Chilled Water Coils Fighting Each Other on a Direct-Return Loop
A secondary chilled water loop serves three AHUs from a common supply main in a direct-return arrangement, with a manual balance valve at each coil. Design is 240 gpm per coil at 12 ft. w.c. coil pressure drop, 720 gpm total at 80 ft pump head. The pump is on a VFD whose differential pressure sensor is installed across the pump, and the drive is holding 78 ft. With all balance valves wide open you measure AHU-1 at 300 gpm, AHU-2 at 245 gpm, AHU-3 at 160 gpm, total 705 gpm. AHU-3 is nuisance-tripping on its low leaving-water temperature safety. The owner's rep says the far coil is starved for pump head and wants the pump speed increased.
1. Which coil is the index circuit, and is the far coil starved for pump head? Support your answer with the numbers.
Read the worked answer
AHU-3 is the index circuit - it is the highest-pressure-loss path (longest supply plus longest return in a direct-return layout).
But it is not starved for pump head: the pump is producing 78 of its design 80 ft and the loop is already passing 705 of 720 gpm design, about 98%. The total is essentially there; the distribution is not. AHU-1 is running 125% of design and AHU-2 about 102%, so the excess taken by the near coils is what the far coil is missing.
If you raise pump speed you will push AHU-1 and AHU-2 further over design, increase wear and noise, and gain very little at AHU-3, because the near-coil circuits are the low-resistance paths. The low-temperature trip on AHU-3 is consistent with low flow (large ΔT across the coil), which is a balance problem, not a head problem.
2. Give the balance sequence and explain what happens to total flow as you work through it.
Read the worked answer
Balance by adding resistance, not by removing it:
- Confirm the index coil valve (AHU-3) is fully open and record its flow - it stays wide open as the reference.
- Throttle AHU-1 toward 240 gpm. Total flow drops; with a constant-speed pump the operating point rides up its curve, so system head rises and AHU-3 flow increases.
- Throttle AHU-2 toward 240 gpm, then re-read AHU-1 and AHU-3.
- Iterate two to three passes until all three are within tolerance of 240 gpm - normally ±10% unless the specification is tighter.
Expect the final total to settle somewhat below 705 gpm because you have added resistance at two coils. That is normal; the question is whether the pump can still deliver the design total at an acceptable head.
Also verify the far coil itself before blaming hydraulics: air-bound coil, fouled strainer, stuck or mis-set valve, and manual air vents. A coil passing 160 gpm against 240 gpm design with the valve wide open and adequate differential is a physical restriction until proven otherwise.
3. What do you recommend regarding the pump DP sensor, and how does that change your balancing work?
Read the worked answer
A differential sensor reading across the pump is measuring pump head, not the critical circuit's need. On a variable-flow secondary loop it belongs across the index circuit (or the far end of the main), so the drive modulates on what the critical coil actually sees. As installed, the drive holds 78 ft regardless of what the coils need, which means the balance valves you set will be fighting the control loop every time load changes - your freshly set balance will drift as soon as the drive reacts.
Actions:
- Balance with the drive temporarily in manual at a fixed speed so the valves are set against a stable system.
- Document the sensor location as a deficiency and recommend relocation by the controls contractor.
- After relocation, re-verify flows at design, at 75%, and at 50% load to confirm the critical coil stays at design flow and the near coils do not over-flow.
- Report actual measured flows, the pump operating point (705 gpm at 78 ft vs. 720 gpm at 80 ft), and the coil ΔT for AHU-3 to document the low-temperature condition.