a live hospital.
Read the worked answer
Stage 1 – Control and hydraulics (low disruption)
- Relocate or re-reference the DP sensor to the index coil, or implement valve-position reset; lower the DP setpoint gradually and trend coil performance.
- Close or automate the bypass; verify it seats fully. Convert or lock open three-way valves only after confirming freeze protection and minimum flow requirements.
- Re-balance secondary branches and verify pump curves against actual flow.
Stage 2 – Targeted coil work
- Clean or replace only coils whose approach remains high after Stage 1. Prioritise clinical critical areas and schedule during low-occupancy periods.
- Use infection-control measures: containment, negative-pressure containment where required, HEPA filtration, and coordination with hospital infection control.
Stage 3 – Verification and handover
- Repeat the M&V test at stable load; update the sequence of operations, commissioning records, and O&M manuals.
Risk controls Maintain chiller redundancy, provide temporary cooling for critical areas, use night or weekend works, and define a rollback plan if clinical temperatures exceed limits. Do not allow the contractor to proceed to wholesale coil replacement until Stage 1 data are reviewed.