Key Takeaways
- Air duct design is the discipline of sizing and routing ductwork so that airflow, pressure loss, and noise stay within the project targets. A system failing any one of these constraints fails in service.
- Velocity is the shared lever across all three constraints. Doubling duct diameter cuts friction loss to about 1/32 at the same airflow, and noise rises with speed — review size and velocity first.
- Branches share one node pressure, so flow splits by resistance. Aim every parallel path at a similar friction rate — about 0.1 in. W.C. per 100 ft as the common starting point — and back dampers with field measurement.
- Round is the efficient shape; rectangular is the space answer. Convert at equal friction with the equivalent-diameter formula, keep aspect ratios under 4:1, and remember that rectangular runs lose noise margin.
- Design values are not measured values. Balance the installed system with field airflow and static-pressure readings at commissioning.
Most failed duct systems are not failed calculations; they are failed reviews. A review that checks only duct sizes will approve a system whose branches starve, whose pressure loss exceeds the fan’s budget, or whose noise violates the project target. Air duct design rests on three physical constraints — airflow balance, pressure drop, and noise — and every sizing, routing, and shape decision moves all three at once. Two assumptions make the problem worse: that a larger fan covers design mistakes, and that rectangular duct is simply a flattened round duct. Both ignore why a duct system behaves the way it does. This guide turns the three constraints plus the shape question into a four-check review, runs one 2,000 CFM fume-exhaust proposal through it, and closes with two reusable checklists.
What Air Duct Design Must Satisfy: Three Constraints, One Geometry Decision
Air duct design is a review discipline before it is a calculation task: for any proposal, ask three constraint questions plus one geometry question. First, balance — will every branch receive its design airflow, or will low-resistance paths overdraw while high-resistance paths starve? Second, pressure — can the fan cover the total pressure loss at the design flow, counting straight-duct friction, fittings, and equipment? Third, noise — do duct velocities stay within limits tied to the project’s acoustic criterion? The geometry question then decides round or rectangular, converted at equal friction with the equivalent-diameter formula. Each mechanism below comes with the numbers you need to apply it in a review.
| Review check | The question it answers | Where it is explained |
|---|---|---|
| Airflow balance | Will every branch receive its design airflow? | Balance section |
| Pressure drop | Can the fan pay the total pressure loss? | Pressure section |
| Noise | Do duct velocities stay inside the acoustic limits? | Noise section |
| Shape decision | Round or rectangular, and does the conversion stay honest? | Shape and conversion section |
Use these four checks as the agenda for every duct proposal you review; you can apply them before any calculation begins, and each later section gives you the numbers to make them precise.
Pressure Drop: Why Duct Resistance Is the Bill the Fan Pays
Pressure drop is the price a duct system charges the fan: every straight run, every fitting, and every component consumes part of the fan’s available pressure, and the fan must pay the total before design airflow exists. If a proposal cannot show a pressure budget, that is the first review finding.
Static, velocity, and total pressure: what the duct numbers mean
Static pressure is the pressure the duct walls feel; velocity pressure is the pressure equivalent of the air’s motion; total pressure is their sum (pt = ps + pv), per ASHRAE’s duct design chapter. Velocity pressure is calculated as pv = (V/4005)², with V in fpm and pv in inches of water column, for standard air at 0.075 lbm/ft³; ASHRAE derives the 4005 constant from that standard density. At 2,000 fpm the result is about 0.25 in. W.C. (example value — use project inputs).
Two review consequences follow. Because velocity pressure is part of total pressure, a static reading only means something together with the velocity at that point, so readings taken at different duct sizes along one run are not directly comparable. And because the formula assumes standard air, hot exhaust or high-altitude installations need a density correction before their numbers are trusted; the closing section flags that correction again.
Friction plus fittings: the two things that eat pressure
Straight duct consumes pressure through friction between the air and the duct wall; fittings — elbows, tees, transitions, dampers — consume more when air changes direction, splits, or speeds up, expressed as loss coefficients against the local velocity pressure. ASHRAE’s chapter sums friction, fitting, and equipment losses section by section to arrive at the total pressure the fan must develop.
Components such as filters, scrubbers, and hoods add their own pressure drop, taken from the manufacturer’s data sheet rather than guessed (see FAQ). The mechanism that matters for review is simple: total pressure loss is the fan’s bill. Ideal fan power tracks flow times pressure (Pi = q·Δp in fan engineering references), the system runs where its resistance curve crosses the fan curve, and throttling with dampers adds loss while lowering total efficiency.
Diameter is the most sensitive lever: the factor-32 rule
At the same airflow, doubling duct diameter cuts friction loss by about a factor of 32 — a widely quoted engineering rule from velocity guidance. The loss belongs mostly to velocity: the same air moves a quarter as fast through a doubled area, and friction climbs steeply from there; shrinking diameter pushes loss up just as steeply. Because fitting losses scale with velocity pressure, total pressure loss trends with the square of velocity — an engineering direction for review use, not a precise law.
Velocity extremes carry their own penalties. Industrial-ventilation health guidance notes that air moving too slowly lets contaminants settle, while excessive speed wastes fan power, can create noise problems, and increases abrasion; typical industrial main-duct velocities run about 8 to 12 m/s (roughly 1,600–2,400 fpm), and guidance tables cap industrial mains near 3,000 fpm supply and 1,800 fpm return, with branches lower at about 2,200 fpm supply and 1,500 fpm return. Treat both bands as review references — the right transport velocity follows the contaminant and the capture conditions of the specific process.
You can now flag the two most common pressure findings in a proposal: diameters that are small for the stated flow, and missing velocity or pressure budgets. Both are grounds to send the design back before anything is ordered.
Airflow Balance: Why Branches Starve and What Restores Design Flow
Balance is the constraint nobody sees at purchase: a duct system delivers its design airflow only when every parallel path can pass its share against the same pressure difference.
Why flow splits the way it does: branches share one node pressure
Parallel branches hang between the same two pressure levels — the same node pressure at their takeoffs and the same pressure at their common outlet — so each branch’s airflow settles where its own resistance consumes exactly that shared difference. A low-resistance path passes more than its design share; a high-resistance path passes less, and that starvation is the effect you notice: weak hoods and quiet-but-useless branches. ASHRAE’s duct design chapter requires the pressure-balancing equations to be satisfied “to attain pressure balancing for design airflow” — those equations are the formal version of the same settlement.
The balance is fragile by design: adding a branch, changing the fan, or rerouting a main changes every path’s resistance and re-splits the flow. Occupational health guidance (CCOHS) warns that a system modified without rebalancing will “self-balance” — airflow is typically reduced in the sections with higher resistance — which is how a new scrubber or an added hood quietly starves an older branch.
Designing for balance: aim every path at a similar friction rate
Designers aim every parallel path at a similar friction rate so branches start close to balanced. Equal-friction design commonly uses about 0.1 in. W.C. per 100 ft for supply ducts and 0.08 in. W.C. per 100 ft for return ducts (roughly 0.85 and 0.65 Pa/m) as starting points. For a single-direction exhaust system, that supply-versus-return pairing collapses: pick one common rate — about 0.1 in. W.C. per 100 ft is the usual starting point — and apply it to every parallel path. When every path is drawn at the same rate, simple systems come close to self-balancing.
Systems with many takeoffs and changing diameters rarely stay self-balanced, which is why balancing dampers exist: a damper adds controlled resistance to a branch that overdraws and throttles it back to its design flow. ASHRAE is explicit that relying entirely on dampers “is not economical and may create objectionable flow-generated noise” — so dampers are the correction, not the design. Finish the correction at field balancing: measure airflow per branch and set each damper to its design value, because the installed geometry never matches the drawing exactly.
Backdraft dampers protect the exhaust direction
Exhaust systems add a direction requirement: the fan pulls fumes from the hood toward the scrubber and stack, and capture fails the moment that direction reverses. When the fan cycles off — or wind, another fan, or an imbalance pushes pressure the other way — a plain duct offers no resistance to reverse flow, and contaminated air can drift back into the room.
A backdraft damper holds the direction: it opens with forward flow and closes against reverse flow, protecting containment whenever equipment stops. Our PP duct range includes backdraft dampers for exactly this job. Specifying damper type, material, and leakage class is a separate selection task in the blog’s airflow-accessory guide; the explanation here stops at why the damper belongs in the system.
You can now require three items in any exhaust proposal: branch balancing dampers, a backdraft damper on the discharge path, and a written field-balancing report — and you can reject the argument that a larger fan covers an unbalanced layout.
Noise: The Third Constraint — Where It Comes From and How Velocity Controls It
Noise is the constraint that arrives at the end of a project and stays for its whole life: duct noise is not tuned out at commissioning, it is set earlier by the velocities and fittings the design chose. If a proposal has no acoustic criterion, noise is already a risk.
Where duct noise comes from: regenerated sound grows with velocity
Flow-generated noise is created inside the duct itself: the faster air moves, the stronger the turbulence, and the louder the regenerated sound that travels with the airstream. ASHRAE’s noise and vibration guidance is direct on the lever — reducing duct airflow velocity significantly reduces flow-generated noise — and its tables note that elbows and other fittings can increase airflow noise substantially, so velocities should be reduced accordingly where fittings are dense.
Dampers throttled hard belong on the review list as a second design-side source: they generate their own flow noise, which is why the balance section treated them as a correction rather than a design tool. Fan sound and duct breakout also travel through the system, but their numbers belong to fan manufacturers and acoustic specialists; noise that shows up as an operating fault belongs to the maintenance guide instead.
Velocity ceilings tied to acoustic criteria: how to read the guidance
Acoustic design criteria such as NC — and the newer RC(N) family — assign a target curve to a space; factory-type spaces typically sit around 40–65 NC on room-criterion references, while offices and labs run lower. ASHRAE’s handbook guidance organizes maximum recommended duct velocities by criterion, duct location, and shape; the table below reproduces the occupied-space rows most relevant to industrial and laboratory review (representative values — confirm project values against the current edition).
| Acoustic design criterion (NC / RC(N)) | Max. velocity, rectangular duct (fpm) | Max. velocity, circular duct (fpm) |
|---|---|---|
| 45 | 2,000 | 3,900 |
| 35 | 1,450 | 2,600 |
| 25 | 950 | 1,700 |
Three notes change how you apply the table. Branch ducts should run at about 80% of the listed values, and final runouts to outlets at 50% or less; duct location matters — the same 35 criterion above a suspended acoustic ceiling allows about 3,000 fpm circular versus 2,600 fpm inside occupied space; and fittings add noise, so dense-fitting sections should run slower than the plain-run number.
Design-side noise control before you buy silencers
The design-side sequence is short: fix the acoustic criterion first, choose velocity bands that respect it (cross-checked against the industrial velocity bands above), keep fittings smooth and dampers few, and only then consider attenuation hardware. Silencers and lined sections exist and work, but their insertion-loss data belongs to the manufacturer and the acoustic engineer — no useful number can be quoted here without a specific product and frequency spectrum.
The procurement consequence is concrete: write the acoustic criterion and the velocity limits into the design brief and the RFQ. A vendor who receives a target before quoting can size the ductwork accordingly; a vendor who learns about noise at startup can only sell you silencers.
You can now do a per-section noise check on any proposal: compare each duct velocity against the criterion row for its location and shape, apply the branch and runout discounts, and send back any section that exceeds the limit with the specific number it must meet.
Round or Rectangular: Efficiency, Space, and the Equivalent-Diameter Swap
Round and rectangular duct move the same air, but not at the same cost in friction, material, or noise margin. The choice is a trade between hydraulic efficiency and space, and the equivalent-diameter formula is the tool that makes the trade honest.
Why round is the hydraulically efficient shape
Round duct is efficient for a geometric reason: of all shapes with the same cross-section area, the circle has the smallest perimeter, so it presents the least friction surface and uses the least material for the area it delivers. As engineering references put it, a circular duct is always more efficient than a rectangular duct of the same cross-section area — the comparison that matters here.
A quick example makes the penalty visible: an 18 in × 9 in rectangle has a perimeter of 54 in, while a round duct of roughly the same area — about 14 in diameter — has a circumference near 44 in (example values — use project inputs). The extra 10 in of perimeter is extra friction surface, extra material, and extra sealing edge. The cost of the efficient shape is spatial: round duct needs circular clearance and does not hug walls or ceiling planes.
When rectangular wins anyway: space, fit, and the decision matrix
Rectangular duct earns its place where space is the binding constraint: limited ceiling height, wall-hugging runs, and tight service zones where a round duct would waste the corners. Many installations need a mix — round mains where height allows, rectangular transitions where it does not. The decision matrix below turns that trade into per-run questions.
| Decision question | Choose round when… | Choose rectangular when… |
|---|---|---|
| Hydraulic efficiency | Height and space allow it — lowest friction and material per area | Only when space forces the shape |
| Space and fit | Clear space is available; circular clearance is acceptable | Clear height is tight; runs hug walls or ceiling planes |
| Noise margin at the same acoustic criterion | You want the higher velocity ceiling of circular duct | You accept the narrower rectangular ceiling and slower runs |
| Material and sealing surface | You want the shorter perimeter and fewer seams | The run is short or the layout leaves no round path |
| Fair comparison | — | You convert with the equivalent-diameter formula before accepting the swap |
Read the matrix per run, not per system: a long straight main in open space stays round, and the same layout’s final low-height section can still go rectangular. The rows also stack — a rectangle chosen for space pays again on noise margin and surface area, which is why the conversion step below exists.
The equivalent-diameter formula and a one-step conversion
The equivalent-diameter formula converts a rectangular duct into the round size an air duct design review compares against it: De = 1.30 × (a·b)^0.625 / (a+b)^0.25, where a and b are the rectangle sides in the same unit — the Huebscher formula carried in standard engineering references. A rectangular duct of sides 300 mm and 500 mm converts to an equivalent round duct of about 420 mm, per the published example.
The formula keeps friction roughly equal, which is its purpose: a run designed as a 14 in round duct, forced into an 18 in × 9 in rectangle, converts to De ≈ 1.30 × (18×9)^0.625 / (18+9)^0.25 ≈ 13.7 in (example value — use project inputs). Because 13.7 in is smaller than the original 14 in, the rectangular version costs slightly more friction — the swap does not preserve the original margin unless the rectangle is sized up. Note what the conversion does not equalize: area, velocity, and noise behavior stay different, and the equivalent diameter is a review approximation, not a replacement for professional sizing.
Aspect ratio: keep it under 4:1 when space allows
Aspect ratio is the rectangle’s self-inflicted penalty: the longer and thinner a rectangle gets, the more perimeter it carries for the same area, and friction, material, and sealing cost all climb with it. Common design guidance keeps rectangular aspect ratios at or below 4:1 where space allows, and treats anything beyond that as a cost trade to justify explicitly.
The extreme case shows why: a 27 in × 6 in rectangle (aspect ratio 4.5:1) has the same area as the 18 in × 9 in example but converts to De ≈ 13.0 in — a thinner, longer box that behaves like a smaller round duct (example value — use project inputs). And because rectangular runs carry the lower velocity ceilings from the noise section, every shape swap also narrows noise margin, which is exactly the interaction the worked example in the next section demonstrates.
You can now decide per run: choose round where height allows it, choose rectangular where the space demands it, and before accepting any rectangle, convert it with the equivalent-diameter formula, check the aspect ratio against 4:1, and re-check the velocity against the acoustic table.
Worked Example: Reviewing a 2,000 CFM Fume-Exhaust Proposal
Proposal under review: a 2,000 CFM corrosive-fume exhaust system in PP duct — one main with two hood branches — running through occupied plant space. The designer proposes a nominal 14 in round main. The review below runs the three constraints plus the shape question; all values are worked example values — use project inputs.
Step 1 — Continuity: does the duct size match the airflow?
Continuity is the first check because everything else depends on it: a 14 in round duct has a cross-section of about 1.07 ft², so the main runs at V = Q/A = 2,000 ÷ 1.07 ≈ 1,870 fpm (about 9.5 m/s), and its velocity pressure at that speed is roughly 0.22 in. W.C. at standard air.
Judgment: 1,870 fpm sits inside the typical industrial main band of 8–12 m/s (about 1,600–2,400 fpm) and under the supply-side guidance maximum near 3,000 fpm. The same guidance tables put the return- and exhaust-side main ceiling near 1,800 fpm, so this exhaust main runs just above that row; for clean gas and vapor fume service the margin is small but workable, and it disappears if the stream carries dust or mist — in that case, transport velocities come from industrial-ventilation references for the specific contaminant instead (see the FAQ).
Step 2 — Balance: what the spec must include
Two branches with different lengths and fitting counts will not split the 2,000 CFM to design unless resistance is managed. The specification must therefore state equal-friction intent — every parallel path sized at one common friction rate, about 0.1 in. W.C. per 100 ft as the typical starting point — include balancing dampers on the branches, and add a backdraft damper so the exhaust direction holds whenever the fan cycles off.
The acceptance clause matters as much as the hardware: require a field-balancing report with measured airflow and static pressure per branch. Design values are not measured values, and this review approves a specification — the specification must end in measurement, not in the drawing.
Step 3 — Noise: check velocity against the acoustic criterion
Occupied plant space typically carries an acoustic criterion in the 40–65 NC range, so test the main at criterion 45 first: circular duct at 3,900 fpm ceiling gives 1,870 ÷ 3,900 ≈ 48% — a pass with real margin. Against criterion 35 (ceiling 2,600 fpm) the same run sits at about 72%; against criterion 25 (ceiling 1,700 fpm) it reaches about 110% and fails.
Two corrections fix a tight target before any hardware is bought: enlarge the main to a nominal 15 in (velocity ≈ 1,630 fpm, now below the 1,700 fpm ceiling), or reroute the run into a shaft or above a drywall ceiling, where the 25-criterion circular ceiling is about 2,500 fpm. Beyond those moves, attenuation belongs to manufacturer sound data and an acoustics engineer.
Step 4 — The round-to-rectangular swap and the verdict
The plant forces a rectangular transition in one section, so the shape question is real. An 18 in × 9 in rectangle — aspect ratio 2:1, inside the 4:1 guide — matches the area of a 14.4 in round duct, but converted at equal friction its equivalent diameter is only De ≈ 1.30 × (18×9)^0.625 ÷ (18+9)^0.25 ≈ 13.7 in, smaller than the 14 in round it replaces, so friction rises slightly unless the section is upsized.
The noise margin narrows at the same time: the rectangular section moves 2,000 CFM at about 1,778 fpm, which is 89% of the 2,000 fpm rectangular ceiling at criterion 45 — a pass with thin margin compared with the round run’s 48% — and about 123% of the 1,450 fpm rectangular ceiling at criterion 35, a fail. Verdict: the round design passes criteria 45 and 35 with margin, fails criterion 25 in occupied space, and a rectangular substitution passes only criterion 45 on thin margin; set the criterion before fabrication, and if criterion 25 is real, order the 15 in main or relocate the run. This review deliberately outputs no friction or pressure-drop figure — that calculation belongs to the duct sizing workflow, because a review’s job is to catch proposals that cannot work, not to replace the designer’s numbers.
You can now run the same four-step review on any proposal — continuity, balance specification, acoustic check, and shape conversion — and each step returns a pass, a fail, or a specific correction path you can hand back to the designer.
Quick answers to residual review questions
Where does the pressure drop of a filter, scrubber, or the hood itself enter the review?
Equipment adds its own pressure drop to the duct friction-and-fittings total. Take the value from the equipment manufacturer’s data sheet and add it to the total before judging whether the fan has margin.
Do I need different duct velocities when the exhaust carries dust or mist?
The velocity guidance above covers air and gas flows. Particulate-laden streams should set transport velocities from industrial-ventilation references for the specific contaminant and capture conditions, because the settling and abrasion behavior changes the design band.
When should a system be re-balanced?
At commissioning, after any layout, fan, or process change, and on a scheduled basis. Measure airflow and static pressure each time — a changed system re-splits flow by resistance until someone rebalances it, and that re-split is rarely where the design intended.
Is the equivalent-diameter method valid for any rectangular shape?
The Huebscher formula is a standard approximation for normal duct proportions. Extreme aspect ratios lose accuracy, which is one reason the under-4:1 guide exists: keep the rectangle reasonable and the conversion stays a fair review tool.
Anything deeper on sizing numbers or damper classes belongs to the sizing and accessory guides referenced above.
What a three-principle review cannot replace
This closing part converts the guide into two reusable documents: a review checklist and an information list. Copy them into your next proposal review and your next request for quotation.
The three-principle review checklist you can copy
Each question maps to a constraint covered in the sections above. A “no” on any row is a written finding, not a preference.
| Check | Ask | A pass means… |
|---|---|---|
| Balance 1 | Is every branch assigned a design airflow? | Each branch has its own CFM figure and the branches sum to the fan flow. |
| Balance 2 | Are parallel paths sized at similar friction rates? | Paths follow equal-friction intent near common starting points (about 0.1 in. W.C./100 ft supply, 0.08 return). |
| Balance 3 | Are dampers, direction, and measurement specified? | Balancing dampers exist, a backdraft damper protects the exhaust direction, and a field-balancing report is a contract item. |
| Pressure 1 | Does a total pressure budget exist? | The fan’s available pressure covers duct, fitting, and equipment losses with margin. |
| Pressure 2 | Are diameters and velocities defensible? | Section velocities sit inside the typical industrial bands and under the maximum guidance values. |
| Pressure 3 | Are shapes and fittings under control? | No extreme aspect ratios; dense-fitting sections are allowed slower velocities, not faster ones. |
| Noise 1 | Is the acoustic criterion fixed? | The project names a criterion (for example NC/RC(N) 45 for factory-type spaces) before fabrication. |
| Noise 2 | Is each section under its velocity ceiling? | Velocities respect the table row for location and shape, including the 80% branch and 50% runout discounts. |
| Noise 3 | Is attenuation a decision, not a default? | Any silencer or lining carries manufacturer sound data and an acoustic review, not an estimate. |
Keep the checklist with the drawings, not in a drawer: run it once at proposal review, again when the layout changes, and once more before the purchase order. Each “no” is a line item for the designer, and the input list below is what you send alongside.
Information to send for review or RFQ
A reviewer or a vendor can only answer the checklist when the inputs are complete. Send these items with every request.
| Input | Why it is needed |
|---|---|
| Total airflow and per-branch design allocation | Feeds the continuity and balance checks |
| Run lengths and a fitting schedule per path | Feeds the pressure check path by path |
| Process medium, temperature, and site altitude | Triggers the density correction for non-standard air |
| Acoustic criterion for each occupied space | Sets the velocity ceiling for the noise check |
| Existing fan data or available static pressure | Tests whether the pressure budget has margin |
| Routing and clear-height constraints | Decides round versus rectangular per run |
Companion decisions sit in other guides: duct wall thickness for the pressure class, installation and support practice, operating faults such as leaks and noise, and damper selection and leakage classes each have their own article in this blog, and thermal insulation and sealing for cold and hot runs are covered in our duct insulation guide.
What a three-principle review cannot replace
A review catches proposals that cannot work; it does not produce the final design. Final sizing, fan selection, and the balance report stay with the responsible design engineer and a field commissioning team; projects with dense fittings or tight acoustic targets should carry an acoustics engineer’s review before fabrication. For hot or high-altitude streams, re-check velocities and pressure against the actual air density rather than standard-air values.
Good air duct design, in short, is a reviewable discipline: balance, pressure, noise, and shape are checkable constraints that every proposal must pass on paper before it earns your order. When your review points to a polypropylene exhaust system, our round duct series spans φ20–600 mm with rectangular sections available for space-constrained runs, and backdraft dampers protect the exhaust direction.
Send the completed checklist with your airflow, process, and space inputs when you request a specification — the full product range is the starting point, and our product specialists will work from your checklist inputs. The sizing hub and the pipe-sizing guide carry the calculation depth this guide deliberately left to them.
