PP Duct Sizing: Velocity, Friction Rate and Fan Static

Written by Bofan Deng · Reviewed by XICHENG EP Engineering Department

Key Takeaways

• PP duct sizing is the result of a loop, not of a formula. Airflow sets the velocity floor, the floor sets the theoretical diameter, the stock series moves that diameter, and the new diameter changes the velocity you started with.
• The diameter is an output, not an input. The four things a fan is selected on are the air volume, the fan static pressure, the contaminant, and how much noise the site will accept — and none of them is a duct size.
• Rounding up is not automatically safe on an exhaust duct. A 5,800 m³/h duty between φ400 and φ500 gives 13.42 m/s in the smaller size and 8.58 m/s in the larger one, so the round-up lands below the transport floor and the mist starts to settle.
• The seven standard production sizes do not tile every flow at every velocity band. The gap between φ400 and φ500 is wide enough to swallow a whole transport band.
• A polypropylene duct loses about 14 to 19% less pressure than a galvanized duct of the same bore at the same velocity, so a steel duct calculator over-reads a PP system.
• The duct is a minor term in the fan selection. In the worked example here the scrubber is 77% of the total static pressure and the whole duct run is 15%.

PP Duct Sizing Is a Loop, Not a Formula

One more input can re-open the loop, and it is a material choice rather than a sizing one. Specifying a flame-retardant grade lowers the tensile strength of the sheet by roughly 10 to 15%, which pushes the required wall thickness up for the same duty; a thicker wall at a fixed outside diameter narrows the bore, and the narrower bore raises the velocity, which sends the designer back to the transport floor and the friction rate. The requirement is worth carrying into the first pass rather than discovering after the fan has been selected. Where it bites, and by how much, is set out in our guide to a flame retardant PP duct.

Search for pp duct sizing and almost every result gives the same two lines of algebra: the area is the airflow divided by the velocity, and the diameter is the square root of four times that area over pi. Both lines are correct. Both take about ten seconds to evaluate. Neither one is the part of the job where a duct gets sized wrongly.

What actually decides the diameter is a set of inputs that sit outside the calculation, and a constraint that only appears after it. The velocity is not a number you pick for convenience — it has a floor below which the contaminant stops being carried. The stocked series does not contain every diameter the arithmetic produces. And the size you finally select changes the velocity you used at the start, which changes the pressure loss, which changes the fan. That circularity is why this page is about a loop rather than a formula.

The Five Numbers the Loop Ties Together

Five quantities are in play, and each one is defined in terms of the others. The airflow is set by the process and the treatment equipment, not by the duct. The velocity floor is set by what the stream is carrying. The theoretical diameter follows from those two. The stocked diameter is the nearest size the factory actually builds, which is where the loop closes because that choice moves the velocity. And the system static pressure is what the fan has to be able to deliver at that airflow, which depends on the velocity and the diameter together.

Read in that order the loop looks linear, and for a simple run it very nearly is. It stops being linear the moment the stocked size is far from the theoretical one, or the moment the stream carries something that settles. Both of those happen routinely in industrial exhaust work, and both are covered below.

Why the Formula Is the Easy Part

The arithmetic is genuinely trivial, and it is worth saying so, because treating it as the difficult step is what produces confident wrong answers. A designer who starts from the formula has to guess the velocity, and the guess is usually inherited from a comfort-air handbook. Those handbooks work in a band suited to offices and homes, where the only costs of moving slowly are noise and a slightly larger duct. An exhaust duct has an additional cost at low velocity, and it is not financial: below a certain speed the droplets and particles stop travelling and start accumulating.

So the useful question is not how to compute a diameter. It is which velocity the computation should start from, and whether the answer the computation gives can actually be built. Everything on this page serves those two questions.

What Changes When You Size It This Way

The purchasing sequence changes too, and it is worth naming because it is the practical payoff. Sizing from the formula downward produces a diameter first and a supplier conversation second, which is how people end up asking whether a size is available after they have already designed around it. Sizing from the loop upward produces a band first, then a stocked size, then a pressure budget, and only then a price. By the time a quotation is requested, the four numbers a supplier needs to check the work — airflow, velocity, diameter and run length — are already settled.

The Three Inputs That Fix a Duct Diameter

Three things between them determine the diameter, and only one of them is under the designer’s control in any comfortable sense. Getting clear on which is which removes most of the frustration from the exercise.

Airflow: Set Upstream of the Duct

The airflow is the volume the system has to move, and it is determined before the duct exists. In a laboratory it comes from the sash opening and the face velocity the hood has to hold. On a process tank it comes from the capture velocity the hood needs to control the emission at a given distance. Where a wet scrubber is in the train, its own rating usually sets the number, because a packed tower is designed around a face velocity through a fixed bed area and will not perform as specified outside a narrow flow range.

The consequence is that the duct is a dependent component. It is sized to carry a flow that somebody else has already determined.

Velocity: a Floor, Not a Target

This is the point where industrial duct work parts company with comfort ventilation, and it is the single most useful idea on the page. In a comfort system, velocity is chosen inside a noise band and there is no lower limit that matters — slow air in an office duct is simply quiet air. In an exhaust duct there is a hard floor, and it is set by the contaminant rather than by the designer.

Below the transport velocity for the material being carried, particles and droplets drop out of suspension. They collect on the bottom of horizontal runs, reduce the free area, raise the local velocity, and in a wet or condensable stream the accumulation can set hard enough to plug the duct. A plugged exhaust duct is not a maintenance inconvenience; it is a loss of containment.

The floor is also contaminant-specific. Dry, clean fume can be carried at the lower end of the industrial range. Fine dust needs substantially more. A condensable vapour that can wet its own particles needs more again, because a wetted particle is heavier and stickier than the dry one it started as. This is why a single published velocity band for “industrial exhaust” is only a starting point, and why the approach velocity a supplier asks about is a question about the process rather than about the duct.

It is worth being explicit about where the divergence starts, because the same word is used on both sides of it. The velocity tables that circulate for supply and return air are drawn for occupant comfort, and they sit in a band well below the industrial transport range: a general reference of that kind lists recommended air duct velocities for comfort service, and reading those numbers into an exhaust duct is one of the most common sources of an under-velocity exhaust run. The number is not wrong. It was written for a system that has no transport requirement at all.

The Stocked Series: Where the Answer Has to Land

The theoretical diameter is almost never a diameter anyone sells. The stocked series moves in nominal steps, and the steps are not even — which is the first place a pp duct sizing calculation has to deal with a discrete ladder instead of a continuous diameter. Polypropylene duct in the standard production range runs 110, 160, 200, 315, 355, 400 and 500 mm, with named sizes above that built to order and gaps in between that have no size at all. The full series, the gaps, and how a metric size maps onto an inch trade label are set out in our guide to PP duct sizes.

Because the steps are uneven, the ratio between one stocked size and the next is uneven as well. Moving from φ400 to φ500 changes the bore by 25%, and since velocity is inversely proportional to the square of the bore, that single step drops the velocity by about 36%. A calculation that says “round to the nearest standard size” is therefore saying something much more consequential than it sounds.

Which Input You Can Actually Move

Of the three, the airflow is usually fixed by the process and the velocity floor is fixed by physics, which leaves the designer negotiating with the ladder. That negotiation is where the real decisions are made: accept a stock size slightly off the theoretical one, specify a made-to-order diameter, revise the airflow, or split the run.

It is worth remembering that the diameter is not the only thing the choice moves. Section length, support spacing and hanger load all follow from the diameter and the wall that goes with it, and the support arrangement for a φ500 run is not simply a scaled version of the one for a φ315 run. That part of the sequence is worked through in our guide to PP duct installation.

What Sets the Airflow You Are Sizing For

Since the airflow arrives as a given, it is worth knowing where it comes from, because knowing the source tells you how much margin exists and who to ask when it changes.

The Scrubber Rating Fixes the System Flow

A packed-bed wet scrubber is sized around a face velocity through a fixed cross-section, and its removal efficiency is quoted at that condition. Push more air through it and the face velocity rises, the contact time falls, and the efficiency the supplier quoted stops applying. Push less through and the bed may not wet evenly. The result is that a scrubber has a preferred flow rather than a maximum flow, and the duct system is designed around the figure the scrubber was sold at.

The reason this matters for sizing is that it removes a degree of freedom people expect to have. A duct can be sized for a range of flows. The scrubber at the end of it usually cannot, so the flow is effectively decided upstream and the duct inherits it.

Hood Capture Volume: the Number the Duct Inherits

Where there is no scrubber, the airflow comes from the hood. Capture volume is set by the face velocity required to control the emission at the working distance, and by how well the hood is enclosed. An enclosure with a small opening needs far less air than an open tank of the same size, which is why enclosure is the cheapest form of ventilation control available. The duct simply carries whatever the hood demands.

Two properties of capture air are worth carrying into the sizing exercise. It is a minimum, not a target, because it is derived from a control requirement rather than a comfort one. And it is sensitive to things that are not in the duct design at all — cross-draughts in the workshop, the position of the operator, and whether the hood stays where it was installed. A sizing calculation cannot recover a capture volume that was never achieved on site.

Adding a Branch Does Not Add Air to the Main

A common error at the enquiry stage is to size the main duct for the sum of what the branches could theoretically take, and then to size each branch as though it had its own dedicated airflow. In a real system the fan delivers a total volume, and the branches divide it according to their resistance, not according to what they were designed for. A branch with less resistance than the design assumed will take more than its share, and the branch at the far end of the run will take less.

Labelling each branch with its design airflow is still necessary — it is what the sizing calculation consumes. But it is a design intent, and it is not self-enforcing. That gap is treated properly in the section on balancing below.

Actual Volume and Standard Volume Are Not the Same Number

Airflow figures arrive in two different currencies and they are not interchangeable. A standard volume is corrected to a reference temperature and pressure, and it is the right currency for a mass balance or an emission calculation. An actual volume is the volume at the temperature and pressure actually inside the duct, and it is the right currency for sizing anything, because the duct has to pass a physical volume of gas at the conditions it is running at.

Sizing a duct on a standard volume while the gas is hot understates the diameter, sometimes badly. The correction is not a refinement; it is a factor that can exceed 30% on a hot stream. It gets its own section further down, because it is also the reason the duct before a scrubber and the duct after it may not need the same size.

The PP Duct Sizing Table: What Each Stocked Size Carries

There is a table that does most of the work in a pp duct sizing exercise, and it is not usually published by duct suppliers. It takes the seven standard production sizes and shows what volume each one carries at each of the transport velocity bands. Read left to right it tells you which size to buy. Read right to left it tells you what an existing run is capable of.

How the Table Is Built: Bore, Not Nominal Diameter

The table below is calculated on the internal bore of each size, because the bore is what the air passes through. A nominal φ400 section has a 4.5 mm wall and a 391.0 mm bore, and the area that wall removes is 4.5% of the cross-section. Building the table on nominal diameters would overstate the capacity of every size, and would overstate the small sizes much more than the large ones, because the wall is a fixed thickness being subtracted from a shrinking diameter: the same nominal-diameter shortcut overstates φ400 by 4.5% and overstates φ110 by more than 10%.

The bore, the wall thickness, the standard section length and the mass per metre for the whole production series are given in our guide to what a PP duct is. The diameters used here are those bores.

Capacity by Stocked Size at Each Transport Band

The columns are the velocity bands used for corrosive fume work, running from the quietest service at 8 m/s to dry particulate at 18 m/s. The figures are volume, in cubic metres per hour, that each size passes at that velocity.

Volumes are rounded to the nearest cubic metre per hour; the areas are the full-precision calculation rounded for display, so re-deriving a cell from the printed area will differ in the last digit.

Two features of the table are worth pointing out before it is used. The first is the size of the step between the last two rungs. At 10 m/s the step from φ400 to φ500 is 2,438 m³/h — more than twice the entire capacity of a φ200 duct at the same velocity. Someone sizing a 5,000 m³/h duty at 10 m/s finds that φ400 is slightly over its limit and φ500 is 35% oversized, and there is nothing in between.

The second is how quickly the small sizes run out. A φ160 branch at the 12 m/s needed for a wet mist carries 805 m³/h, which is a single fume cupboard’s worth. Anything larger than a couple of hoods moves the branch to φ200 or beyond, and the main to φ315 or more.

Reading the Table Backwards: Sizing a Duct You Already Have

The table is more useful in reverse than in forward for one common job: establishing what an installed run can carry. A φ315 duct already on site, asked to run at 12 m/s, carries 3,190 m³/h. If the process now needs 4,000 m³/h, the existing duct is not adequate at that velocity, and the options are to accept a higher velocity inside the band, to replace the run, or to add a parallel leg.

Working in this direction also exposes the constraint that catches people out on retrofit projects. A duct that is adequate at the design flow can be pushed out of its band by a modest process increase, and because the velocity rises with the square of the diameter change, a 20% flow increase raises velocity by 20% — enough to move a run from the quiet middle of its band to the erosion end of it. The duct does not fail; the stream starts wearing it away and the fan starts drawing more power.

What the Table Leaves Out

The table is a geometry table, and three things it does not carry can change the answer. It assumes the stream is at ambient conditions, so a hot or saturated gas needs the volume correction described later. It assumes a single run with no branch interaction, so a multi-branch system needs one of the balancing methods rather than a straightforward lookup. And it says nothing about pressure — a size that will pass the volume may not be available with a wall thick enough to hold the vacuum the fan creates.

The Duct Sizing Calculation, in Five Steps

With the inputs understood and the capacity table in hand, the calculation itself is short. The five steps below are the whole of a pp duct sizing method; the length comes from what each step refuses to assume.

Step 1: Theoretical Diameter From Airflow and Velocity

Convert the airflow to cubic metres per second if it is not already, divide by the design velocity, and the result is the required cross-sectional area. The diameter follows from the area of a circle. Working in cubes per second matters: dividing cubic metres per hour by metres per second gives an area that is wrong by a factor of 3,600, and the error survives a casual sanity check because the intermediate number still looks plausible.

For 4,600 m³/h at 12 m/s: the flow is 1.278 m³/s, the area is 0.1065 m², and the diameter is 368.2 mm. That number is a bore, and it is the first place the calculation has to be read carefully.

Step 2: From Nominal Diameter to the Bore You Flow Through

A 368.2 mm answer does not correspond to a 368.2 mm product, because products are sold on outside diameter. The comparison has to be made bore to bore. A nominal φ400 has a 391.0 mm bore, so it is larger than the requirement; a nominal φ355 has a 346.6 mm bore, so it is smaller. The theoretical figure sits between two stocked sizes, which is the normal outcome.

Skipping this step is the most common arithmetic error in duct sizing, and it always errs in the same direction: the nominal diameter is larger than the bore, so comparing a required bore against nominal sizes makes the smaller size look adequate when it is not.

Step 3: Round to a Stocked Size, in Either Direction

Standard advice at this point is to round up, on the reasoning that a larger duct is quieter and wastes less fan energy. That reasoning comes from comfort ventilation, where a larger duct has no downside except material cost. On an exhaust duct the downside is real: a larger duct lowers the velocity, and the velocity has a floor.

So the rounding direction is a decision rather than a rule. Round up when the velocity stays above the floor. Round down when it does not, provided the resulting velocity stays under the erosion and noise ceiling. And when neither neighbour in the stocked series satisfies both limits, the calculation has found a genuine problem, which the next section deals with.

Step 4: Re-read the Velocity and Test It Against Both Limits

The velocity at the chosen size is the airflow divided by that size’s actual area, and it is not the velocity that was used in step 1. This is the step that closes the loop, and it is the one most often left out, because the calculation looks finished once a size has been chosen.

Test the result against both ends of the band. Below the floor the stream settles; above the ceiling the friction cost climbs steeply and a particulate or mist-laden stream begins to erode bends from the inside. A calculation that only checks the upper end has checked the less likely failure.

Step 5: Repeat Until the Diameter Stops Moving

When the velocity changes, so does the friction loss, and when the friction loss changes the fan selection can change, which occasionally changes the flow and therefore the diameter. In practice this converges in one or two passes, because the velocity at a stocked size is usually within about 10% of the design velocity and because the friction term is a modest share of the total pressure. The loop rarely oscillates, but it does need to be closed rather than abandoned at step 1.

When Rounding Up Puts You Under the Floor

Here is the case that makes the difference between the two philosophies concrete, and it happens often enough in exhaust work that it is worth having a method for it rather than a preference.

The 5,800 m³/h Case: Both Neighbours Sit Outside the Band

Take a pickling line producing 5,800 m³/h of acid mist, and take the 10 to 12 m/s band that wet mist needs. Depending on where in the band the design velocity is taken, the theoretical bore falls between 413 mm and 453 mm — comfortably between the φ400 and φ500 bores. Now evaluate both stocked neighbours at the actual flow.

In φ400 the bore is 391.0 mm, the area is 0.1201 m², and the velocity is 13.42 m/s. That is above the 12 m/s ceiling for a mist-laden stream, in the region where erosion of bends and droplets re-entraining off the duct wall become real. In φ500 the bore is 489.0 mm, the area is 0.1878 m², and the velocity is 8.58 m/s. That is below the 10 m/s floor, and 8.58 m/s is not marginally below it: the mist will begin to drop out in the horizontal sections and pool at the low points.

Both stocked neighbours fail, in opposite directions. The theoretical diameter was right and the ladder cannot deliver it.

When the Band Falls in a Ladder Gap

The cause is the shape of the series rather than bad luck. The production series steps by 100 mm at this point in the range, and a 100 mm step on a 400 mm duct is a 25% change in bore and roughly a 36% change in velocity. Any velocity band narrower than about a third of its own value can fall entirely inside one step. The 10 to 12 m/s band spans only 20% of its value, so it fits inside the gap with room to spare.

This is not a defect in the standard range, which is built around the sizes that most work falls on. It is a property that has to be designed around, in the same way as any other discrete series. The gaps are described as part of the size system in the guide to PP duct sizes, and the transport bands that produce them are described in the guide to what a PP duct is. What this page adds is the arithmetic that shows when a gap is going to bite.

The Three Ways Out

There are four practical resolutions, and they are worth listing in the order they are normally tried.

Specify the size that is not stocked. A φ450 duct is a named size in the wider series and a made-to-order item in the production range. It has a bore that takes 5,800 m³/h at close to 11 m/s, which is inside the band. It costs a forming setup and it lengthens the lead time, which is the honest price of landing on the right velocity.

Change the airflow. Where the scrubber or the hood has any tolerance, trimming the flow moves the theoretical diameter. Reducing 5,800 to 5,200 m³/h drops the theoretical bore to 428.8 mm, which is still in the gap, but it moves φ400 to 12.03 m/s, close enough to the ceiling to be worth checking against the actual erosion duty.

Split the run. Two parallel φ315 legs carry 5,800 m³/h at about 10.9 m/s each, squarely inside the band, at the cost of a duplicated run, twice the fittings and twice the supports. On a long route the material cost of the second leg is usually what kills this option.

Accept the smaller size with a documented reason. If the mist is fine, non-erosive and not prone to wetting solid particles, running φ400 at 13.42 m/s is a defensible engineering decision — it costs fan power and it wears the bends faster, but it does not break containment. What makes it defensible is that the trade was written down, not that it was overlooked.

Why This Rarely Shows Up in Comfort-HVAC Sizing

The reason this section has no equivalent in a residential duct guide is that comfort systems have no velocity floor. Slower air in a supply duct is quiet air, so every stocked step larger than the theoretical diameter is an improvement, and “always round up” is a rule that cannot fail. Import that rule into an exhaust system and it can fail immediately, because the direction that reduces noise and friction is the direction that lets the contaminant settle.

The rule is not wrong in its own setting. It is a comfort rule applied to an industrial problem, which is a different failure from a wrong calculation and a harder one to catch, because the arithmetic will look correct all the way through.

The Friction Rate of a PP Duct

The pressure a duct loses along its length is the part of the system the diameter actually controls, so it is worth having the number for each stocked size rather than a chart built for steel. The figures below are computed for polypropylene at the transport velocities, and they are the second table in a pp duct sizing exercise that is normally missing.

Roughness, Reynolds Number and the Friction Factor

Friction loss in a duct is governed by the Darcy–Weisbach relationship, in which the loss per unit length is proportional to the velocity pressure and to a friction factor, and inversely proportional to the duct diameter. The friction factor itself is not a constant: it depends on the Reynolds number, which describes how turbulent the flow is, and on the relative roughness, which is the surface roughness divided by the diameter.

Reynolds numbers in an exhaust duct run into the hundreds of thousands, which is deep in the turbulent regime. At those numbers the flow is hydraulically smooth up to a size-dependent point, and the relative roughness is what separates one material from another. Extruded polypropylene is smoother than most duct materials and its relative roughness stays small even in the larger sizes, which is why its friction factor sits in the 0.014 to 0.020 range across the series rather than climbing the way a rougher material’s does.

Pressure Loss per Metre and per 100 Feet, by Size

Applying that friction factor at the transport velocities gives the loss per metre of straight duct. At 10 m/s the pressure loss ranges from 11.39 Pa/m in the smallest size down to 1.77 Pa/m in φ500, and the ratio between the ends of the series is a little over six to one.

Figures are computed for ambient air at 1.2 kg/m³. Both columns matter to the design and neither is a target: they are what the duct does at the velocity the sizing selected. The imperial column is included because a great many duct charts are drawn in inches of water per hundred feet, and the comparison in the next subsection depends on being able to read one against the other.

Why a Galvanized Ductulator Over-Reads a PP Duct

Almost every duct friction chart in circulation is drawn for galvanized steel. The comparison in the other direction is given in our guide to PP duct versus galvanized duct: polypropylene’s surface is materially smoother, and at the Reynolds numbers an exhaust duct runs at, that difference produces a friction factor roughly 14 to 19% lower than galvanized at the same bore and velocity.

The practical effect is a systematic error in the safe-looking direction. Sizing a PP duct off a steel chart overstates its pressure loss by about a sixth, which makes a duct look marginally undersized and pushes the designer toward a larger size. On a comfort system that miscost material. On an exhaust system, where a larger duct lowers the velocity, the same error can push a run toward the settling floor. The fix is simply to use a friction rate computed for the material rather than borrowed from a chart for another one.

Turning Fittings Into Equivalent Metres

Straight duct is only part of the resistance. Every bend, tee and transition adds a local loss expressed as a loss coefficient multiplied by the velocity pressure, and the practical way to add those to a length calculation is to convert each one into an equivalent length of straight duct of the same size. That conversion, and the loss coefficients of the fitting families, are treated in our guide to PP duct fittings.

Two things about that conversion matter for sizing. The first is that a fitting’s loss is fixed by the velocity rather than by the diameter, so the pressure a fitting costs depends on the size choice only through the velocity it produces. The second is that local losses are not minor. On a short run with several bends, the fittings can outweigh the straight duct they sit in, and a budget built from the length alone will come up short.

The System Static-Pressure Budget, Term by Term

The budget is what turns a sized duct into a specified fan, and it is the part of a pp duct sizing exercise that most published guides decline to attempt. It is not difficult. It is six terms, and only one of them is the duct.

Hood Entry Loss: K Times the Velocity Pressure

Air entering a hood turns a corner and accelerates, and both cost pressure. Industrial ventilation practice expresses the loss as a loss factor multiplied by the duct velocity pressure, with published loss factors for the standard hood types. A plain flanged opening is at the higher end; a well-designed hood with a smooth transition into the duct is at the lower end. The convention used here follows OSHA’s ventilation investigation guidance, which states the entry loss as the loss factor times the velocity pressure and tabulates the coefficient by hood geometry.

The term belongs in the budget rather than in the hood design discussion because it is a real pressure the fan has to supply, and because it is set by the velocity the sizing chose. Increasing the duct diameter lowers the velocity pressure and therefore lowers the entry loss — one of the few places where a larger duct helps a term other than its own friction.

Straight-Run Friction: the Only Term the Diameter Moves

This is the term the friction table above gives, multiplied by the length of the longest path. It is the only term in the budget that responds materially to the diameter choice, and because it is a modest share of the total, the diameter’s influence over the whole budget is modest in proportion.

Put a number on that early, because it shapes the rest of the design. On a 25 m run at 10 m/s the straight duct contributes tens of pascals. The treatment device a few sections down contributes an order of magnitude more. A designer who has internalised the comfort-airframe expectation — that duct diameter is the main lever on fan pressure — will spend the design budget in the wrong place.

Fittings and Transitions

The local losses from the previous section, summed along the longest path, expressed either as loss coefficients times the velocity pressure or as equivalent lengths. Either route gives the same number, and the choice is a matter of which the designer finds easier to audit.

The Scrubber

A packed-bed wet scrubber at the scale of a few thousand cubic metres per hour typically drops somewhere between several hundred pascals and well over a thousand, depending on the packing depth, the face velocity and whether there is a mist eliminator downstream. It is usually the largest single term in the system and it can exceed the entire duct run several times over.

It is also the term that makes the diameter choice less financially consequential than it looks, a point that follows from the budget arithmetic rather than from any preference about how to design. The reasoning behind it, and how a treatment device sits in the exhaust train, is set out in the guide to what a PP duct is. The budget below uses a single figure for it and shows what share of the total it takes.

Stack Exit: Velocity Pressure You Pay For and Throw Away

The last term is the one most often forgotten. Air leaving the stack has kinetic energy, and that energy came from the fan and is not recovered. The exit loss is the velocity pressure at the stack outlet, so the stack outlet velocity should be a deliberate design value rather than a by-product of the stack diameter.

There is a lower limit on it as well as a cost. Industrial ventilation guidance puts the minimum stack exit velocity at about 1.4 times the local wind velocity, so that the plume is ejected clear of the stack rather than bent back down around it. On a site with a 5 m/s design wind that means a minimum exit velocity around 7 m/s, which sets a maximum stack diameter for a given flow — a stack sizing constraint that runs in the opposite direction to the duct sizing one.

Adding the Column Up

Sum the six terms and the result is the fan static pressure the system demands at the design flow. That single number is what a fan is selected against, and it is worth more than any of the intermediate values, because it is the only figure that can be checked against a fan curve.

It is also worth keeping the terms separate rather than only keeping the total. When the fan comes back from the supplier oversized or the flow comes back low, the distribution of the budget is what tells you which term is worth revisiting, and the duct is rarely the one that repays the effort.

The Fan Duty Point, Not the Fan Rating

A fan has a rating and a curve, and they are not the same thing. The rating is a single point, usually quoted at a nominal static pressure that may or may not resemble the system it is going into. The curve is the set of flow and pressure pairs the fan can produce, and where the system actually operates is the point at which the curve crosses the system’s own resistance curve.

What a Fan Is Selected On

Industrial ventilation practice names four inputs for fan selection: the volume to be moved, the fan static pressure, the type and concentration of the contaminant, and the importance of noise as a limiting factor. The contaminant enters because it drives the fan’s materials of construction, and a corrosive or wet stream is not handled by an ordinary steel fan. Noise enters as a weight rather than a number, because a fan at the quiet end of a range may be a different machine from one at the loud end.

What is absent from that list is instructive. The duct diameter is not a fan selection input. It reaches the fan only through the velocity it produces and the static pressure that velocity implies, which is another way of saying that the diameter matters to the extent that it changes the budget — which, as the arithmetic below shows, is less than most people expect.

The System Curve, With the Scrubber in It

For a run of duct and fittings, the resistance rises with the square of the flow, because velocity pressure does. Double the flow and the duct alone demands four times the pressure. That is the familiar steep system curve, and it is what makes a duct-only system behave predictably.

Put a scrubber in the train and the composite curve flattens. The duct and fittings still rise with the square of flow, but the scrubber’s own loss is closer to constant over a modest flow range, because it is set by the packing and the liquid loading rather than by the gas velocity alone. When a large constant is added to a term that grows as the square, the sum grows much more slowly. The system becomes one where a large change in pressure produces a comparatively small change in flow.

Where the Fan Actually Runs

The operating point sits where the fan curve crosses the composite system curve. Two consequences follow from the flatness of that curve, and both catch people out at commissioning.

The first is that a scrubber system holds its flow more tenaciously than a duct-only system. A 10% error in the assumed scrubber pressure drop moves the operating point along the fan curve by far less than 10% of the flow, because the curve is being crossed by a shallow line. What looks like a significant budget error turns out to be a modest flow error.

The second is the opposite of reassuring. Because the operating point moves so little when the pressure changes, the only way the duty point ends up badly wrong is if the flow itself changes, and the flow is the number that other people control.

Why the Operating Point Drifts Right

Several ordinary events move the operating point along the fan curve, and they do not all move it the same way. Filters loading, a damper being closed or the bed of a scrubber fouling all add resistance, which moves the operating point toward lower flow. A branch being added, a hood being opened up or a damper being removed all reduce the system’s resistance, which moves the operating point toward higher flow.

Higher flow is the direction that costs money and causes trouble. The fan draws more power, since power is the product of flow and pressure, the velocity in the duct rises above the value it was sized for, and on a particulate stream that means accelerated wear. Because the motor was selected against a design duty rather than a maximum one, there may be little margin for the drift, which is a reason to check the fan’s curve at the right-hand end and not only at the design point.

Air Volume Is Not Constant

Everything above assumes the volume entering the duct is the volume leaving it. In a hot or wet exhaust that assumption is wrong, and the correction can be larger than the difference between two stocked sizes. It is also the reason a system can need two different duct sizes on either side of the same treatment device.

Volume Changes With Temperature

For a fixed mass flow, the volume of a gas is proportional to its absolute temperature. Working in kelvin rather than celsius, a stream at 60 °C is at 333 K and the same stream at 40 °C is at 313 K, so cooling it from 60 to 40 shrinks its volume by 6%. Cooling a stream from 150 °C to 40 °C shrinks it by 26%. None of these are subtle effects, and none of them appear in a duct chart drawn at ambient conditions.

The correction applies to sizing directly, because both the velocity and the diameter are volume quantities. A duct carrying 10,000 m³/h at 150 °C that is cooled to 40 °C before the fan needs to pass a substantially smaller volume after the cooling, and if that whole volume has to be sized at the same velocity, the downstream diameter is smaller.

What a Wet Scrubber Does to the Volume

A wet scrubber is a heat exchanger as well as a separator. The gas leaves a packed bed close to the adiabatic saturation temperature, which for a hot process stream is usually far below its inlet temperature. The volume at the outlet is therefore smaller than the volume at the inlet for the same mass flow.

For a near-ambient pickle fume at around 20 to 30 °C, the effect is a few percent and can be ignored for sizing. For a hot process — a dryer, an oven, a kiln off-gas — the effect is large enough to change the answer by a whole size. A useful working approximation on such a system is that the duct between the scrubber and the fan can be one stocked size smaller than the duct upstream of the scrubber, for the same mass flow and the same velocity target. That is worth checking rather than assuming in either direction, because it is also the duct that sits under the greatest vacuum.

Density, Friction and the Cold-Air Chart

Temperature changes density as well as volume, and the friction loss depends on density through the velocity pressure. Hot gas is less dense, so at the same volumetric flow and the same velocity, a hot duct loses less pressure per metre than the same duct at ambient conditions. At 20 °C air sits at about 1.20 kg/m³; at 90 °C, near the ceiling for polypropylene duct in continuous service, it is about 0.97 kg/m³ — a fifth lighter, and therefore about a fifth less friction at the same volume.

The temperature limits themselves, and what continuous service means for polypropylene grades, are covered in the guide to polypropylene duct. For sizing, the practical point is that a hot PP duct is cheaper in pressure than a steel chart drawn at ambient conditions suggests, and the two reasons behind that — the smoother surface and the thinner air — push in the same direction rather than offsetting one another. Neither is large compared with the scrubber, and neither is worth ignoring when the duct is the term under discussion.

The Volume the Fan Must Be Specified At

The single most common error in this area is specifying the fan at the hood rather than at the fan. The fan sees the volume at its own inlet, at the temperature and pressure and moisture content that arrive there. On a hot stream with a wet scrubber that is a different number from the hood’s, and the fan’s performance curve is drawn for the conditions at the fan.

Specifying a fan on the hood volume overstates the volume it has to move, which pushes the selection toward a larger machine, a higher duty and a larger motor. The error is safe in the sense that the fan will work, and expensive in the sense that the site pays for capacity it does not need for the life of the installation.

Make-Up Air Is a Sizing Input

The system does not end at the stack. Every cubic metre an exhaust system removes has to be replaced, and where it is not replaced deliberately it comes in through whatever opening is available. That mechanism is part of the duct sizing problem, even though it belongs to the building rather than to the ductwork.

A Room Under Negative Pressure Moves Less Air

If the make-up air path is inadequate, the exhaust fan pulls the room into a slight negative pressure relative to outside. Once that happens, the fan is no longer working against the duct system alone; it is also lifting air against the pressure difference across the building envelope. The operating point slides back along the fan curve, and the system moves less air than it was sized for.

Industrial ventilation guidance is explicit on the mechanism: without replacement air, a slight negative pressure is created in the room and the airflow through the exhaust system is reduced. The reduction is not a design abstraction. It is measured, it appears as reduced capture at the hood, and it is one of the standard findings when an installed system under-performs against its design figures.

How a Shortfall Eats Your Velocity Margin

This is where make-up air connects to the sizing arithmetic. The velocity in the duct is proportional to the volumetric flow, so a shortfall in flow is a proportionate shortfall in velocity. If the make-up path delivers 90% of what the exhaust removes, the duct velocity is about 90% of its design value.

A run sized with a comfortable margin — say 12 m/s against a 10 m/s floor — survives that. A run that was already close to the floor does not, and the failure appears as material settling in the horizontal sections rather than as anything obviously to do with the building. Where a system is designed tight against the floor for good reason, the make-up air is not an optional comfort provision; it is part of the containment design.

Sizing the Replacement Path

The replacement air needs a path with enough free area to pass the exhaust volume at an acceptable speed. Velocities through doorways, louvres and wall openings are limited by comfort and by the force they exert on doors, so a large exhaust volume implies a genuinely large free area. A dedicated make-up air unit with a fan and a tempering coil is the controlled answer, and it is normally interlocked with the exhaust so that neither can run alone.

Two design habits improve the outcome. Admitting make-up air close to the hood it serves — rather than across the room — reduces cross-draughts that disturb the capture, and tempering it avoids the draught complaints that otherwise lead to the make-up system being shut down in service.

What to Ask the Building Services Engineer

Three questions belong in the sizing enquiry once the exhaust duty is known. What is the free area of the make-up path at the design exhaust volume? Is the exhaust interlocked with the make-up supply, or can one run without the other? And does the make-up air arrive near the hoods, or across the room? The answers change the margin the duct sizing can safely carry, and none of them can be answered from the duct drawing.

Sizing a Run Is Not Sizing a System

Every method on this page sizes a single run. A system with several branches needs something different, and the difference is worth understanding before a single-run answer is applied to a multi-branch layout. Single-run pp duct sizing is what most of the preceding sections have been doing, and it is the right tool for a great deal of installation work; it is not the right tool for every layout.

The Equal Tee and the Starved Branch

Where two branches leave a main at the same point through a tee of equal diameter, the branch nearer the fan sees a different pressure than the one further away, and the flow divides according to the resistances rather than according to the design airflows. The branch with the shorter run and fewer fittings takes more than its design share, and the branch with the longer run takes less. Sizing each branch individually for its own design airflow assumes a division that the layout does not produce.

The effect is worst at the end of a long main, which is where the least-favoured branch normally is. That is the branch that shows a capture failure first, and it is why branch take-offs deserve more attention in layout than their individual pressure losses suggest.

Two Methods for Multi-Branch Mains

Professional practice uses one of two methods rather than repeated single-run sizing. The equal-friction method sizes every section to the same pressure loss per unit length, which tends to produce roughly uniform velocity and noise across the system at the cost of more size changes. The static-regain method deliberately oversizes downstream sections so that regained velocity pressure offsets the friction loss, which suits systems where branches must balance without dampers.

Both need the same transport-velocity floor to be checked on every branch, since neither method has any knowledge of what the stream carries. And both produce a design intent that still requires verification, because the assumption of clean duct and nominal fan performance does not survive contact with an installed system.

Balancing Dampers Are a Sizing Decision

Where branches cannot be balanced by geometry, dampers do the work at commissioning. A fixed orifice or a manual damper sized for the design pressure drop balances the branch by construction and needs no later adjustment, at the cost of a permanent pressure loss. An adjustable damper gives the commissioning engineer something to work with, at the cost of a device that can be moved by anyone who finds it.

Whichever is chosen, it is a sizing decision rather than an accessory decision, because the pressure it drops has to be in the budget. A design that balances with dampers needs the fan to supply the pressure those dampers will dissipate, and a design that balances geometrically does not.

Where the Sizing Has to Hand Over

Single-run sizing covers a great deal of real industrial work — one hood, one duct, one fan. It stops being sufficient when a system has several branches that must deliver their design airflows simultaneously, when the static pressure is high enough that the vacuum on the duct becomes a structural matter, when the stream is explosive or highly toxic, or when failure means an unsafe workplace. Those are the cases for a specialist designer, and the value of the calculation here is that it produces the inputs such a designer needs rather than replacing their work.

Worked Example: 4,600 m³/h Through a Scrubbed Pickling Line

A worked case ties the sections together. The duty is a pickling line exhausting 4,600 m³/h of acid mist through a 25 m duct run into a packed-bed scrubber, with the fan downstream of the scrubber. The site wind design speed is 5 m/s and the gas is at ambient conditions, so the volume correction does not apply here.

Steps 1 and 2: Flow, Band, Theoretical Diameter

Acid mist needs a transport velocity high enough to carry entrained droplets. The band used is 10 to 12 m/s, and the design velocity is taken at the top of the band because the loop will move it downward when a stocked size is chosen. The flow is 1.278 m³/s, the required area is 0.1065 m², and the theoretical bore is 368.2 mm.

Compared bore to bore, φ355 at 346.6 mm is smaller than required and φ400 at 391.0 mm is larger. The theoretical figure sits between the two stocked sizes, which is normal.

Steps 3 to 5: Choosing φ400 and Re-reading the Velocity

In φ355 the velocity would be 13.54 m/s, above the 12 m/s ceiling for a mist-laden stream. In φ400 it is 10.64 m/s, inside the band and close to the design velocity. The rounding goes up, and here it is the right direction, because the larger size still leaves the velocity above the floor.

The loop closes with one pass: 10.64 m/s is within 11% of the design velocity, the friction rate at that velocity is about 2.5 Pa/m, and the resulting budget does not change the flow enough to move the diameter. Nothing in the calculation needs repeating, which is the normal outcome and not an accident.

The velocity pressure at 10.64 m/s in ambient air is 67.9 Pa. That single number drives three of the six budget terms below, which is why it is worth computing explicitly rather than keeping the velocity as a bare figure.

The Budget: 1,164 Pa From Six Terms

The six terms are evaluated along the longest path, with a loss factor of 0.5 at the hood entry, three long-radius bends and one branch tee giving a combined loss coefficient of 1.05, and a packed-bed scrubber at 900 Pa.

Three observations follow from the column of shares, and they are the reason this section exists.

First, the scrubber is 77% of the system. Its figure comes from the equipment supplier, and a 10% error in that figure moves the total budget by more than the entire straight-run friction of the duct.

Second, the whole duct — straight run, fittings and hood entry together — is 14.6% of the budget. The diameter decision controls part of that 14.6%, and the part it controls is the 5.6% that the straight run accounts for.

Third, the stack accounts for 8% and is usually designed last. A stack outlet that is oversized to reduce the exit loss is also a stack whose exit velocity may fall below the 7 m/s that the site wind speed demands, so the two constraints on stack sizing genuinely pull against each other.

The Fan and the Shaft Power

The fan has to deliver 1.278 m³/s against 1,164 Pa. At an overall fan efficiency of 60% the shaft power is 2.48 kW, which is the product of the flow and the pressure divided by the efficiency. A reasonable motor selection sits above that figure with margin for the operating point drifting right and for the fan not achieving its nominal efficiency in service.

The power figure is also the number to carry into any comparison, because it is the one that accumulates. At 8,760 hours a year and an electricity price of 0.10 per kilowatt hour, 2.48 kW is about 2,170 per year, which is a large enough annual figure to be worth optimising and a small enough one to be dwarfed by the capital cost of getting the scrubber wrong.

What the Next Size Up Would Have Cost

The obvious optimisation is to try φ500. At 4,600 m³/h in a φ500 duct the velocity falls to 6.80 m/s and the velocity pressure to 27.8 Pa. The straight-run friction drops to 0.88 Pa/m and the duct terms fall sharply: the hood entry to 13.9 Pa, the fittings to 29.1 Pa over the same 25 m, the stack friction to 8.8 Pa and the stack exit to 27.8 Pa. The scrubber, unchanged, remains at 900 Pa.

The new total is 1,001 Pa rather than 1,164, a reduction of 14%. At the same flow that is about 2.13 kW of shaft power instead of 2.48 — a saving of 0.35 kW, or roughly 3,070 kilowatt hours a year. And the entire saving is unavailable, because 6.80 m/s is below the transport floor and the acid mist will settle in the horizontal run.

The exercise is worth running anyway, because it shows the size of the prize. The duct-side terms fall by about 62% when the diameter grows by 25%, and the system’s total pressure falls by 14%, because the duct was never the term that mattered. The same φ500 duct also weighs 7.78 kg/m against 5.09 kg/m for the φ400, so the capital cost of chasing that 14% moves in the wrong direction at the same time as the velocity moves out of band.

The Same Duty at 60 °C

Now suppose the same pickling line runs hotter, with the gas arriving at the hood at 60 °C and leaving the scrubber saturated at 40 °C. The mass flow is unchanged, and the standard volume is unchanged, but the actual volume at the hood is now larger than the figure above by the ratio of absolute temperatures, 333/293, or 14%.

The design flow at the hood therefore becomes about 5,230 m³/h of actual gas, and the theoretical bore rises from 368 mm to 393 mm. That is a more uncomfortable answer than it first looks. The φ400 still serves the hood end, but at 12.1 m/s it now sits hard against the ceiling of the 10 to 12 m/s band instead of in the middle of it, and the next size up — the made-to-order φ450 — would push the velocity back under the 10 m/s floor. The hot case turns the squeeze described earlier into an everyday one, and the ways out of it are the same.

Downstream of the scrubber the position is easier. The gas leaves the bed close to saturation at around 40 °C, so the volume reaching the fan is about 4,910 m³/h — some 6% below the hood volume rather than 14% above it — and the ductwork between the scrubber and the fan is comfortably served by the same φ400.

The fan, meanwhile, must be specified on the volume and density at the fan, which is the 40 °C figure and not the 60 °C one. A fan specified on the hood volume would be selected about 6% larger than the duty requires, and would run for the life of the plant away from its best efficiency point.

What to Put on a PP Duct Sizing Enquiry

The sizing work produces a short list of numbers, and sending all of them is what turns a request for a price into a checkable design. A supplier receiving the list can verify the arithmetic and flag a problem before a quotation is issued rather than after the duct is fabricated.

The Sizing Numbers

Five numbers belong in every enquiry. The design airflow, as a volume rather than a mass, with the temperature it applies at. The target velocity and the band it comes from, so the supplier can see whether the floor was respected. The selected diameter, expressed as a nominal size with its wall thickness. The total developed length of the run. And the number of bends, tees and transitions, because the fittings determine a large share of the pressure.

Two of those deserve emphasis. Stating the band rather than only the velocity lets a supplier propose a different size without guessing at the constraint, and it is the fastest way to find out that a site’s flow sits in a ladder gap. Stating the wall thickness alongside the diameter prevents the quotation from being built on a nominal bore that the finished duct will not have.

The Process Conditions

The stream decides the material before the diameter decides anything else. The chemistry, the concentration and whether the gas is wet or dry determine whether polypropylene is the right choice at all; the comparison of PP against the other common duct materials is set out in PP duct versus PVC duct for the solvent and chlorinated cases and in PP duct versus FRP duct for hot and fluoride-bearing streams.

Alongside the chemistry, three physical conditions belong on the enquiry. The continuous and peak temperature, because they set the material grade as well as the volume correction. The pressure class, meaning whether the run is under suction and how much, because the wall that resists a vacuum is a different selection from the wall that carries a size. And the contaminant’s own character, because whether it is a fume, a mist or a dust is what sets the velocity floor the design was built on.

The Questions to Ask Back

Three questions are worth asking a supplier who quotes a sized run. Which velocity does the quoted diameter produce at the stated flow? What wall thickness does the quoted size come with, and what negative pressure is that wall rated for? And what lead time applies if the design settles on a diameter outside the standard production series?

The first question catches a quotation built on nominal diameters rather than bores. The second catches a price built on the lightest wall in the range for a duty that needs a heavier one. The third is the practical cost of the ladder gap discussed earlier, and it is better known at enquiry stage than after the layout has been frozen.

Frequently Asked Questions About PP Duct Sizing

What size PP duct do I need for 5,000 m³/h?

It depends on the velocity the stream requires, and that is the whole answer rather than an evasion. At the 12 m/s appropriate to a wet mist, 5,000 m³/h needs a bore of 384 mm, which falls between the φ355 and φ400 production sizes; φ400 gives 11.57 m/s and φ355 gives 14.72 m/s. At the 8 m/s that a clean, non-condensing fume can tolerate, the same flow needs only a 470 mm bore, and φ500 carries it at 7.39 m/s. The flow alone does not determine the size, which is why a pp duct sizing enquiry that contains only a volume cannot be answered properly.

What velocity should a PP exhaust duct run at?

Between about 6 and 18 m/s, with the working band set by what the stream carries rather than by the duct material. Clean fume in a noise-sensitive location runs at 6 to 8 m/s. General chemical fume and laboratory exhaust runs at 8 to 10 m/s. Plating and pickling mist, which carries droplets that must not settle, runs at 10 to 12 m/s. Light dust and condensable vapour run at 15 to 18 m/s, above the settling velocity of the particles being carried. Above the top of the range the friction cost climbs with the square of velocity and a particulate stream begins to erode bends.

Should I round the duct size up or down?

Round toward the size that keeps the velocity inside the band, which is not the same as rounding up. On a comfort system a larger duct is always the safer choice, because the only costs are material and space. On an exhaust duct a larger duct lowers the velocity, and below the transport floor the contaminant settles out. In a 5,800 m³/h pickling-mist duty the theoretical bore falls between φ400 at 13.42 m/s and φ500 at 8.58 m/s, so the round-up lands below the floor and the round-down lands above the ceiling. When both neighbours fail, the answer is a made-to-order diameter, a revised airflow or a split run.

Why does my figure differ from an online duct sizing calculator?

Three reasons, and they are all worth checking. Most calculators are drawn for galvanized steel, whose roughness gives a friction factor about 14 to 19% higher than polypropylene at the same bore and velocity, so they overstate the pressure a PP duct loses. Most are drawn at ambient conditions, so they overstate the volume a hot stream presents and understate the volume after a wet scrubber. And most assume the nominal diameter is the flowing diameter, which ignores the wall and overstates the capacity of every size, the small ones most of all.

Can I size an exhaust duct from airflow alone?

No, for two separate reasons. The first is that the airflow determines a diameter only once a velocity is chosen, and for an exhaust duct the velocity is a constraint rather than a preference: too slow and the stream settles, too fast and the duct erodes. The second is that the diameter is not the whole of the sizing problem. The same 4,600 m³/h can be carried in φ400 with a modest fan, or in a duct that is correctly sized and still fails because the vacuum class of the wall is wrong or the make-up air path cannot supply the volume being removed.

How do I calculate the static pressure of a duct system?

Add six terms along the longest path: the hood entry loss, the straight-run friction over the full developed length, the fittings expressed as loss coefficients times the velocity pressure or as equivalent lengths, the treatment device’s pressure drop, the stack friction, and the stack exit loss. The third term is where most hand calculations go wrong, because fittings on a short run can outweigh the straight duct they sit in, and the fourth term is usually the largest of the six. In the worked example on this page the six terms come to 1,164 Pa, of which the scrubber alone is 900 Pa.

How much does the duct diameter change the fan power?

Less than the friction arithmetic suggests, on any system with a treatment device in it. Stepping the worked example from φ400 to φ500 cuts the duct-side pressure terms by about 62% and the total system pressure by 14%, which is 0.35 kW out of 2.48 kW. The reason is the distribution of the budget: the duct contributes about 15% of the total pressure and the scrubber about 77%, so a large proportionate improvement in the small term produces a small improvement in the whole. The saving is also unavailable here, because the larger duct drops the velocity to 6.80 m/s, below the transport floor for an acid mist.

Get a PP Duct Sized Against Your Actual Duty

If you have a flow, a contaminant and a route, the sizing can be worked through against sizes that are actually built rather than against a chart drawn for another material. Bring the airflow with its temperature, the velocity band or the reason for it, the developed length and the fitting count, and the answer comes back as a nominal size, a wall thickness and the velocity it produces.

The standard round duct that most industrial exhaust work lands on, from φ110 to φ500 in the production series, is listed with its bores and wall thicknesses in the industrial ductwork range. The two sizes most commonly specified on a sub-main of this kind, the φ315 duct and the φ355 duct, show how the nominal size, the wall and the connecting flanges are specified together. Send the duty figures through the contact page and the velocity and the wall can be checked against your run before anything is cut.