VAV System Starved at the End of the Run
You are balancing AHU-3 in a 12-story office building. Design supply is 12,000 cfm but the VAV terminal schedule sums to 12,600 cfm. The fan is on a VFD controlled to a duct static pressure sensor mounted roughly 15 ft from the fan discharge, in a straight run, setpoint 1.50 in. w.c. All boxes are calling for maximum cooling. At 60 Hz the fan delivers 13,200 cfm and the sensor reads 1.50 in. w.c. Boxes on floors 1 and 2 are running 105–115% of design, while boxes on floors 9–12 cannot reach design even with dampers wide open - several sit at 65–80%. Motor current is 34 A against a 40 A nameplate. The mechanical contractor tells you the fan is "fine" and wants the report issued as-is.
1. Is this a terminal-unit adjustment problem, a fan problem, or something else? Explain what the readings tell you.
Read the worked answer
Neither, in the simple sense. The evidence points to a control-location problem sitting on top of a schedule problem.
- The fan is moving 13,200 cfm, which is 110% of the 12,000 cfm air-side design and 105% of the 12,600 cfm box schedule. Air quantity at the fan is not the shortage.
- Flow is front-loaded: near boxes are over-served while far boxes are starved. That is the classic signature of a static pressure sensor installed too close to the fan. The sensor sees 1.50 in. w.c. early, the drive backs off, and the pressure actually available at the far end of the main is well below what the index box needs.
- Every box being wide open means the dampers have no authority left to fix the far end. Throttling near boxes would only make the fan back off further.
You cannot balance this. Report the measured condition, identify the index box, and document what static pressure the index box actually needs.
2. Describe the procedure you would use to determine the correct static pressure setpoint, with the numbers you would produce.
Read the worked answer
Work from the critical path, not from the sensor:
- Take manual control of the VFD and hold the fan at 60 Hz.
- Drive every VAV damper to 100% open (max flow override).
- Identify the index terminal - the one requiring the highest inlet static (typically the longest run/highest loss box, here in the floors 9–12 group).
- Slowly reduce fan speed until the index box delivers its scheduled airflow with its damper approximately 90–100% open. Note the fan speed, the static at the sensor, and the static at the index box inlet.
- Verify no other terminal requires more static than that condition.
- Release boxes to normal control, balance each terminal to design by damper position working from the index box back toward the fan, and re-check at the resulting setpoint.
Sanity check with fan laws: reducing 13,200 cfm to the 12,600 cfm box total implies speed x (12,600/13,200) = 0.955, i.e. about 57 Hz, static x 0.91 and power x 0.87. If the required speed to satisfy the index box is close to that, the fan and drive are adequate and the fix is sensor location/setpoint.
3. What do you put in the report, and what do you request in writing?
Read the worked answer
- Report the measured data as found: fan cfm, fan speed, sensor static, index-box static, and every terminal reading, flagged against design.
- Note the deficient terminals (the 65–80% group) with measured values and the reason they could not be brought to design at the available static - do not write "passes" or leave them unaddressed.
- Raise an RFI asking the engineer to confirm relocation of the static sensor to approximately two-thirds of the way down the main (or to a box-damper-position reset strategy), and to confirm the correct static setpoint. The sensor location is a design decision, not a TAB decision - you recommend, you do not relocate it unilaterally.
- Address the schedule overshoot: boxes total 12,600 cfm against 12,000 cfm fan design. The fan can physically do it, but someone must confirm the fan and motor are intended to deliver the sum of the boxes. Include motor amps vs. nameplate (34 A / 40 A) so the owner sees the margin being consumed.
- Record the condition in the deficiency log with a retest commitment once the sensor is relocated.