Why Pipe Size Decides More Than Compressor Size

Why this matters

A compressor is sold on flow. The number on the nameplate is cubic feet per minute, and that is the number the customer remembers paying for. But nobody in the plant ever runs out of cubic feet. They run out of pressure at a tool, and pressure at the tool is set by the pipe between it and the receiver, not by the machine. A shop that answers a pressure complaint with a bigger compressor has bought more air and none of the thing it was short of, and it now pays for the mistake on every electric bill for the life of the system.

Flow is what you buy, pressure is what you pay for

Two separate quantities, and they get confused constantly.

Flow is how much air the machine can make. If total demand exceeds it, header pressure falls everywhere at once and stays down until demand drops. That is a capacity problem and a bigger machine genuinely fixes it.

Pressure at a point of use is header pressure minus everything the air lost getting there. That loss is friction, and friction is local. It can be severe at one drop and near zero at the next. A bigger compressor cannot fix it, because the loss is subtracted after the compressor has done its work.

The reason this matters commercially is that the compressor's discharge setpoint is always chosen to cover the worst point of use. Whatever the largest loss in the plant is, the machine carries it as extra discharge pressure, continuously, whether or not that tool is running. Compression work rises with the logarithm of the pressure ratio rather than with pressure itself, so the cost of one more psi depends on where you already are: at a discharge of about 100 psig from an atmospheric intake, one additional psi is roughly four tenths to half of one percent of input power. That coefficient is derived at about 100 psig and does not travel far. At a lower discharge the same psi costs a larger percentage, because the same increment is being divided by a smaller absolute pressure; at a higher discharge it costs less.

The scaling that makes bore the dominant term

For a round pipe carrying fully turbulent flow with the friction factor treated as roughly constant across the range in question, friction loss scales with the square of volumetric flow and inversely with about the fifth power of the inside diameter. The fifth power comes from two places at once: velocity is flow divided by area, so it goes as one over diameter squared, and the loss goes as velocity squared, which is one over diameter to the fourth; the remaining power comes from the pipe wall area the flow is rubbing against.

Two consequences follow directly, and both are worth carrying in your head:

• Halve the flow and you quarter the loss. So a pressure drop measured at idle tells you almost nothing. Loss must be measured during the demand peak, because that is the condition the setpoint has to cover.
• A bore 25 percent larger cuts the loss to about a third, because 1.25 raised to the fifth power is about 3.05. Nominal pipe sizes do not step in neat 25 percent increments of actual bore, so read the real inside diameters off the schedule rather than assuming the nominal jump gave you what you wanted.

This is a scaling estimate, not a design calculation. The friction factor drifts with Reynolds number and with wall roughness, and a badly corroded steel main has an effective bore smaller than its schedule says. Use the scaling to decide whether a size change is worth investigating and hand the actual sizing to the pipe manufacturer's tables or a mechanical engineer.

What a bigger main does not buy

This is the part that gets skipped, and skipping it is how a shop spends a week repiping and fixes nothing.

It does not reduce leaks. A leak is an opening venting to atmosphere. Above roughly 13 psig any such opening is choked, meaning air leaves the throat at the speed of sound and the mass flowing out is proportional to the absolute upstream pressure. Pipe bore is not in that relationship anywhere. Bigger pipe changes leak flow only through whatever setpoint reduction it allows.

It does not add capacity. If the machine cannot make enough air, larger pipe delivers the same shortfall with less friction on it.

It does not add ride-through. Larger pipe does add some storage volume, but a header sized to hold up a two-minute demand event is a receiver with a very expensive shape. Ride-through is storage's job, and it is sized from volume and usable pressure band.

It does not fix a regulator set wrong, a plugged filter element, or a dryer that is not drying. Each of those is a loss or a fault at a specific component, and each has its own measurement.

It does not fix the last ten feet. This is the exclusion that catches the most shops. The Q-squared-over-D-to-the-fifth scaling bites hardest where the bore is smallest, and the smallest bore in most systems is a quick-coupler, a hose, or a drop nipple, not the main. A plant with a generous header and a small coupler at the tool has spent its money at the wrong end of the run.

Where the loss actually concentrates

Walk a run and the loss sits in a predictable order. Fittings and valves behave like extra length of pipe: a globe valve, a sharp elbow, or a tee taken on the branch each add an equivalent length that can dwarf the straight run they sit in. Filters and dryers carry a rated drop clean and a much larger one loaded, and the loaded figure is the one your setpoint has to cover, so read the element's change-out differential from the manufacturer's data rather than the clean rating.

Then the drop itself: the vertical branch off the header, its shutoff, its filter-regulator-lubricator unit, its coupler, and the hose. Every one of those has a bore smaller than the main, and the scaling punishes each of them harder than it punishes the header.

Worked example: fifteen psi you are paying for continuously

Before any of what follows, the measurement itself: to fit a test gauge into a live line, isolate that section at its upstream valve, open a bleed or vent and confirm the gauge reads zero before breaking the joint, and keep the isolation under lock and tag per 29 CFR 1910.147, which governs stored mechanical energy of this kind. A compressed-air line that is opened while still pressurized ejects the fitting and whips the hose, and that is a struck-by injury from a system everyone treats as harmless. Do not clear the line or the fitting with compressed air; 29 CFR 1910.242(b) permits compressed air for cleaning only where it is reduced to less than 30 psi and then only with effective chip guarding and personal protective equipment.

A fabrication shop complains that its plasma table starves in the afternoon. Both readings are taken with the same gauge, moved from one port to the other, during the peak: the loss you want is a difference, and a bourdon gauge's dominant error at a given point on the dial is a fixed offset, which cancels in a difference taken with one instrument. Read the two ends with two different gauges and those offsets do not cancel; if all you know about each is a worst-case bound, the bounds add linearly and you have doubled your uncertainty on the only number you came for.

• At the receiver outlet, under peak load: 102 psig
• At the far drop feeding the table, same load: 87 psig
• Loss through the distribution: 15 psi

The table needs 90 psig at its inlet. To deliver that through a 15 psi loss the machine has to hold 105 psig. If the loss were 5 psi instead, 95 psig would do.

Now price the difference in the two currencies it is paid in.

Compression side. Ten psi off the setpoint, at roughly four tenths of a percent of input power per psi in the neighbourhood of 100 psig, is about 4 percent of input power for the same air made.

Leak side. Leak mass flow is proportional to absolute upstream pressure under choked flow. Dropping from 105 psig to 95 psig takes the absolute from 119.7 psia to 109.7 psia, a ratio of 0.916, so every leak and every unregulated opening in the plant passes about 8 percent less air. In an unmanaged system leak load commonly runs a fifth to a third of compressor capacity, so that 8 percent is being applied to a large term.

What it takes to get there. Cutting 15 psi to 5 psi is a factor of three, and a factor of three under the fifth-power scaling needs a bore about 25 percent larger, because 1.25 to the fifth is 3.05. Length and fittings unchanged, same flow, same turbulent regime. Whether the next schedule size up actually gives you 25 percent more bore is a question for the pipe table, and the answer is often more or less than you wanted.

The failure mode. The shop that skips the measurement and reasons from the complaint installs a larger compressor. Header pressure at the receiver was already 102 psig, so it was never short of air; the new machine holds the same 102 and the table still sees 87. The loss is untouched, the setpoint is untouched, and the plant now runs a larger machine at a lower load fraction, which for most control strategies is the expensive end of its curve.

How to verify you got this right

Do not accept the repipe on the drawing. Re-run the same two-port measurement with the same gauge at the same peak condition and confirm the loss actually fell to what the scaling predicted. If it did not, the loss was never in the length you replaced, and the next place to look is the smallest bore in the path: the coupler, the hose, and the filter element's loaded differential rather than its clean rating.

Then close the loop on the setpoint. A repipe that reduces the loss and leaves the discharge setpoint where it was has captured nothing at all. The saving is realised only when the setpoint comes down, and it comes down only to the level the new worst point of use still tolerates, which is a fresh measurement and not an assumption.

References

• 29 CFR 1910.147, control of hazardous energy, for isolating and verifying the depressurisation of a compressed-air line before any joint is broken
• 29 CFR 1910.242(b), which permits compressed air for cleaning only when reduced to less than 30 psi and used with effective chip guarding and personal protective equipment
• Pipe manufacturer's published inside diameters and equivalent-length data for fittings and valves; filter and dryer manufacturer's loaded pressure-drop data
• See related: What a Loop Header Does That a Dead-End Main Cannot; Why Raising System Pressure to Fix a Symptom Costs More Than the Symptom; How Pressure and Flow Relate in a Real System