- Duct design is the point where a careful load calculation is either delivered or thrown away. The load says how much air a space needs. The duct decides whether it arrives, at what noise level and at what permanent fan energy cost.
- There are three recognised sizing methods. Equal friction suits most commercial supply and return, static regain suits long high velocity trunks, and constant velocity is for exhaust systems that must maintain transport velocity.
- Velocity limits in occupied buildings are set by acoustics, not by airflow physics. Air will move at almost any speed you ask. The room noise criterion is what fails first.
- Fitting losses and system effect routinely outweigh straight duct friction. A badly detailed elbow or a fan boxed into a plenum can cost more static pressure than a hundred feet of straight duct.
- SMACNA pressure classes run from plus or minus one half to 10 inches of water gauge, and each maps to an expected leakage class using the relationship F equals CL multiplied by P to the power 0.65.
- The five most expensive duct design failures are all coordination failures: an oversized fan, an undersized shaft, a flattened duct, no straight length at the fan, and a leakage test failed after the ceilings are closed.
- What Is HVAC Duct Design?
- How Duct Sizing Actually Works
- The Three Sizing Methods
- Which Projects and Systems It Governs
- What Duct Design Asks of Every Other Trade
- Codes, Standards and Pressure Classes
- Healthcare, Laboratories, Data Centres and Aviation
- Choosing a Sizing Method
- Static Pressure, System Effect and Where It Fails
- Where Budlong Works
- Related Reading
- Frequently Asked Questions
A load calculation is a piece of arithmetic that says how much air each space needs. It can be done beautifully, with real occupancy schedules, real glazing data and a properly modelled envelope, and still produce a building that is noisy, unbalanced and expensive to run. The reason is almost always the same. The distribution network that was supposed to deliver that air was sized late, flattened in coordination, detailed by someone else, and connected to a fan that was picked from a static pressure figure nobody could defend.
Duct design is where the engineering either lands or evaporates. It is also unusually unforgiving, because almost every mistake in it becomes permanent the moment the ceilings close. This article sets out how duct sizing works, how total static pressure is actually calculated, which losses matter and which are noise, and the specific failures that cost the most money on California projects. Budlong delivers HVAC design services across healthcare, aviation, education, laboratory and mission critical work, and the failure catalogue at the end of this article is drawn from those projects rather than from a textbook. You can see the wider range of that work on our expertise overview.
1. What Is HVAC Duct Design?
HVAC duct design is the process of turning a schedule of airflow requirements into a physical network, and then calculating what that network costs the fan. It has three outputs, and all three matter.
The first is geometry: the size, shape, route and material of every duct section, along with every fitting, transition, damper, terminal and diffuser. The second is a pressure calculation: the total static pressure that the worst case path through that network imposes on the fan, which becomes the basis for fan selection and therefore for the largest single mechanical energy consumer in most commercial buildings. The third is a construction specification: the pressure class each portion of the system is built to, the leakage class it must achieve, its sealing and insulation requirements, and its acoustic treatment.
Those three outputs are usually produced by different people at different times, which is the structural reason duct design goes wrong. The geometry is drawn by a designer, the pressure calculation is done by an engineer, the fittings are detailed by a contractor, and the route is renegotiated by whoever loses the coordination meeting. Each step is defensible on its own. The accumulated result is a system that behaves nothing like the one that was calculated.
The single most useful discipline in duct design is to identify the critical path early, mark it on the drawing, and protect it through coordination. Everything else in the system can be rerouted, flattened and compromised without changing the fan. The critical path cannot.
2. How Duct Sizing Actually Works
Duct sizing rests on a single relationship. For a given airflow, the pressure lost to friction depends on the duct’s cross sectional area, its shape and its internal roughness. Make the duct bigger and friction falls, velocity falls, and noise falls with it. Make it smaller and all three rise. Everything else is a way of choosing where on that curve to sit.
The friction chart
The classic tool is the friction chart, which plots airflow, duct diameter, velocity and friction rate on a single set of axes so that fixing any two fixes the others. It is published in the ASHRAE Handbook of Fundamentals and reproduced in every duct calculator and sizing tool. The chart is drawn for round galvanised duct at standard air density, so two corrections are always needed in practice: an equivalent diameter conversion for rectangular or flat oval duct, and a density correction for high altitude or high temperature work. Both are routinely skipped, and both matter.
Choosing a friction rate
No code specifies a friction rate. Conventional commercial practice for low pressure supply, return and exhaust sits in a band of roughly 0.05 to 0.10 inches of water gauge per 100 feet, and the choice within that band is economic rather than technical. A lower friction rate means larger ducts, more sheet metal, deeper ceiling voids and higher capital cost, in exchange for permanently lower fan energy. A higher rate does the reverse. Because fan power varies with the cube of speed, small differences compound: a 10 percent increase in fan speed produces roughly a 33 percent increase in energy consumption, which is why the friction rate decision has a much longer tail than its capital cost implies.
The friction rate is a thirty year decision made in an afternoon. The extra ceiling depth needed to drop from a high friction rate to a moderate one is negotiated once, in a coordination meeting, against an architect defending floor to ceiling height. The fan energy difference is paid every hour the building operates. Framing it as a life cycle cost question rather than a space question is the only way that argument is ever won.
Velocity limits are an acoustic decision
Air will travel through a duct at almost any velocity you ask of it. What breaks first is the acoustic environment. High velocity generates broadband noise within the duct, regenerates additional noise every time the airstream passes a damper blade, a takeoff or an abrupt fitting, and breaks out through the duct wall into whatever space the duct passes over. Duct velocity limits are therefore derived backwards from the room noise criterion rather than forwards from airflow physics.
The practical consequence is that acceptable velocity falls sharply as the duct approaches the occupied space. A shaft riser or a mechanical room trunk can run fast because nobody occupies the space it passes through. The same air in a runout above a private office or a classroom must run far slower, because there is nothing between the duct wall and the ceiling tile. Systems that are sized to a single velocity from the fan to the diffuser are almost always too fast at the end and too slow at the beginning.
Aspect ratio and the permanent penalty
Rectangular ducts of equal cross sectional area do not perform equally. A duct at a high aspect ratio, meaning a much wider dimension than depth, presents a longer perimeter to the airstream. More wetted surface means more friction for the same airflow, and it means more sheet metal, heavier gauge and more reinforcement because SMACNA construction requirements escalate with the larger dimension rather than with the area. It also means more surface area for heat gain and loss and more panel to radiate breakout noise.
Aspect ratios beyond about four to one are widely discouraged for exactly these reasons, and the further past that a duct is pushed, the worse each penalty gets. The trap is that flattening a duct in a coordination meeting looks free. The plan area is unchanged, the airflow is unchanged, the drawing still works. The cost is invisible until the fan is selected, and permanent thereafter.
📖 Also Read: HVAC Load Calculation Methods for the analysis that produces the airflow figures duct design starts from.
3. The Three Sizing Methods
Three methods are in general use. They are not interchangeable, and applying the wrong one is a common cause of systems that balance badly or waste energy.
Equal friction
Equal friction holds the pressure loss per unit length constant across the system. The designer picks a friction rate, and each section is sized so that it loses that amount per 100 feet regardless of how much air it is carrying. Duct sizes therefore step down as air is dropped off along the run.
It is fast, easy to check, easy to hand to a drafting team, and defensible at plan check. Its weakness is that it takes no account of static regain and produces meaningfully different available pressures at takeoffs on runs of very different length. A short branch near the fan will be over pressured and a long branch at the end will be starved, which is what balancing dampers exist to correct. In a badly proportioned system, that damper throttling both wastes fan energy and regenerates noise.
Static regain
Static regain exploits the fact that when velocity falls, some of the velocity pressure converts back into static pressure. The method sizes each downstream section so that the regain from the velocity reduction roughly cancels the friction loss of the next section, which leaves approximately the same static pressure available at every takeoff along the trunk.
The result is close to self balancing, which is why it suits long high velocity trunk systems: large variable air volume distribution, terminal buildings, arenas and anything where a single main run serves many takeoffs over a long distance. Its costs are that the calculation is iterative rather than direct, that downstream sections come out larger and therefore need more space at the far end of the run where space is often tightest, and that it is sensitive to fitting quality because poorly detailed transitions destroy the regain the method depends on.
Constant velocity
Constant velocity fixes the velocity and lets pressure loss fall out of the calculation. It is not a comfort air method. It exists for systems where the airstream is carrying something and must keep carrying it: kitchen grease exhaust, dust and particulate collection, fume and process exhaust, and anything where dropping below a transport velocity means material settles in the duct and becomes a fire or blockage risk. In those systems the transport velocity is the governing requirement and energy optimisation is secondary.
Most real buildings use more than one method. A large VAV system might use static regain on the medium pressure trunk from the air handler and equal friction on the low pressure runouts downstream of each terminal unit, while the kitchen and laboratory exhaust are sized to constant velocity. Specifying one method for a whole building is usually a sign the system was sized by a tool rather than designed.
4. Which Projects and Systems It Governs
Duct design applies to any project that moves conditioned or exhausted air, but the depth of analysis required varies enormously with what triggers it.
New construction
New buildings get the full exercise: route planning against structure and ceiling voids, sizing, static pressure calculation, pressure and leakage class specification, acoustic analysis and fan selection. This is where the sizing method choice is genuinely open, because the shafts and ceiling depths have not yet been fixed. The single highest leverage moment on any project is the point at which duct routing is tested against structural depth and ceiling height, and that moment sits earlier than most programmes assume.
Tenant improvements and fit outs
Fit outs inherit an existing trunk, an existing shaft and an existing fan, and the design question changes from what should this be to what will this system tolerate. The available static pressure at the connection point becomes the binding constraint, and it is frequently unknown. Measuring it rather than assuming it is the difference between a fit out that balances and one that does not.
Alterations, additions and system replacement
Replacing an air handler onto existing ductwork is the most common duct design problem in the California market and the most commonly mishandled. The new unit has a different fan curve, different internal component losses and often a different discharge configuration. Reconfiguring ductwork also brings sealing and leakage verification obligations into play under the energy code, which is a compliance trigger owners rarely budget for.
Specialist air systems
Smoke control ductwork, stair pressurisation, laboratory fume exhaust, kitchen grease duct and industrial process exhaust all follow their own rules on construction, material, velocity and fire rating, and are sized and specified separately from comfort air. They should never be folded into a general supply air sizing exercise.
📖 Also Read: Why HVAC Design Belongs Before Architectural Finalisation for how early routing decisions protect ceiling heights.
5. What Duct Design Asks of Every Other Trade
Duct is the largest physical object in the ceiling void, and its requirements land on almost every other discipline.
Architecture and structure
Duct routing sets shaft sizes, ceiling void depth and therefore floor to floor height. It requires penetrations through beams and walls that must be structurally coordinated and fire rated. Shaft dimensions in particular are effectively irreversible once the structural grid is set, and a shaft sized from a preliminary airflow figure rather than a routed duct layout is one of the most reliable sources of late redesign. This is exactly the coordination work that mechanical design services should be resolving during design development, not during construction.
Electrical and controls
Fan power drives the mechanical electrical load, and a fan selected against an inflated static pressure figure oversizes the motor, the starter, the feeder and sometimes the upstream distribution. Variable speed drives, control dampers and terminal unit controllers all need power and network connections routed alongside the duct they serve. Where a duct system carries smoke control or pressurisation duty, damper actuators become life safety circuits with their own power and monitoring requirements.
Fire protection
Every duct penetration of a rated assembly needs a fire or combination fire and smoke damper, and every damper needs access for inspection and testing. Damper locations are a duct design output, not a fire protection afterthought, and their pressure loss belongs in the static calculation. NFPA standards governing air conditioning and ventilating systems set the construction and installation rules those dampers follow.
Acoustics and interiors
Three separate noise mechanisms come out of ductwork. Fan noise travels along the inside of the duct and is treated with lining, silencers and length. Regenerated noise is created locally at dampers, sharp takeoffs and terminal units, and cannot be silenced upstream because it is generated downstream. Breakout noise passes through the duct wall into a space the duct merely crosses, and is treated with mass rather than absorption, which means external lagging or double wall construction rather than lining.
The distinction matters because the wrong treatment does nothing. Adding a silencer upstream of a damper that is regenerating the noise is a common and expensive error. So is specifying internal lining after sizing, which reduces free area and raises velocity in the section that was already the noisiest.
Is Your Ceiling Void Deep Enough for the System You Have Specified?
The answer is cheap to establish during schematic design and extremely expensive to establish during clash detection. A short routing study against the structural grid usually settles it in days.
6. Codes, Standards and Pressure Classes
Duct design is governed by a layered set of documents. The California Mechanical Code, Part 4 of Title 24, governs construction and installation. The California Energy Code, Part 6, governs fan power, sealing, insulation and leakage verification. SMACNA construction standards define how duct at each pressure class is actually built. ASHRAE provides the design methods and, through Standard 90.1, the national energy baseline that Title 24 exceeds.
Pressure class and leakage class
SMACNA pressure classes describe what the duct is built to withstand, and the recognised classes are plus or minus one half, 1, 2, 3, 4, 6 and 10 inches of water gauge. The construction standards are explicit that contract drawings must assign a pressure class to each portion of the system or assign one to the whole system, and that where nothing is stated the duct defaults to the 1 inch class. Leaving pressure class unspecified on a medium pressure trunk is therefore not a neutral omission. It is an instruction to build it too light.
Leakage class describes the air the resulting construction is expected to lose. The relationship is F equals CL multiplied by P raised to the power 0.65, where F is leakage in cubic feet per minute per 100 square feet of duct surface, CL is the leakage class constant and P is static pressure in inches of water gauge. The standard classes are 3, 6, 12, 24 and 48, with lower numbers meaning tighter duct.
| Construction Pressure Class | Rectangular Metal Duct | Round Metal Duct | Where It Typically Applies |
|---|---|---|---|
| One half, 1 and 2 in. wg | Leakage Class 24 | Leakage Class 12 | Low pressure supply downstream of terminal units, return air, general exhaust |
| 3 in. wg | Leakage Class 12 | Leakage Class 6 | Medium pressure supply upstream of VAV terminal units |
| 4, 6 and 10 in. wg | Leakage Class 6 | Leakage Class 3 | High pressure trunks, industrial and process air systems |
Two things follow from that table. Round duct achieves a tighter leakage class than rectangular duct at every pressure class, because it has fewer longitudinal seams and fewer transverse joints per unit of surface. And a leakage class is a property of construction quality, not of intention, which is why leakage is verified by test rather than accepted on a submittal.
Sealing, insulation and California requirements
Section 120.4 of the California Energy Code sets the mandatory requirements for air distribution system ducts and plenums, covering sealing, insulation and, for the systems it captures, leakage verification through the acceptance test process. Duct insulation requirements vary with where the duct runs, with ducts outside the conditioned envelope carrying the heaviest obligations. Sealing materials must be listed for the service, and the acceptance testing procedures for air distribution systems are set out in the reference appendices to Part 6, published by the California Energy Commission. Where testing applies, the certificate of occupancy depends on the result.
📖 Also Read: Duct Leakage Testing Requirements Under California Title 24 for what the test involves and which systems it captures.
7. Healthcare, Laboratories, Data Centres and Aviation
Four building types change the duct design problem enough to deserve separate treatment.
Healthcare
Hospital ductwork is governed by pressure relationships between spaces, not just by airflow quantities. Operating rooms, isolation rooms, sterile processing and pharmacy compounding suites each carry directional airflow requirements that must survive filter loading, damper drift and system rebalancing over years of operation. Internal acoustic lining is commonly restricted or prohibited on cleanability grounds, which removes the easiest acoustic tool and forces the solution towards double wall duct, external lagging and inline silencers. Our healthcare facilities work sits under a separate plan review track, which makes late duct changes considerably more painful than they are on a commercial project.
Laboratories
Fume exhaust is a constant velocity problem, not an equal friction one. Duct material is dictated by the chemistry being exhausted rather than by cost, exhaust systems frequently run at high negative pressure requiring a high pressure class, and discharge velocity at the stack is a dispersion requirement rather than a duct sizing outcome. High air change rates mean laboratory duct is physically enormous relative to floor area, which makes shaft coordination the governing constraint in most laboratory and industrial buildings.
Data centres and mission critical
Air distribution in data halls is a containment problem more than a duct problem, but where ducted systems are used the static pressure calculation carries redundancy implications that comfort systems do not. A fan sized on the assumption that all units run simultaneously behaves differently when one is out for maintenance, and the duct network has to hold its pressure profile in both states. See our mission critical capability for context.
Aviation and large public buildings
Terminal buildings, concourses and large halls involve very long duct runs, high ceilings, exposed ductwork that becomes an architectural element, and acoustic environments where announcement intelligibility is a design requirement rather than a preference. These are the systems where static regain earns its complexity, because a single trunk serves takeoffs over distances at which equal friction produces genuinely unworkable imbalance.
📖 Also Read: Enhancing Indoor Air Quality Through MEP Solutions for how distribution design affects delivered air quality rather than just supplied quantity.
8. Choosing a Sizing Method
The method decision should be made when the system type is fixed and the routing concept is understood, which in practice means at the end of schematic design. The table below sets out how the choice usually resolves.
| Factor | Equal Friction | Static Regain | Constant Velocity |
|---|---|---|---|
| Sizing rule | Constant pressure loss per unit length | Velocity reduction regain offsets next section friction | Velocity held fixed, pressure loss falls out |
| Best suited to | Low and medium pressure commercial supply, return and general exhaust | Long high velocity trunks, large VAV systems, terminals, arenas | Grease, fume, dust and process exhaust that must maintain transport velocity |
| Balance behaviour | Needs balancing dampers, imbalance grows with run length difference | Close to self balancing along the trunk | Not a balance method, transport velocity governs |
| Calculation effort | Direct and fast | Iterative, benefits from software | Simple, but material and code driven |
| Space demand | Predictable, reduces steadily along the run | Larger downstream sections, often where space is tightest | Constant section for constant flow, large in long runs |
| Sensitivity to fitting quality | Moderate | High, poor transitions destroy the regain the method relies on | Moderate, but fitting geometry affects deposition |
| Energy outcome | Good when runs are reasonably symmetrical, poorer when they are not | Strong on long trunks, less damper throttling | Not optimised for energy, safety governs |
| Typical failure mode | Starved far branches, throttled near branches, damper noise | Downstream duct will not fit the space available | Excessive pressure loss on long runs, oversized fan |
The pattern that causes most trouble is a system sized on equal friction that has one branch three times the length of the others. Everything looks correct on the schedule, and the balancing contractor is left to throttle the short branches down to match the long one, which wastes the fan energy the friction rate was chosen to save and generates damper noise in exactly the spaces that are closest to the fan.
Ask one question before choosing a method. What is the ratio between the longest and the shortest run from the fan to a terminal? If it is close to one, equal friction will behave well. If it is three or more, either the layout needs rethinking or the trunk needs static regain, because no amount of damper adjustment will make that system quiet and efficient at the same time.
9. Static Pressure, System Effect and Where It Fails
Fan selection depends on one number: the total static pressure the system imposes at design airflow. That number is the sum of every loss along the single worst path from the fan to the most remote terminal, which is the critical path. Losses on any other path are irrelevant to fan selection, because those branches will simply be dampered back.
What actually makes up total static pressure
Engineers new to the discipline expect straight duct friction to dominate. In a compact commercial system it is often the smallest contributor. The table below sets out the components in the order they usually surprise people.
| Loss Source | Where It Occurs | Typical Magnitude | Why It Gets Missed |
|---|---|---|---|
| Air handler internal components | Coils, filters, heat recovery, humidifiers, mixing section | Usually the largest single group in a comfort system | Submittals quote clean filter loss, but the fan must be selected at final filter resistance |
| Fittings, elbows and transitions | Every direction and size change on the critical path | Expressed as equivalent length of straight duct; a square unvaned elbow can be worth many times its physical length | Fitting geometry is left to the shop drawing stage, after the fan is bought |
| System effect at the fan | First few diameters either side of the fan | Can be a substantial share of total static where the fan is boxed in or discharges into a wall | Appears on no schedule, no friction chart and no submittal |
| Takeoffs and branch connections | Every branch off the trunk | Conical and bellmouth takeoffs cost a fraction of a straight tap; the difference compounds across dozens of branches | Drawn as a single line on a plan with no detail called out |
| Terminal units and control dampers | VAV boxes, fan powered boxes, volume dampers | Fixed component loss plus regenerated noise when throttled | Minimum inlet static requirements are read from a catalogue rather than confirmed for the selected size |
| Silencers and acoustic treatment | Usually near the fan or the terminal | Buys attenuation and spends static pressure in the same component | Added late for acoustic reasons after fan selection is complete |
| Grilles, registers and diffusers | At the room | Sets the minimum pressure that must remain at the end of the critical path | Taken from a catalogue value without checking the actual neck velocity |
| Straight duct friction | The full length of the critical path | Proportional to length at the chosen friction rate; often smaller than expected in compact buildings | Rarely missed, because it is the only loss the friction chart gives directly |
| Duct leakage | Distributed across the whole system | Governed by construction pressure class and leakage class, verified by test | Assumed to be zero in the calculation and then discovered in commissioning |
Fittings and equivalent length
Fitting losses are usually expressed as an equivalent length of straight duct, which makes them easy to add to a friction calculation and easy to underestimate. The important insight is not the arithmetic but the sensitivity: the difference between a well proportioned radius elbow and a square mitred elbow with no turning vanes, in the same physical position, is large enough that a handful of poorly detailed fittings can outweigh a hundred feet of straight duct on the same run. The same applies to abrupt versus gradual transitions and to straight taps versus conical takeoffs.
This is why fitting details belong in the contract documents rather than being left to the sheet metal shop. A calculation performed on radius elbows and installed with square ones is not a calculation, it is a hope.
System effect, the loss nobody schedules
System effect is the gap between how a fan performs in the laboratory, on a clean straight test setup, and how it performs installed in a real plant room. An elbow hard against the inlet, an inlet that draws from a plenum with swirl in it, or a discharge that meets a wall within a duct diameter all reduce the pressure the fan can actually develop at a given airflow. The fan does not know it has been treated badly. It simply fails to deliver its curve.
The mitigation is straight duct at the inlet and the outlet. AMCA Publication 201 guidance is that the effective length of outlet ducting should be no less than 2.5 duct diameters at a velocity of 2,500 feet per minute or less, with one further diameter added for every additional 1,000 feet per minute. That is a real dimensional requirement that has to be reserved in the plant room at layout stage, and it is almost always the first thing given up when the mechanical room shrinks.
System effect is the most commonly missed loss in the entire system. It does not appear on the friction chart, it is not in the equipment submittal, and it survives every review because there is no line item to check it against. The only defence is to draw the fan inlet and outlet at scale, in plan and section, and confirm the straight length exists before the plant room is signed off. Lawrence Berkeley National Laboratory research on fan and air distribution performance has consistently found installed system efficiency falling well short of component ratings for exactly these reasons.
The failure catalogue
Five failures account for most of the money lost to duct design on California projects.
The oversized fan chasing a pessimistic static calculation. When nobody is confident in the static pressure number, the instinct is to add margin. Margin is added at the coil, at the filter, at the fittings and again as a global safety factor, and the compounded result is a fan selected far to the right of the real system curve. That fan runs off its efficiency peak, draws a larger motor and feeder than needed, generates more noise than the acoustic analysis assumed, and on a constant volume system simply delivers too much air. Accurate calculation is cheaper than margin, and it starts with an honest load calculation feeding an honest airflow schedule.
The shaft discovered to be undersized in coordination. Shafts are sized early, often from a rule of thumb, and fixed by the structural grid. When the routed duct layout arrives and does not fit, the options are all bad: flatten the duct, split the system, take ceiling height, or move the structure. This is precisely the failure that MEP BIM coordination and clash detection is meant to prevent, and it is only prevented if the model contains real routed duct with real fitting geometry early enough to matter.
Aspect ratios forced by the ceiling void. When the void loses depth, ducts flatten. Each flattening decision is small and individually reasonable, and the cumulative effect is a system whose friction, sheet metal cost and breakout noise are all permanently higher than designed. The counter is to fix a maximum aspect ratio in the specification and treat exceedances as changes requiring engineering review rather than as field decisions.
No straight duct at the fan. The plant room is laid out for equipment footprints, not for airflow, and the fan ends up with an elbow at the inlet and a wall at the discharge. The building then underperforms in a way that is very difficult to diagnose after the fact, because every component tests correct in isolation. Catching this needs early MEP coordination, at the point where the mechanical room is being sized rather than after it has been.
Duct leakage testing failures under Title 24. Where the energy code requires sealing and leakage verification, a failed test arrives at the worst possible moment: after installation, usually after ceilings are closed, with the certificate of occupancy waiting. The remedial cost is dominated by access and reinstatement rather than by sealant. The practical defences are to specify the pressure and leakage class explicitly on the drawings, to test a representative section early in the installation rather than the whole system at the end, and to treat sealing as an installation discipline rather than a rectification activity. Our article on duct leakage testing under California Title 24 covers the test procedure in detail.
Testing, balancing and commissioning
Duct design is proved in the field, not on the drawing. Testing, adjusting and balancing confirms that each terminal delivers its design airflow. Leakage testing confirms the construction met its specified class. Acceptance testing confirms the installation matches what the energy code compliance documents claimed. Functional performance testing under a commissioning services scope confirms the system behaves correctly across its operating range rather than only at the design point. Running these as one coordinated sequence, through the building commissioning process, avoids the duplicated site visits and contradictory findings that come from running them separately. The U.S. Department of Energy Building Energy Codes Program has consistently found that the cost of correcting energy performance problems rises steeply the later in the project they are discovered.
Key Takeaways
- Duct design converts an airflow schedule into a physical network and a static pressure number. Both outputs are load bearing, and the second one selects the fan.
- The friction rate is an economic decision with a thirty year tail, not a technical constant. Conventional commercial practice sits in a band of roughly 0.05 to 0.10 inches of water gauge per 100 feet.
- Velocity limits come from the room noise criterion, which is why acceptable velocity falls sharply as the duct approaches occupied space.
- Equal friction suits symmetrical commercial layouts, static regain suits long high velocity trunks, and constant velocity is for exhaust that must maintain transport velocity. Most real buildings use more than one.
- High aspect ratio ducts cost more sheet metal once and more fan energy forever. Fix a maximum in the specification and defend it in coordination.
- Fitting losses and system effect routinely outweigh straight duct friction. Detail the fittings in the contract documents rather than leaving them to the shop.
- Reserve straight duct at the fan inlet and outlet at plant room layout stage. AMCA Publication 201 guidance calls for at least 2.5 duct diameters of outlet duct at 2,500 feet per minute or less, plus one diameter per further 1,000 feet per minute.
- Assign a SMACNA pressure class and leakage class on the drawings. Unspecified duct defaults to the 1 inch class, and leakage is verified by test rather than accepted on a submittal.
10. Where Budlong Works
Duct design decisions are national in principle and local in practice. Ceiling void depth is negotiated against local height limits and zoning envelopes, plan check expectations for compliance documentation vary between jurisdictions, and local reach codes in several California cities add requirements above the state baseline. Budlong maintains teams across the state so that local plan check experience and local construction practice sit behind every set of duct drawings. More on how the practice is structured is on our company story page.
Sectors where duct design decides the outcome
Healthcare carries pressure relationship requirements that outrank energy optimisation. Laboratories are dominated by exhaust volumes that make shaft coordination the governing constraint. Mission critical facilities need a pressure profile that holds under redundancy scenarios as well as at design. Schools and universities combine tight acoustic criteria with tight budgets, which is where sizing method and fitting detailing pay for themselves most visibly.
11. Related Reading
12. Frequently Asked Questions
What is HVAC duct design?
HVAC duct design is the process of converting a room by room airflow requirement into a physical distribution network of sized ducts, fittings, dampers and terminals, and then calculating the total static pressure that network imposes on the fan. It sits between the load calculation, which says how much air each space needs, and the fan and air handler selection, which says how that air gets moved. Get it wrong and the load calculation becomes irrelevant, because the air never arrives where the calculation said it should.
What is the difference between equal friction and static regain duct sizing?
Equal friction sizes every section of duct to lose the same pressure per unit of length, so the duct steps down in size as air is dropped off along the run. Static regain sizes each downstream section so that the static pressure recovered from the reduction in velocity roughly offsets the friction of the next section, which leaves approximately the same static pressure available at every takeoff. Equal friction is faster to produce and easier to check. Static regain is close to self balancing and suits long high velocity trunks, but it is iterative, produces larger downstream ducts and needs more space.
What friction rate should be used for duct sizing?
There is no code mandated friction rate. Conventional commercial practice for low pressure supply, return and exhaust sits in a band of roughly 0.05 to 0.10 inches of water gauge per 100 feet, with tighter values chosen where duct runs are long, ceiling space is generous or fan energy is being optimised. The friction rate is an economic decision, not a technical constant. A lower rate means larger ducts, more sheet metal and more ceiling depth but permanently lower fan energy. A higher rate saves capital cost and space and spends it on electricity for the life of the building.
Why are duct velocity limits set by noise rather than by physics?
Air will move through a duct at almost any velocity you ask of it. What fails first is the acoustic environment, not the airflow. High velocity generates broadband noise inside the duct, regenerates additional noise every time the air passes a damper, a takeoff or a sharp fitting, and breaks out through the duct wall into the space below. Velocity limits in duct design are therefore derived backwards from the room noise criterion, which is why acceptable velocities fall sharply as the duct gets closer to the occupied space and why a runout above a private office is held far slower than a shaft riser.
What is system effect at a fan and why is it missed so often?
System effect is the loss in fan performance caused by disturbed airflow at the fan inlet or outlet, compared with the clean laboratory conditions under which the fan was rated. An elbow hard against the inlet, a fan discharging straight into a plenum wall, or swirl induced by an upstream fitting all mean the fan produces less than its catalogue curve says. AMCA Publication 201 guidance is that the effective length of outlet ducting should be no less than 2.5 duct diameters at a velocity of 2,500 feet per minute or less, with one additional diameter for every further 1,000 feet per minute. It is missed because it appears on no schedule, no friction chart and no equipment submittal, and because the space it needs is the first thing sacrificed in coordination.
What are SMACNA duct pressure classes and leakage classes?
SMACNA pressure classes describe the construction standard a duct is built to, expressed as the static pressure it is designed to withstand. The recognised classes are plus or minus one half, 1, 2, 3, 4, 6 and 10 inches of water gauge. Leakage classes describe how much air the resulting construction is expected to lose, using the relationship F equals CL multiplied by P raised to the power 0.65, where F is leakage in cubic feet per minute per 100 square feet of duct surface. The standard classes are 3, 6, 12, 24 and 48, with lower numbers meaning tighter duct. Round duct achieves a tighter leakage class than rectangular duct at the same pressure class because it has fewer seams.
Why does duct aspect ratio matter so much?
A flat duct and a square duct of the same cross sectional area do not carry air equally. The flatter duct has a longer perimeter, so it presents more wetted surface to the airstream and generates more friction for the same airflow. It also uses more sheet metal, needs heavier gauge and more reinforcement because of its larger single dimension, and loses more heat through its larger surface. The capital cost penalty is paid once. The fan energy penalty is paid every hour the system runs for the life of the building, which is why aspect ratio is one of the few design decisions worth defending hard in a ceiling void negotiation.
What happens if a duct system fails leakage testing in California?
Where the California Energy Code requires duct sealing and leakage verification, the work is documented through the acceptance test process and the certificate of occupancy depends on it. A failure means the duct has to be found, opened up, resealed and retested, on a building that is already built and frequently already has ceilings closed. The cost is not the sealant. It is the access, the ceiling demolition and reinstatement, the retest and the schedule. Sealing to the specified class during installation, and testing a representative section early rather than the whole system at the end, is the only reliable way to avoid it.
Should ducts be internally lined or externally lagged for acoustics?
They solve different problems. Internal acoustic lining attenuates noise travelling along the inside of the duct, which is what you need between a fan or a terminal unit and the room it serves. External lagging adds mass and a damping layer to the duct wall, which is what you need where noise is breaking out through the duct into a space it passes over rather than serves. Lining also reduces the free area and adds friction, so it has to be in the sizing calculation rather than added afterwards. In healthcare and laboratory work, internal lining is often restricted or prohibited on infection control and cleanability grounds, which pushes the solution towards double wall duct, external lagging and inline silencers.
