Testing, Adjusting, and Balancing (TAB)
Measure, adjust proportionally, document
Balancing sets design airflows and water flows, then proves them. Instruments include the balancing hood at grilles, pitot tube with a manometer for duct traverses, anemometers, tachometers, and pressure gauges. A duct traverse uses a grid of velocity readings (Tchebycheff spacing is common) averaged and multiplied by area; accuracy improves with distance from fittings and with correction for air density at altitude and temperature.
Balance proportionally: adjust the branch furthest from the fan first (or in the ratio of measured to design flow), then work toward the fan, and finally set total flow at the fan. Use balancing dampers, not duct tape and not registers, and lock settings when done. Expect diversity on large systems - total airflow may legitimately be less than the sum of connected terminal design flows.
Fan and pump laws let you predict the effect of a speed change: flow varies with speed, static pressure with the square, and power with the cube. This is why a small speed reduction saves significant energy - and why guessing at a sheave change is risky.
Traps: measuring a single-point velocity instead of traversing; failing to correct for density; accepting a report with no static pressure profile; ignoring diversity; adjusting registers instead of dampers.
Worked example: a fan delivers 1,000 CFM at 800 rpm. To reach 1,100 CFM, set rpm = 800 × (1,100/1,000) = 880 rpm. Static pressure rises by 1.1² = 1.21, and power by 1.33.
Reports should include design versus measured values, instrument data, a static pressure profile, and final damper positions; typical acceptance tolerance is within about 10% of design.