Compressed Air Pipe Networks & Boiler Steam Piping Design

A garment factory depends on two pipe networks that most managers never see, and they obey opposite logics. Compressed air works best as a ring: a loop around the floor so that every tool can draw from two directions. Steam works as a one-way system: steam flows from the boiler to the iron or press, condenses as it gives up heat, and the hot condensate should return to the boiler to be used again. Designing both networks well comes down to the same discipline: calculate demand honestly, design for the worst point on the floor, and maintain the small components that waste the most energy.

This lesson works through one demand sheet for both networks, so the method can be reused on any floor. Where a value depends on a code, a tool rating, or a manufacturer specification, the lesson says so: verify against the current code and manufacturer data before building anything. Pipe diameters, boiler safety fittings, and pressure relief arrangements are installation design work for qualified engineers, and this lesson gives the demand logic that feeds that design, not a substitute for it.

Start with an inventory, the same discipline as the electrical load list. List every air user on the floor with its rated consumption in cubic metres per minute. The worked example floor has 30 air tools at 0.12 cubic metres per minute each, which is 3.60; 6 auto-trimmer stations at 0.25 each, which is 1.50; and cleaning points at 0.80. Total installed consumption is 3.60 plus 1.50 plus 0.80, which is 5.90 cubic metres per minute.

Installed consumption is not demand, because not every tool runs at once. The simultaneity factor converts one to the other: it is the fraction of tools running simultaneously at peak. At a factor of 0.70, demand is 5.90 multiplied by 0.70, which is 4.13 cubic metres per minute. The factor should reflect the real floor: a floor where operators share tools and work in sequence justifies a lower factor than a floor where every station runs continuously. As with the electrical demand factor, a clamp-style measurement beats a copied number: log tool use during a peak hour and see what fraction truly runs together.

Then add a leakage allowance. No real network is leak-free, and a system sized with zero allowance is undersized the day the first joint starts hissing. Ten percent is a common planning allowance: 4.13 multiplied by 1.10 is 4.54 cubic metres per minute. Select the next compressor size above that, which here is a compressor rated 5.0 cubic metres per minute at 7 bar. The rating must be checked at the working pressure the tools need, because a compressor rated at a lower pressure delivers less usable air at the tool.

The most common air complaint on a sewing floor is weak tools at the far end, and the most common wrong fix is raising compressor pressure. Raising pressure increases energy use across the whole floor, increases leakage at every joint, and still leaves the far end weakest, because pressure falls along the length of a dead-end line in proportion to distance and simultaneous flow. The layout is the problem, not the pressure setting.

A dead-end line feeds every tool from one direction. During simultaneous use, the tools nearest the compressor take their share first, and the pressure available at the far end is whatever remains after friction loss along the whole run. A ring main changes the geometry: the loop connects back to the compressor or receiver at both ends, so every drop point can draw from two directions. The effective distance to the farthest point roughly halves, and the pressure at that point rises without any increase in compressor setting. Receivers placed near high-demand zones add a second benefit: they store air locally for peak bursts, so a row of trimmers starting together does not starve the rest of the floor.

Pipe velocity and pressure drop guides exist for sizing the main, and they are worth knowing as planning ranges: air mains are commonly designed for velocities around 6 to 10 metres per second, with total pressure drop from compressor to farthest tool kept small, often around 0.1 bar for the main run. Treat those as planning guides and verify against manufacturer data and the current code for the actual installation, because the correct diameter depends on flow, length, fittings, and acceptable drop together.

Steam demand is simpler than air demand because steam users run more predictably: at peak finishing, every iron and press is in use. Sum the consumption of every steam user at peak simultaneous use, add a safety margin, and select the boiler above the result.

The worked example finishing section has 12 steam irons at 8 kilograms per hour each, which is 96 kilograms per hour, and 4 steam presses at 25 kilograms per hour each, which is 100 kilograms per hour. Total steam demand is 196 kilograms per hour. A 20 percent margin gives 196 multiplied by 1.20, which is 235.2 kilograms per hour, and the next boiler size above that is 250 kilograms per hour. Sizing from average use instead of peak simultaneous use is the classic steam error: the boiler copes all morning and starves every iron at the busiest hour, when finishing is already the bottleneck before packing.

Steam pipe velocity guides are higher than air, commonly around 15 to 30 metres per second for distribution mains, because steam is a gas at low density and high volume. Again, treat that range as a planning guide and verify against manufacturer data and the current code. What matters most in the layout is not only velocity but drainage: steam lines must fall toward drain points so condensate does not pool, water hammer does not occur, and irons receive dry steam instead of wet steam that spots garments.

A steam trap is a small valve with an outsized effect on fuel cost. It sits at equipment outlets and low points, lets condensate out, and holds live steam back. When a trap fails open, it passes live steam straight to the condensate line, continuously, silently, and invisibly, because the steam disappears into the return system instead of escaping as a visible leak. A factory can lose a significant share of its boiler fuel through a handful of failed traps and never see a symptom except a rising fuel bill.

Testing traps is therefore a routine, not an occasional task. Test at least quarterly, and monthly on high-use finishing lines, using temperature measurement or an ultrasonic tester: a trap passing live steam runs hot on both sides and hisses at ultrasonic frequencies. Log the results trap by trap, so a trap that fails repeatedly gets replaced with a different type or gets its installation checked, instead of being replaced with the same trap that will fail again.

Condensate return closes the loop. Condensate is hot water the boiler already paid fuel to heat, and returning it to the feed tank saves the fuel needed to heat cold make-up water from scratch. A well-maintained trap system commonly returns 70 to 80 percent of condensate; treat that range as a planning guide and verify against manufacturer data for the installed traps and boiler. The return system needs its own care: return lines corrode if condensate is acidic, pumps must handle hot water without cavitation, and the feed tank needs level control that copes with returning surges. When return stops working, fuel use rises and nobody can say why, which is why return flow belongs on the monthly energy review.

Insulation is the third fuel saver. Every metre of uninsulated steam line loses heat along its whole route, increases condensate before steam reaches the iron, wastes fuel, and creates a burn hazard for anyone brushing past it. Insulate every steam line, inspect the insulation during the trap routine, and repair damaged sections promptly, because insulation damaged by a trolley collision stays damaged until someone is assigned to fix it.

A knit factory in Gazipur ran its sewing floor air from a single dead-end line. Tools near the compressor worked well, but tools at the far end lost power during peak hours, auto-trimmers stalled mid-cycle, and operators blamed the machines. Maintenance raised compressor pressure to compensate, which increased energy use across the whole floor without fixing the far end, because the pressure drop along the line grew with the flow.

The engineering team converted the dead-end line into a ring main, so every tool could draw from two directions, and added a second receiver tank near the far end. Far-end pressure during peak hours rose from 5.2 bar to 6.5 bar on the same compressor, and the pressure setting was returned to normal. Trimmer stalls stopped, energy use fell, and far-end operators stopped keeping a private stock of hand tools for the hours when air was weak. The ring main cost less than one month of the overtime the stalls had been causing. The moral is the same as the electrical lesson: fix the layout and the sizing, not the pressure setting.

Both networks degrade gradually, which makes routine the only defence. For air: run a monthly leak audit during a quiet hour with an ultrasonic detector, repair every leak found, drain receivers and low-point drains on schedule, and change intake filters so the compressor does not work harder for the same air. For steam: test traps quarterly or monthly, check insulation, verify condensate return flow, and log boiler fuel against production so a rising fuel-per-piece trend triggers investigation before it becomes a crisis.

Log everything against energy use. Compressor electricity per thousand pieces and boiler fuel per thousand pieces are the two numbers that reveal network health over time. When either rises without a production change, the cause is almost always in the network: a new leak, a failed trap, a clogged filter, or a lost insulation section. The log turns a vague sense that energy is rising into a dated, assignable fault.

Inventory every air tool and steam user with rated consumption. Calculate air demand with simultaneity and leakage allowance, and steam demand at peak simultaneous use with margin. Select the compressor and boiler at the next sizes above the calculated demands. Draw the air ring main with receivers near high-demand zones, and the steam route with fall toward drains, traps at low points, condensate return, and insulation throughout. Set the monthly leak audit, the trap testing routine, and the energy-per-piece log. Verify pipe sizing, boiler capacity, and safety fittings against the current code and manufacturer data before building, because demand logic tells you how much, and qualified installation design tells you how safely.