top of page

Fan-Inlet Airflow Measurement: Check K-Factor Units and Density Compensation

Oct 19
5 min read
Original violet fan-inlet measurement schematic showing paired pressure pickups connected to a differential-pressure instrument and a labeled K-factor calculation.

An AHU reports 5,000 m³/h from its fan-inlet pressure signal, while a separate airflow test indicates a higher value. Before changing fan speed or adjusting a stored coefficient, inspect the calculation behind the displayed flow. A nozzle factor belongs to a particular inlet geometry, pressure definition, unit system and reference density. The correct number in the wrong equation can produce a convincing but misleading result.


This article concerns pressure-based airflow sensing at a fan inlet. It does not apply a grille’s effective balancing area or a general duct fitting-loss coefficient. ebm-papst’s explanation of airflow measurement describes inlet-nozzle differential-pressure sensing and its sensitivity to air density. The original calculations below are deliberately independent of any named fan rating.


1. Identify the pressure pair and coefficient definition


A manufacturer may relate volume flow to the difference between a pressure upstream of the inlet and a pressure at a specified inlet-nozzle location. Those are the defined measurement points. AHU external static pressure, fan total pressure and a convenient nearby duct pressure difference are different quantities and cannot be substituted without supporting instructions.


For the coefficient convention used here, Qactual = K × √[Δp × (ρref/ρactual)]. Qactual is actual volume flow at the inlet condition, Δp is the specified positive differential pressure, ρref is the coefficient’s reference density and ρactual is density at the measurement location. The reference-density factor is inside the square root.


The ZIEHL-ABEGG centrifugal-fan application catalog illustrates this coefficient convention and explains that its nozzle factors depend on the specified inlet arrangement. Other suppliers can absorb density or conversion constants into their coefficient. Copy the complete applicable formula, not only the symbol K.


2. Work a fully declared hypothetical example


Assume a fictional approved coefficient K = 250 (m³/h)/√Pa, referenced to ρref = 1.20 kg/m³. The measured differential is 400 Pa and independently established actual inlet density is 1.00 kg/m³. The coefficient is an invented example input, not a value to use for an installed product. Assume the pressure points and inlet arrangement match its definition.


At the reference density, the uncorrected indication would be 250 × √400 = 5,000 m³/h. At the actual density, Qactual = 250 × √(400 × 1.20/1.00) = 5,477.2 m³/h, or 1.52145 m³/s. Lower density requires greater volume flow to produce the same specified nozzle differential under this relationship.


An independent inverse check gives Δp = (Qactual/K)² × (ρactual/ρref). Substituting 5,477.2 m³/h returns approximately 400 Pa. Multiplying 5,000 directly by 1.20 would instead give 6,000 m³/h. That applies a linear density ratio where this pressure-to-flow conversion requires a square root.


3. Distinguish density correction from reference-volume conversion


The 5,477.2 m³/h result is actual volume, not standard volume. Under the assumed constant-composition density basis, mass flow is 1.00 × 5,477.2/3600 = 1.52145 kg/s. Expressing that mass at the declared 1.20 kg/m³ reference density gives 1.52145/1.20 × 3600 = 4,564.4 m³/h reference volume.


Three different figures now exist: 5,000 from an uncorrected pressure calculation, 5,477.2 actual, and 4,564.4 referenced to the stated density. None should be relabeled merely “corrected flow.” Preserve the raw calculation and state the physical meaning of each output. If humidity is material, ensure density and mass-flow definitions describe the same moist-air composition.


Original pressure-to-flow calculation chain: 400 Pa with fictional K 250 and density ratio 1.2 gives 5477.2 cubic metres per hour actual; a 5 Pa bias creates 0.62 percent error at 400 Pa and 9.54 percent at 25 Pa.

Original example only: K=250 (m³/h)/√Pa, reference density 1.20 kg/m³ and actual density 1.00 kg/m³. Use the exact fan manufacturer’s factor, units and measurement arrangement.


Check whether the instrument or controller already uses live temperature and pressure compensation. Applying the square-root correction twice would overstate actual flow. Conversely, a controller labeled m³/h may still rely on a fixed reference-density assumption. Confirm its configured algorithm and product documentation rather than inferring behavior from the display units.


4. Keep pressure units attached to K


The example K accepts pascals. Entering 0.400 kPa as though it were 400 Pa produces only 173.2 m³/h after the same density correction, an error by a factor of √1000. Converting the pressure value is sufficient; alternatively the coefficient can be transformed consistently. Do not convert both and thereby duplicate the correction.


For output in m³/s with pressure still in pascals, K becomes 250/3600 = 0.069444 (m³/s)/√Pa. It then gives exactly the same 1.52145 m³/s. Coefficients stated for cfm and inches of water cannot be carried unchanged into the SI equation. Their numerical values encode the associated units.


Keep the original manufacturer sheet in the commissioning record, along with the selected model, inlet size, coefficient revision and controller entry. A rounded coefficient copied from another fan can look reasonable across many operating points while producing a persistent bias. Repeated stable readings do not establish that the stored coefficient is correct.


5. Check the low-pressure end separately


At fixed density and K, flow is proportional to the square root of differential pressure. Assume an illustrative positive measurement bias of 5 Pa. At a true 400 Pa, the flow overstatement is √(405/400) − 1 = 0.623%. At a true 25 Pa, it is √(30/25) − 1 = 9.54%. These are sensitivity calculations, not the accuracy specification of a particular sensor.


The same absolute pressure error matters much more near the bottom of the operating range. Include zero stability, instrument range, pressure tubing condition, pickup integrity and the manufacturer’s minimum usable measurement signal in the review. A nominal fan-flow turndown ratio does not automatically establish a usable airflow-measurement turndown ratio.


Do not remove a low-flow discrepancy by editing K until the display matches one preferred reading. A coefficient change might hide a pressure offset at one duty and create a different error elsewhere. Establish the cause with a suitable independent reference method and several approved stable operating points.


6. Verify the installed configuration and close the record


Inlet obstructions, guard arrangements and nearby walls can change the measurement relationship. Follow the current instructions for the exact fan and accessory arrangement; the cited catalog’s factors and limitations are product-specific. Any installation-derived coefficient needs a documented reference measurement and a defined scope of validity. It is not automatically portable to another AHU or a revised inlet.


For a fan array, retain each fan’s own coefficient, pressure and density basis, convert each valid signal to flow, then sum comparable flows. Do not add nozzle pressures as though they were airflow. Verify how the control system treats stopped fans, reverse flow or isolation devices under the manufacturer’s approved method.


If the independent test reports a different volume basis, use actual versus standard airflow reconciliation before interpreting the remaining discrepancy.


Record raw pressure, calculated density, formula, coefficient units, final output basis and independent test result together. Apply the project’s agreed acceptance method and uncertainty requirements separately. The paid Air Balancing (TAB) Toolkit provides a workbook and field handbook for traceable airflow results, instruments and deficiencies; use the fan supplier’s instructions for the nozzle conversion itself. The objective is a reproducible measurement chain before any balancing decision is made.



Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page