TECHNICAL ARTICLE

Key Considerations in Air Duct Design

Duct area and velocity, fitting losses, critical paths, fan power, controls and practical acceptance checks.

1. Objectives of good duct design

Ductwork must deliver required terminal flows while controlling fan energy and noise, fitting within the building, supporting fire/smoke safety, allowing cleaning and remaining buildable. Evaluate space, sound, leakage, balancing, safety, capital and operating costs together.

2. Inputs before design

Obtain current room/terminal flows, supply/return/outdoor/exhaust schematics, ceiling and shaft dimensions, acoustic criteria and terminal data, pressure/leakage classes and testing requirements, fire zones/dampers/access, thermal and condensation insulation needs, and available external static pressure.

Update sizing after airflow changes. Reusing dimensions based on old flows with newly selected equipment is a coordination error.

3. Flow, area and velocity

Q = V × A

Q is m³/s, V is m/s and A is actual free area in m². At 3,600 m³/h=1.0 m³/s and an initial 5 m/s, area is 0.20 m²; 500×400 mm provides that area. Still check friction, fittings, sound and fabrication.

Velocity pressure varies with velocity squared. Doubling velocity approximately quadruples it; smaller ducts can increase fan duty, electricity, fitting losses and noise disproportionately.

4. Sizing methods

Equal friction sizes main and branch ducts near a target loss per metre. It is a useful starting point for low/medium-pressure comfort systems, but does not automatically balance terminal pressure.

Static regain uses reduced velocity after flow take-offs to recover static pressure against downstream losses. It can improve distribution in large systems and requires iterative calculation.

Velocity limits may govern acoustics, erosion, particle transport or restricted shafts. Different network sections can use different constraints.

5. Straight-duct friction

Δp = f × (L/Dh) × (ρV²/2)

f is friction factor, L length, Dh hydraulic diameter and ρ density. Use validated software or an accepted equivalent-diameter method for rectangles. A very flat section with the same area has more perimeter, material and insulation and potentially more friction. Include altitude and temperature effects on density.

6. Fitting losses

Δpfitting = C × ρV²/2

C depends on geometry and the reference section's velocity pressure. Use the ASHRAE database, an accepted design source or validated manufacturer data. Appropriate-radius bends, gradual transitions and well-oriented branches reduce loss. Calculated geometry must match fabricated fittings.

7. Critical path and fan pressure

The highest-resistance route is not necessarily the longest. Include filters, coils, recovery devices, silencers, terminal boxes, dampers, grilles and louvres at the defined fan boundary.

ComponentIllustrative loss
Ducts and fittings210 Pa
Dirty design filter120 Pa
Coil85 Pa
Silencer45 Pa
VAV box and terminal90 Pa
Critical total550 Pa

These values illustrate the method; obtain actual product losses. Arbitrary large margins can create oversized motors, noise and permanent damper throttling losses.

8. Sound and vibration

Velocity alone is not an acoustic criterion. Assess fan sound power, regenerated fitting/damper noise, breakout, grille losses and cross-talk between rooms. Reduce terminal approach velocity/pressure where necessary. Include silencer loss, self-generated noise, length and access.

9. Leakage and insulation

Leakage wastes delivered airflow and fan/heating/cooling energy. Negative-pressure return/extract ducts can draw contaminants in. Specify pressure class, airtightness, joint sealing and test scope.

Check insulation against surface temperature and condensation as well as energy loss. Preserve vapour-barrier continuity on cold ducts.

10. Fire, hygiene and access

Place fire/smoke dampers under adopted rules and approved scenarios. Keep actuators and service openings accessible. Kitchen extract, smoke exhaust, laboratories and corrosive streams need specific materials, cleaning, drainage, velocity and fire-safety design.

11. Drawings and BIM coordination

Nominal duct dimensions are not the installed envelope. Include insulation, flanges, supports, tolerances and servicing space; model actual bend/transition lengths.

Verify structural openings early, provide damper/filter/cleaning access, coordinate sprinklers/lights/cable trays and show flexible connections and vibration isolation at equipment.

12. Testing, adjusting and balancing

TAB verifies site flows. Provide accessible measurement points with suitable straight lengths. Record fan rotation, filter state, damper positions, equipment pressures, terminal flows and controls. Test minimum/maximum VAV flows and static-pressure reset together.

13. Design checklist

Check current terminal demand; critical paths in each mode; all straight/fitting losses; matching drawing geometry; dirty-filter loss; efficient fan operating point; noise/cross-talk; pressure and leakage classes; insulation/vapour barrier; damper access; and accessible measurement/balancing points.

14. Common mistakes

Using m³/h as m³/s; choosing velocity without total losses; crude fitting estimates; calculating long-radius bends but installing sharp bends; missing filter/internal losses; throttling excessive fan pressure; omitting flange/insulation space; concealing fire-damper access; crushed or long flexible ducts; and no TAB/leak-test access.

15. Compare sections and fittings at the same flow

At 1 m³/s, 500×400 mm gives 0.20 m² and 5 m/s. A 600×400 mm section gives 0.24 m² and 4.17 m/s. At assumed density 1.2 kg/m³, velocity pressures are 15 and 10.4 Pa. With illustrative K=1.5, fitting losses are 22.5 and 15.6 Pa. These are comparisons, not final size recommendations.

Use coefficients for actual geometry and reference sections; tee main and branch coefficients need not match. A 1,000×200 section and 500×400 section have equal area and velocity but different perimeter, friction, fabrication and acoustics. Equivalent diameter may serve the friction method; calculate actual velocity from actual area. Assess reinforcement and insulation costs for flat sections.

16. Critical route and boundary: worked example

Assume 35 m×0.8 Pa/m=28 Pa straight friction, 90 Pa fittings, 45 Pa terminal and 20 Pa balancing loss: external distribution loss is 183 Pa.

ComponentLossBasis
Straight duct28 PaLength and friction
Fittings90 PaActual geometry coefficients
Terminal45 PaProduct selection
Balancing element20 PaDesign setting
External path183 PaDefined boundary

State whether AHU internal filter/coil losses are already included in fan selection. Counting them twice and omitting them entirely are both errors. Record clean/dirty conditions separately.

If another branch needs 220 Pa, the first route does not govern. Check all candidates and the correct return/exhaust boundary. Do not combine separate fans' paths incorrectly or inflate every component with unjustified factors.

17. Fan power and variable-flow control

In a separate example, 1 m³/s at 500 Pa total pressure requires 500 W air power. At 60% total efficiency, electricity is about 833 W. This 500 Pa is not the earlier 183 Pa external distribution figure. Use consistent pressure and efficiency definitions, adding motor/drive losses only if excluded.

For similar fan conditions, 80% speed suggests approximately 80% flow, 64% pressure and 51.2% power. These affinity relations are not guaranteed installed consumption: dampers, minimum outdoor air and control matter. Verify critical terminal requirements before reducing speed. Sensor location and setpoint form part of design; a fan-discharge sensor alone does not prove distant-terminal adequacy.

Allow installation and service space beyond the duct body. Avoid measurements immediately beside bends. Record terminal flows, sound and modes at acceptance. Investigate leakage, filters, returns and dampers before increasing fan size. Reference: https://handbook.ashrae.org/Handbooks/F25/IP/F25_Ch21/F25_Ch21_ip.aspx

18. Sources and scope

Verify adopted editions, local rules and project requirements. This educational/preliminary guide does not replace final coordinated engineering calculations and product data.

← All articlesTechnical feedback →