Why is a fan delivering less air? AMCA system-effect diagnosis

Why does an installed fan miss its selected airflow?

Match model, wheel, rotation, accessories, RPM, density and pressure definition before comparing duty. ANSI/AMCA 210-25/ASHRAE 51-25 governs laboratory aerodynamic ratings; AMCA Publication 203-90 (R2011) addresses field measurement; Publication 201-23 addresses fan/system interaction. S1, S2, S3.

Read the installation category on the rating. ANSI/AMCA 210-25/ASHRAE 51-25 distinguishes A, free inlet/free outlet; B, free inlet/ducted outlet; C, ducted inlet/free outlet; and D, ducted inlet/ducted outlet. Check whether guards, dampers, inlet boxes and transitions were included in the selected assembly. S2.

First establish whether total fan output is low. A fan meeting its required CFM, or cubic feet per minute, can still leave rooms short because of imbalance, leakage or recirculation through an idle fan. Use the TAB report guide and duct-leakage guide to follow that air downstream.

What is system effect?

System effect is the installed performance consequence of disturbed flow at the fan interface compared with the rated setup. Ordinary duct friction and fitting losses remain separate accounting items; the additional interaction between a nearby fitting and the fan profile requires its own treatment. Count each allowance once, including corrections already included by the manufacturer. S1, S4, S5.

Is the problem geometry, resistance, speed or the measurement?

Check rotation, actual wheel RPM, belt condition, commanded mode, filters, coil condition, damper positions and closed access panels before assigning a system-effect loss. A VFD frequency display does not establish wheel speed on a slipping belt drive. The belt and sheave guide and AHU components guide support those checks. S6.

Compare each component differential with its published pressure-drop curve at the measured airflow. A dirty coil changes resistance; inlet swirl changes the aerodynamic conditions at the blades. High static pressure or low motor current alone cannot distinguish them. S4, S6.

A constant-airflow controller can conceal loss by raising RPM to its limit; log RPM and command together. Confirmed low duty plus adverse geometry supports a matched-speed correction trial. Resolve conflicting airflow measurements first.

Is the fan stalled, or are the controls hunting?

Trend flow, pressure, command and RPM when noise or duty pulsates. A stable command with oscillating duty directs attention to aerodynamic instability; an oscillating command also requires control-loop diagnosis. Check the complete manufacturer curve and stable selection region through every operating state, including staged parallel operation. S14.

Why does acceptable TESP fail to prove fan performance?

Total external static pressure is the signed outlet static minus inlet static across the specified equipment boundary. That boundary can include coil or filter effects differently from a bare-fan rating. Use the external-static guide to establish the equipment boundary and ESP basis before comparing it with fan performance.

Installation geometry matrix

Photograph and dimension both connections. Sketch rotation, elbow direction, internal free area, first disturbance and the OEM outlet-to-blast-area relationship. Compare configurations with AMCA's 2023 troubleshooting presentation. S6, S8, S13.

D in the outlet-recovery rows means outlet duct diameter; the plenum clearances use wheel diameter. Twin City Fan FE-3500 (2018), Figure 1, recommends inlet clearance to satisfy both the one-wheel-diameter minimum and the approach-convergence condition. Anvilfield estimates 54 in. clearance for a 144-in. approach height and 36-in. wheel: subtract wheel diameter from approach height and divide by two, then apply the wheel-diameter minimum. S7.

Method: the manufacturer's 30,000-CFM, 4-ft takeoff example gives 12.57 ft² and 2,387 fpm. Its velocity pressure before display rounding is about 0.355 in. wg; multiply by the applicable takeoff factor. S7.

How do inlet swirl and obstructions reduce airflow?

According to ASHRAE's 2025 Fundamentals chapter 21, nonuniform fan inlet flow changes fan performance and cannot be treated as a simple percentage reduction in airflow and pressure. The blades receive unequal velocities and approach angles, so a downstream-resistance correction alone does not describe the changed fan characteristic. S4.

Twin City Fan's FE-100 (2018) describes prerotation in the wheel's direction reducing developed pressure and power, while opposite-direction swirl can substantially increase power. Low current therefore does not establish a clear inlet. Record swirl direction relative to wheel rotation, rather than writing only 'turbulence observed.' S6.

Inspect elbows, drop boxes, asymmetric entries, collapsed flexible connectors and inlet-throat probes. A measuring device in the high-velocity throat can itself obstruct airflow. S6.

Will straighteners or an inlet bell fix the problem?

Straighteners address spin; an inlet bell smooths open-inlet entry. Choose the correction that matches the defect and include its own resistance and required clearance. The measured-case table shows successive straightening and probe corrections. Retain approved permanent airflow measurement and required guards after the repair. S6, S7.

How much outlet duct is needed for static regain?

AMCA's 2020 inMotion guidance recommends 2.5 outlet duct diameters at velocities up to 2,500 fpm, adding one diameter for each additional 1,000 fpm to obtain effective outlet recovery length. Greenheck's FA/101-99, Understanding Fan System Effects, ©2017, gives the same application rule. The geometry matrix shows its velocity-dependent lengths; this is recovery guidance rather than a building-code clearance. S5, S9.

Static regain is conversion of velocity energy into static pressure as the discharge profile develops. Total pressure still decreases through passive ductwork because of losses, so a higher downstream static reading does not show that the duct added energy. AMCA's 2020 article places approximately 80% of regain at about half the effective outlet length; that observation concerns pressure recovery, not recovered airflow. S4, S5, S6.

Measure straight length to the first disturbance. Compare a nearby discharge elbow with scroll rotation and the manufacturer's supported arrangement; unfavorable orientation compounds recovery loss. S6.

Which diameter and length should I use?

For the rectangular diagnostic layout here, use equal-area diameter: the diameter of a circle with the same internal area. Anvilfield estimates 3.33 ft² and 2.06 ft equal-area diameter for a 24 by 20 in. duct using the TSI-106 method. Use fan outlet area for recovery length and local duct area for traverse spacing; recalculate after a size change. S8.

Greenheck's one-wheel-diameter inlet extension improves the specific elbow arrangement pictured in its guide; Twin City Fan separately recommends one wheel diameter of minimum plenum inlet clearance. Neither inlet recommendation replaces the velocity-dependent outlet recovery rule. S7, S9.

Where and how should airflow be traversed?

TSI Application Note TSI-106 Rev. C (2021) recommends an airflow traverse at least 7.5 duct diameters downstream and 3 diameters upstream of disturbances, with individual Pitot pressures converted to velocities before averaging. Its 2D-after/1D-before allowance impairs accuracy. Fluke's undated live guidance prefers 10D/3D and describes 5D/1D as potentially adequate. Preserve the selected method and available distances in the report. S8, S10.

The plan uses Fluke's 10D/3D spacing and TSI's five-line setout in each direction. Mark ports and the insertion-depth origin for repeatable return measurements. S8, S10.

Record temperature, absolute station barometric pressure, probe coefficient and calibration; zero the instrument. For a unity-coefficient Pitot tube, multiply the standard-density velocity by the square root of reference density divided by actual density. Then average point velocities and multiply by internal area; retain probe orientation and reading stability. S3, S6, S8.

What if no straight traverse location exists?

Document the available lengths and observed profile, then use another suitable plane, an approved calibrated station or an agreed field method. Compare an independent airflow measurement under the same system state and reconcile leakage between stations. FE-100 reports static readings varying by 20% across one swirling plenum; use the OEM's approved plenum pickup location rather than importing that case variation as a tolerance. S3, S6, S8.

How do you reconcile measured duty with the fan curve?

Mark the selection axis as fan total pressure, fan static pressure or equipment ESP, then identify configuration and included corrections. Under ANSI/AMCA 210-25/ANSI/ASHRAE 51-25, fan static pressure is fan total pressure minus outlet velocity pressure, so the difference between two duct static readings does not by itself establish fan static pressure. S2.

Pressure method: add velocity pressure to signed static pressure at each boundary, subtract inlet total from outlet total, then subtract outlet velocity pressure for fan static. In symbols, P_tF = (P_s2 + P_v2) - (P_s1 + P_v1); P_sF = P_tF - P_v2. For remote stations, add both intervening total-pressure losses to the remote total-pressure difference. S2, S3, S4.

Normalize to one RPM and density using the fan laws for the same wheel at corresponding points. S6, S11.

Normalization method: multiply measured volume by reference RPM divided by measured RPM, then apply the density ratio and squared speed ratio to pressure and the cubed speed ratio to shaft power. In symbols, r = N_ref/N_meas; d = density_ref/density_meas; Q_ref = Q_meas r; P_ref = P_meas d r²; BHP_ref = BHP_meas d r³. Use the unrounded ratios 1,200/1,080 and 0.075/0.060 for the table's separate worked point. S6, S11.

Do hot air and altitude change airflow or just pressure?

At unchanged wheel size and RPM, corresponding-point volumetric airflow stays the same while pressure and shaft power scale with density. Temperature, elevation, humidity and gas composition set actual density; controls or a fixed pressure requirement can move the operating point. Keep actual inlet volume separate from standard-density volume. S2, S6, S14.

Anvilfield estimates that 10,000 actual CFM at 0.060 lb/ft³ equals 8,000 standard CFM at the stated reference density of 0.075 lb/ft³: multiply actual volume by actual density divided by reference density. The same density mismatch in an uncorrected Pitot conversion reads velocity 10.56% low, calculated from the square root of 0.060/0.075. Resolve this conversion before assigning a geometry penalty. S6, S8.

What does a real duty reconciliation look like?

FE-100 assigns the residual below revised expectation to test error and upstream-elbow effect; the full design deficit is not a geometry loss. S6.

Resistance-curve method: multiply each anchor pressure by the squared flow ratio, holding density and geometry constant. In symbols, P_original = 1.70(Q/15,000)² and P_installed = 1.68(Q/13,600)². Anvilfield estimates 2.04 in. wg at design flow on the unchanged installed curve. Moving VAV dampers or terminal pressure requirements need the actual control-dependent relationship; use the VAV commissioning guide to define those states. S6.

Accept against the project flow band, ventilation duty and equipment limits with documented measurement uncertainty. FE-100's laboratory check-test discussion does not establish a universal field acceptance tolerance. Plot the verified point with its uncertainty before deciding what remains unexplained. S3, S6.

System-effect curve and installed operating point

Add the geometry allowance to clean-system resistance at the same airflow, then find where the combined requirement meets the fan curve. An equivalent allowance or depressed fan characteristic can represent the deficit; use one method to avoid double-counting. S1, S4, S6.

Method: Anvilfield estimates use chosen inputs of 3.00 in. wg shutoff pressure, 1.70 in. wg at 15,000 CFM and a 0.50-in. wg geometry allowance at design. Divide airflow by design airflow, square that ratio and scale the pressure terms: with x = Q/15,000, P_fan = 3.00 - 1.30x², P_clean = 1.70x², P_SE = 0.50x² and P_combined = 2.20x². Solving their intersection gives 13,887 CFM at 1.89 in. wg, including 1.46 for the clean system and 0.43 for the allowance. All curves share a total-pressure boundary, including terminal losses.

For an actual allowance, identify the arrangement, geometric ratios, effective length and reference velocity required by AMCA 201 or the manufacturer. Retain the selected figure or factor with its calculation. Use the actual fan curve within its stable range. An unexplained field residual is not a transferable factor; inlet distortion need not follow this quadratic approximation. S1, S4.

For speed recovery, multiply the shutoff-pressure term by the squared speed ratio while retaining the flow-dependent term: P_new = 3r² - 1.30x², where r is speed ratio. Anvilfield estimates 8.01% more RPM to reach 15,000 CFM with this geometry and 26.01% more shaft power relative to deficient operation at unchanged efficiency. At equal 15,000-CFM duty and total efficiency, speed-only operation needs 2.20 versus 1.70 in. wg after geometry correction, making its power 29.41% higher.

Measured geometry-to-duty cases

Twin City Fan's FE-100 (2018) publishes the following geometry interventions and measured results. S6.

Twin City Fan's FE-100 (2018) reports that straightening cyclone inlet swirl increased airflow from 12,800 to 14,500 CFM while shaft power increased from 39.4 to 53 BHP. The source derives horsepower from electrical readings and notes belt-loss and measurement uncertainty; verify input kW for an energy comparison. S6.

In the open-inlet case, the numeric 4,503-CFM reading followed throat-screen removal during testing. The source then reports an immeasurable effect from a new permanent screen near the bellmouth, without a separate final numerical reading. Preserve permanent protection in the repair. S6.

What should a measured case record include?

Record fan ID, model, RPM, density, pressure basis, measurement methods, geometry and commands. FE-100 omits some as-built clearances. Calculated changes in the table are Anvilfield estimates: divide each increase by its stated starting value. Include permanent guards and measurement devices in the final acceptance configuration. S6.

Which remedy restores airflow with the best power tradeoff?

Match the remedy to the mechanism and compare at equal required duty. Obtain corrected manufacturer selections for geometry that differs from the rating setup. S1, S6, S11.

Use target divided by actual flow when estimating speed recovery on an unchanged quadratic system. A 20% shortfall leaves 80% of target; recovering to target requires 25% more flow and RPM, with about 95% more shaft power at corresponding points. These are Anvilfield estimates from the FE-100/DOE fan laws. Electrical input also includes motor and drive efficiency, so current does not obey an exact cubic rule. S6, S11.

Before changing speed, verify permissible RPM, motor load, drive rating, bearings, wheel limits, sound, vibration and allowable duct pressure against the actual equipment documents. Resolve instability and obtain the required design approval before increasing speed. S6, S11, S14.

Does a better FEI eliminate installed losses?

Greenheck's FEI FAQ, displayed January 7, 2026, states that FEI represents standalone fan performance and does not directly include system effect. Compare revised selections at corrected duty, then compare actual input kW in the installation. Multiplying catalog FEI by a guessed geometry percentage does not produce a verified installed rating. S12.

What changes for plenum fans and fan arrays?

Apply the plenum clearances and takeoff allowances in the geometry matrix to the installed fan family.

With fans in parallel, add their flows at the same pressure using the installed array selection; multiplying one fan's free-delivery airflow by fan count misses the shared system operating point. ASHRAE's 2024 HVAC Systems and Equipment chapter 21 explains parallel-fan interaction and instability. Record fan count, actual RPM and isolation-damper position in each required state. S14

In the one-fan-unavailable test, check for reverse flow through the idle opening and confirm that remaining fans meet airflow without exceeding their limits. Test the failure and isolation sequence specified for the project. N+1 duty is the measured capacity of that remaining assembly.

Use the CRAC EC retrofit acceptance guide for electrical/control tests and the fan-wall guide for staging and isolation. Retain matched-duty baselines through retrofits.

How should diagnosis and verification be scoped and priced?

Anvilfield estimates 14 technician-hours and $1,400 to $2,100 in labor for an initial diagnosis and return verification, using $100 to $150 per technician-hour as estimating inputs. The base case covers an accessible unit, existing suitable ports and one controlled operating state; repair work is priced separately.

Method: multiply staffing by elapsed hours and the rate; the table uses $120 per technician-hour. Scope includes selection review, geometry survey, traverse, pressure/RPM/electrical readings, diagnosis and a written remedy.

What drives the estimate up?

Price roof access, escorts, permits, ceiling removal, new ports, controls support, off-hours work, rental equipment, engineering, crane work and sheet-metal alteration separately. Coordinate access and operating states through the existing-building commissioning program.

Contractors planning rooftop airflow surveys receive building lists selected for two or more visible rooftop HVAC units through PorchRocket's Ground Control.

Anvilfield estimates $1,986 in added annual electricity cost for a 10% speed increase from a 10-kW baseline, using 4,000 hours/year, $0.15/kWh, a fixed quadratic system and unchanged overall efficiency. Multiply baseline power by the cubed speed ratio, then by hours and tariff: the annual budget rises from $6,000 to $7,986. Substitute metered kW, schedule and tariff when pricing the actual remedy. S6, S11.

Retest after changing fan-adjacent geometry, drive ratio, controls or array layout, and when duty departs from the accepted baseline. Include affected operating states rather than assuming one full-speed test covers the system. Calibrate instruments on their documented manufacturer/project interval. S3, S6.

Verification traverse plan and field checklist

Repeat the marked traverse at original RPM/state after repair, then at final duty. This separates geometry recovery from speed changes. Retain raw and rejected readings with reasons. S3, S6, S8.

Prepare for inspection and running tests

For covered general-industry servicing, OSHA 29 CFR 1910.147 requires hazardous-energy isolation, control of stored energy and verification before work; a VFD stop command is not an isolating device. Apply the equipment energy-control procedure before reaching inside or making ports, including control of windmilling and stored drive energy. S15

Restore guards, close panels and clear personnel before the controlled restart. OSHA 29 CFR 1910.212(a)(1) requires protection from exposed rotating parts; take running readings through suitable external ports and protected measurement points. Reapply energy control before further internal adjustments. S15, S16

Field checklist

• Complete energy isolation for inspection and port work, then the controlled restart described above.
• Obtain the exact selection revision, configuration and pressure basis; define airflow, ventilation and pressure-critical acceptance requirements.
• Inspect rotation, drive, wheel, filters, coil and dampers; photograph and dimension both connections.
• Hold the agreed control state; record command, actual RPM, temperature, barometer and instrument calibration; zero the instruments.
• Traverse the complete grid, convert each point before averaging, and retain raw pressure and velocity readings with the area calculation.
• Measure matching fan pressure and actual electrical input kW; reconcile leakage and the independent airflow check.
• After repair: repeat original-RPM and final-duty tests, preserving guards and ventilation settings; verify mechanical and electrical limits.
• Where specified, record all-fan and one-fan-down states; retain the remedy drawing, readings and signed deficiency disposition.

Record the result as accept, correct/retest or measurement unresolved. The design or contract authority resolves duty deviations within applicable ventilation requirements; the manufacturer supplies permissible RPM and loading limits; the TAB team documents method and findings. When uncertainty overlaps the acceptance limit, apply the agreed decision rule instead of adding instrument accuracy to the allowed shortfall.

Keep dimensioned photos, selection revision, signed pressures, density, RPM, kW, operating settings and the acceptance decision with the TAB report.

Sources checked and article reviewed November 5, 2026. Update log: verified source editions and calculations; clarified measurement, power comparisons, array testing and energy isolation. Byline: Anvilfield.

How to cite this page: Anvilfield, Why is a fan delivering less air? AMCA system-effect diagnosis, 2026, https://anvilfield.com/field-guides/hvac/fan-system-effect-airflow-diagnosis/

Related tools

FAQ

People also ask

• What is a fan array in an HVAC air handler?
• What is air balancing?
• How do I measure total external static pressure?
• What are the three fan laws?
• How do you tension a fan belt?

Sources checked

• S1. AMCA Publication 201-23, Fans and Systems, official publisher page. Rating setup, fan/system interaction and selection allowances.
• S2. ANSI/AMCA 210-25 / ANSI/ASHRAE 51-25, Laboratory Methods of Testing Fans for Certified Aerodynamic Performance Rating, full publisher PDF. Laboratory rating, installation categories and pressure definitions.
• S3. AMCA Publication 203-90 (R2011), Field Performance Measurement of Fan Systems, official scope page. Field-performance measurement scope for ducted and nonducted fans.
• S4. ASHRAE Handbook, 2025 Fundamentals, chapter 21, Duct Design, public publisher text. Inlet distortion, outlet recovery and pressure accounting.
• S5. AMCA inMotion, Mitigating System Effect to Optimize Fan Performance and Efficiency, 2020 magazine article, posted 2024-06-04. 2020 magazine article, posted June 4, 2024; outlet recovery guidance.
• S6. Twin City Fan, Fan Engineering FE-100, Fan Performance Troubleshooting Guide, ©2018, manufacturer PDF. 2018 manufacturer guide; measured cases 1, 2, 4 and 5.
• S7. Twin City Fan, Fan Engineering FE-3500, Plenum Fan Application Guidelines, ©2018, manufacturer PDF. 2018 manufacturer guide; plenum clearances, selection and takeoffs.
• S8. TSI Application Note TSI-106 Rev. C, A4, April 7, 2021, Traversing a Duct to Determine Average Air Velocity or Volume. Rev. C, April 7, 2021; traverse layout and velocity averaging.
• S9. Greenheck Product Application Guide, Fan Application FA/101-99, Understanding Fan System Effects, ©2017, manufacturer PDF. FA/101-99, ©2017; connection geometry and outlet recovery.
• S10. Fluke, Duct traversal airflow measurement, undated live manufacturer article. Undated live manufacturer guidance; traverse-plane clearances.
• S11. U.S. Department of Energy / AMCA, Improving Fan System Performance: A Sourcebook for Industry, April 2003. April 2003; fan laws, system improvements and duty-based energy comparisons.
• S12. Greenheck, Is FEI affected by system effects?, live manufacturer FAQ, displayed January 7, 2026. Live manufacturer FAQ displayed January 7, 2026; FEI scope.
• S13. AMCA, Minimizing & Troubleshooting Fan System Effects, 2023 educational presentation. 2023 educational presentation; installed versus rated configurations.
• S14. ASHRAE Handbook, 2024 HVAC Systems and Equipment, chapter 21, Fans, public publisher text. Fan curves, stability, density and parallel operation.
• S15. OSHA, 29 CFR 1910.147, The control of hazardous energy, current regulation checked November 5, 2026. General-industry servicing: energy isolation, stored energy, verification and controlled restart; paragraphs (b), (d), (e) and (f)(1).
• S16. OSHA, 29 CFR 1910.212, General requirements for all machines, current regulation checked November 5, 2026. Guarding against rotating parts under paragraph (a)(1); supports retaining protection during fan operation.