Low Delta-T Diagnosis in a Primary-Secondary Chilled-Water Plant
A 300,000 ft², 12-story office building has two 500-ton centrifugal chillers and a primary-secondary chilled-water plant. The secondary pumps are constant-speed, and the design supply-return ΔT is 12°F. Trend data show a 44°F supply, 51°F return, and a bypass valve 35% open at design flow. Most air-handling units still have three-way coil valves; 20 were converted to two-way but were never balanced. Chillers are staged manually, supply temperature reset is disabled, and the outdoor-air economizer has been locked out since a coil freeze. The plant runs 6,000 hours per year and electric demand costs $18/kW. Occupants complain of warm afternoon floors, while the plant operator adds a second chiller whenever the first reaches 90% load. A pump VFD retrofit has been proposed, but the owner asks whether it will solve the problem.
1. What are the likely root causes of the low ΔT, and what measurements would confirm them?
Read the worked answer
Low ΔT is a symptom, not a single fault. Likely causes are: three-way valves and an open bypass allowing water to return without delivering cooling; fouled or air-bound coils; two-way valves converted without balancing, so some coils are starved and others are over-flown; temperature sensors out of calibration; and manual chiller staging that keeps a second machine online at low load.
Confirm with a week of trend data: calibrated supply and return temperatures at each AHU, valve position, coil differential pressure, plant flow, bypass position, chiller amps, and load. Compare design flow to actual flow. A coil with a high ΔP and low ΔT is likely fouled or air-bound; a coil with a wide-open valve and low ΔT may be over-flown. Check sensor calibration in an ice bath or against a reference. Do not accept a VFD proposal until the hydraulic causes are separated from the control causes.
2. Recommend a corrective sequence for controls and retrofits. Why does the order matter?
Read the worked answer
Sequence the fix. First, calibrate sensors and close the bypass under control. Second, clean and air-vent coils, then replace remaining three-way valves with two-way and balance all AHUs to design flow. Third, restore supply-water temperature reset, stage chillers on total plant load rather than a single machine, and enable the economizer with proper freeze protection. Fourth, add VFDs to the secondary pumps and commission the control curve.
The order matters. A VFD on a low-ΔT plant may save little because flow stays high to meet load, and it can make coil starvation worse if the curve is poorly set. Fixing valves and coils raises ΔT, reduces required flow, and only then creates pump savings. The controls must maintain minimum flow and chiller evaporator limits.
3. Estimate the potential pump and demand savings from restoring ΔT, and state the main uncertainty.
Read the worked answer
Use the affinity law only after confirming the flow reduction is physically available. Suppose the two secondary pumps draw 180 kW combined at full flow. If valve and coil fixes plus balancing let flow fall to 70% at the same load, pump power falls to about 0.7³ = 0.343 of full power, or 62 kW, saving 118 kW. At 6,000 hours and $0.10/kWh, that is about 708,000 kWh and $70,800 per year. Demand savings at $18/kW add about $2,100 per month if the peak is coincident.
Do not add this to chiller savings without checking overlap. A second chiller that no longer runs at 40% load may save 0.3-0.5 kW/ton, but the exact savings need trend data. Present a range and require commissioning and M&V. The main risk is that low ΔT persists because coil cleaning was incomplete or sensors remain biased.