AHU Fan-Array Redundancy: Proving Airflow with One Fan Out
AI-generated editorial illustration; not a verified installation or construction detail.
Define the duty that must survive
An AHU submittal shows four fans and the description “N+1.” The useful next question is precise: what airflow will the installed unit deliver after one fan becomes unavailable, at the specified resistance and within every remaining fan’s limits?
Write the required surviving airflow, pressure basis, filter condition, air density, allowable recovery time and acceptable transient shortfall into the review. Identify whether all fans normally run or one remains on standby. These arrangements have different transition sequences. A requirement to maintain minimum ventilation is also different from a requirement to preserve full cooling airflow.
Nexora’s fan-curve and system-curve guide explains the basic intersection. The redundancy review repeats that exercise after changing the number of available fans and the failed module’s airflow path.
Use a common pressure across parallel fans
For identical fans operating stably in parallel, add their flows at the same pressure. Do not add their pressures. ASHRAE’s Fans chapter describes this combination and warns that unstable curve regions and installation geometry can undermine the ideal result.
Obtain both normal and one-fan-out selections from the supplier. They should include the actual array arrangement, speeds, power, motor or electronics limits and sound data. Match the pressure basis at defined boundaries; AHU external static pressure alone omits the internal resistance the fan assembly must overcome.
A synthetic four-fan example
The following curves are invented teaching data, not a manufacturer selection. Assume four identical fans, equal loading, constant density and a fully isolated failed module. Represent the available static-pressure rise between common low-velocity chambers by p = 1300(N/1500)² − 64q². Here p is Pa, N is rpm and q is each operating fan’s flow in m³/s; the numerical coefficients carry the corresponding units.
Assume the complete fixed-configuration system requires p = 9Q², with Q in m³/s. This simplified resistance curve includes the assumed AHU and external losses. It is valid only for this illustration; an actual AHU’s filters, coils and controls may not follow one square-law curve.
With four fans at 1500 rpm and total flow Q = 10 m³/s, each fan moves 2.5 m³/s. Available pressure is 1300 − 64 × 2.5² = 900 Pa. Required pressure is 9 × 10² = 900 Pa. The normal duty therefore balances.
Remove one fan and solve again
With three fans at the same speed, each carries Q/3. Equating available and required pressure gives 1300 − 64(Q/3)² = 9Q². Therefore Q = √[1300/(9 + 64/9)] = 8.983 m³/s, approximately 89.8% of design flow.
Losing one quarter of the installed fans has not removed one quarter of the delivered flow. The system pressure falls with flow, allowing each surviving fan to move more than before. That observation still falls short of proving full-duty redundancy.
At the original 10 m³/s duty, each survivor must move 10/3 = 3.333 m³/s while the system still requires 900 Pa. Solving 900 = 1300(N/1500)² − 64(10/3)² gives N = 1669.9 rpm. The required speed increase is approximately 11.3%, although each fan’s flow increases 33.3%.
Now assume the selected assembly is limited to 1600 rpm. At that speed, Q = √[1300(1600/1500)²/(9 + 64/9)] = 9.582 m³/s. The array cannot restore the full 10 m³/s in this model. A supplier would need to revise the selection or obtain approval for a lower degraded-mode duty. A computed 1670 rpm is not permission to exceed an equipment limit.
Original Nexora synthetic example. Four normal fans meet 10 m³/s; three limited to 1600 rpm deliver only 9.582 m³/s. Full recovery requires approximately 1670 rpm, above the assumed selection limit. Actual acceptance requires manufacturer data and installed verification.
Include the stopped fan’s reverse-flow path
A failed fan can connect the higher-pressure chamber back to the lower-pressure chamber. Surviving fan flow then includes recirculation rather than entirely useful supply. For example, 10.0 m³/s through the working modules with 0.4 m³/s returning through an inactive opening leaves only 9.6 m³/s net delivery.
Continental Fan’s array application guide explicitly includes increased surviving-fan speed and prevention of backflow in its redundancy discussion. Establish the installed isolation strategy during design. An opening that requires a technician to fit a cover cannot be credited with automatic, immediate recovery.
Where dampers provide isolation, verify their orientation, opening pressure, pressure loss, closing behavior and permitted leakage. Greenheck’s backdraft-damper guidance distinguishes gravity-operated and motorized arrangements. Include the selected device’s resistance in the normal and degraded selections. Do not improvise blanking or reach into an operating array.
Review common failures beyond the impeller
A four-fan arrangement can still share one electrical feeder, controller, network segment or pressure sensor. Make a failure matrix naming each shared item and the function lost when it fails. Decide which failures the project’s redundancy requirement actually covers; one motor failure does not demonstrate resilience to loss of the common supply.
Check what the controller calls “airflow.” The ebm-papst CN1116 controller manual describes an arrangement that can substitute assumed readings from healthy sensors after a sensor failure. This is product-specific, but it illustrates why a displayed total needs a known measurement and fault-handling basis.
Specify an observable recovery test
Build the functional test around an approved, manufacturer-compatible method of making one module unavailable. Qualified personnel should establish safe boundaries, preserve protective functions and set abort criteria before testing. Do not simulate failure by obstructing a running impeller or disconnecting energized equipment.
Trend independently verified net airflow, chamber pressure, each fan’s speed and status, and motor input or current as appropriate. Timestamp the fault, alarm, isolation action and restored duty. Record the minimum airflow during recovery and the time until the specified acceptable band is regained.
Test the required operating combinations and the specified resistance condition. A clean-filter demonstration cannot establish dirty-filter reserve without supporting selection data or an approved equivalent test. Verify the remaining fans stay within thermal, electrical, vibration and sound limits. Nexora’s fan commissioning guide supplies the broader TAB context.
Close the submittal with a short acceptance checklist
State the surviving duty and recovery-time requirement, including density and filter condition.
Submit normal and failed-module curves on the same pressure basis, with all enforced speed and power limits.
Identify the backflow barrier and account for its opening loss and closed leakage.
List shared electrical and control failures, the permitted degraded response and alarm routing.
Provide a safe test method, independent airflow evidence, trend records and agreed acceptance tolerances.
Record the permitted repair state. Continuing ventilation does not by itself make access to an isolated module safe.
The calculation is a screening tool. Final acceptance belongs to the complete installed unit, its manufacturer-supported operating envelope and the project’s defined failure scenarios. The strongest evidence is a documented duty that remains achievable when the specified failure occurs.
Related learning resource
For the field record, Nexora’s Air Balancing (TAB) Toolkit provides airflow-deviation, instrument-calibration and deficiency reporting tools. Use it alongside a project-specific failed-fan test sheet containing the recovery timeline and isolation checks above.
Sources and further reading
Engineering education only. Verify project conditions, applicable requirements and manufacturer data; use qualified professional judgment before design, selection, operation or maintenance decisions.
Comments