Dust Collection Ductwork Design for Stone Fabrication Shops

Most shops buy a dust collector the way they buy a compressor. They look at horsepower, look at price, and put the unit in a corner. Six months later the polishing station is still hazy, the drums fill slower than expected, and someone decides the collector is undersized. Nine times out of ten the collector is fine. The problem lives in the pipe between the machine and the fan — the hoods, branches, main trunk, flex hose, elbows and gates.

Stone dust punishes lazy layout in a way sawdust does not. It is dense, abrasive, and it drops out of a moving airstream the moment the air slows. A run that would happily carry planer shavings will silt up solid with granite and engineered stone fines in weeks. The system then gets worse every day while nothing visibly breaks. What follows is about the duct itself: sizing, routing, slope, gates and upkeep. Not respirators, not slurry, not wet cutting.

What the Duct Actually Has to Do

A dust system has two separate jobs and they fail in different ways. The first is capture: pulling particles into the hood before they reach room air. The second is transport: keeping those particles moving inside the pipe all the way to the collector. Excellent hoods on a badly routed trunk will look good for a month and then choke. A well engineered main serving hoods that sit ten inches from the cut never grabs the dust at all.

Capture is governed by geometry and distance. Dust leaving a blade or cup wheel has real momentum and a direction, and the air a hood can pull toward itself falls off sharply with distance from the opening. A hood that works at three inches may do almost nothing at ten. The most valuable change in most shops is not a bigger fan but moving the pickup closer and shaping it so the dust is already headed into the opening.

Transport is governed by air velocity inside the pipe. Every material has a minimum conveying velocity below which particles settle and accumulate. Heavier particulate requires higher conveying velocity to stay suspended, which is why a stone shop cannot borrow duct sizing from a cabinet shop. Guidance for heavy industrial dust generally sits in the 3,500 to 4,500 feet per minute range, and the ACGIH Industrial Ventilation Manual gives empirical transport velocities spanning roughly 3,000 to 5,500 feet per minute depending on the material.

Round mains in heavy-dust service are commonly designed around 4,000 feet per minute, though the right figure varies by configuration. Understand what happens below that threshold, because it is a one-way process. Fines drop along the bottom of a horizontal run. The deposit adds roughness and resistance, airflow drops, more material settles, and the run keeps filling until a length of pipe is functioning as a storage bin instead of a duct.

There is a compliance dimension as well. Respirable crystalline silica is regulated by OSHA at a permissible exposure limit of 50 micrograms per cubic meter as an eight-hour time-weighted average, with an action level of 25 micrograms per cubic meter, under 29 CFR 1910.1053 and 1926.1153. Engineering controls come first in that framework, and a dust system that has quietly lost half its airflow is a control that is no longer doing its job.

Sizing and Routing the Run

Duct sizing runs in one direction only: from the hood back to the fan. Establish the airflow each hood needs, add up what has to move through each section of pipe, then pick a diameter that puts the air in that section at or above transport velocity. Picking a pipe size first because it matches a machine port, or because a length is leaning against the wall, produces branches that will never work regardless of the fan.

Start From Airflow at Each Hood

Begin with the machine builder’s recommended airflow for each pickup and treat it as a floor, not a target. Where no figure exists, size the hood by its opening and the capture distance you actually have, then verify with smoke and a meter once it runs. Write the number for every pickup on a sketch of the shop. That sketch, not the collector nameplate, is your design document.

Decide now whether the system serves all hoods at once or a limited number of open gates. Both are legitimate, but they produce different pipe. A shop that gates down to two or three active machines can run a smaller main and fan, at the cost of discipline. A shop that wants everything live needs the main sized for the full sum. Choose one, write it down, and label the gates so the choice survives staff turnover.

Size Branches for Velocity, Not the Machine Port

A branch carries one machine’s airflow, so it should be the smallest diameter that still delivers that airflow at transport velocity. Oversizing a branch is the most common mistake in shop-built systems and it is invisible: a six-inch branch where four inches belonged still looks like working duct, but the air inside is crawling and material drops out a few feet past the hood. Keep branch velocity at or slightly above main velocity.

Branches should enter the main at a shallow angle aimed downstream, tying into the side or top of a horizontal main rather than the bottom. A branch dropping into the bottom of a trunk becomes the catch basin for everything the trunk cannot carry. Where two branches enter close together, offset them along the run instead of landing them opposite each other, which creates turbulence exactly where the merging air is trying to keep moving.

Slope, Drops and Cleanouts

Horizontal runs are where stone dust accumulates, so minimize them and give the ones you need a consistent slope back toward the collector. Drops to a machine should come off the top or side of the main and run down to the hood, not up from a floor-level trunk. Long horizontal branches that dead-end at a machine are the classic problem: the last few feet see the lowest velocity, and that is where debris parks.

Cleanouts are not optional in stone service. Put an access port at the base of every vertical, at the end of every horizontal branch, and at each major direction change on the main. They cost almost nothing at installation and they are the difference between a fifteen-minute clearing job and cutting pipe open with a grinder. Make them reachable without a lift; a port twenty feet up behind a bridge saw will never be opened.

Static Pressure, Fan Curves and the Starving Branch

Every foot of duct, every elbow, gate and filter takes something out of the fan’s budget. That budget is static pressure, and a fan only produces its rated airflow at a specific point on its curve. As total system resistance rises, the operating point slides along the curve and airflow falls. Two shops with identical collectors get wildly different results because one built a short, smooth, correctly sized system and the other built a resistance obstacle course.

Fittings drain that budget fastest. A tight ninety-degree elbow costs the equivalent of a substantial length of straight pipe, and stacking three in a corner can consume more of the fan than the entire main run. Use long-radius sweeps. Where layout allows, replace a single ninety with two forty-five-degree bends separated by a short straight section, which turns the air more gently and keeps flow attached to the pipe wall instead of separating.

Flex hose is the other budget killer and it is worse than most people assume. The corrugated interior creates far more resistance than smooth pipe of the same diameter and length, and the penalty multiplies when the hose is compressed, kinked or looped on the floor. Treat flex as a short, straight, final connection where movement is genuinely required. Ten feet of coiled hose behind a machine is a throttle valve you installed on purpose and then forgot.

This is also the mechanism behind the most common upgrade failure: adding a machine to an existing trunk. The new branch does not create airflow, it divides it. The fan sees lower total resistance because another path opened, so it moves somewhat more total air, but that air is split across more openings and every existing hood loses velocity. A system that was marginal at three machines becomes ineffective at four, and the shop blames the collector.

Blast gates manage that division deliberately. One gate per branch, closed by default, opened only for the tool in use, concentrates available airflow where dust is being made. Position gates near the main where an operator can reach them without walking around a slab, and label each with the machine it feeds. An unlabeled gate in a busy shop is a gate that stays open, and an open idle branch is a hole your capture velocity escapes through.

Balancing is the discipline of making sure every branch gets its intended share. The simple approach for a small shop is to size each branch correctly on its own and use gates to control which are live. The engineered approach adjusts branch sizing so pressure loss along each path is roughly equal at design flow, letting several branches run together without one dominating. Either way, the short branch nearest the fan will always want more than its share.

Verify with instruments rather than opinions. A manometer and pitot tube, or an anemometer at the hood face, settles in minutes what an hour of arguing will not. Take readings at each hood and a few points on the main when the system is new and clean, write them on a card taped to the collector, and repeat quarterly. Trends matter more than absolute numbers: a hood reading down a third since installation is telling you something.

Keeping the System Working Over the Long Haul

Filter loading is the slow leak in every dust system. As cake builds on the media, resistance climbs, the fan slides down its curve, and duct velocity falls — so the ductwork starts dropping material at the same moment the collector is least able to pull. Cleaning cycles and filter change intervals are duct-performance items, not just consumable costs. If nobody is reading and logging a differential pressure gauge, install one that will be read.

Abrasion is the failure mode people forget until they find a hole. Stone fines moving at transport velocity erode the outside heel of every elbow and the wall opposite every branch entry. In a busy shop those are consumable components. Use heavier gauge or wear backs at elbows and at the collector inlet, inspect on a schedule, and keep a spare sweep on the shelf. A pinhole bleeds air continuously and shows up as lost capture at the farthest hood.

Leaks in general deserve more respect than they get. Every unsealed joint, loose clamp and open cleanout downstream of a hood steals airflow budgeted for capture. Seal joints, use proper clamps, and make sure gates seat rather than merely slide. A system assembled with tape and optimism loses a meaningful fraction of design flow through the seams alone, and that loss is spread evenly enough that no single spot ever looks like the culprit.

Build a short routine and make it somebody’s job. Open the cleanouts on a fixed interval and check what is inside. Tap the horizontal runs. Read the gauge and the hood velocities. Confirm gates match the operating plan. Ten minutes a week catches problems while they are still adjustments. Update the duct sketch every time a machine moves or a branch is added, because the numbers change and the balance shifts with it.

Getting the air right pays off everywhere else. Cleaner capture at the saw means cleaner surfaces for layout and marking, better visibility during a cut, and less abrasive grit migrating onto polished faces and into machine ways. If you are reworking the shop around a new machine, review the tooling and equipment range while you plan the duct run, since hood geometry and tool choice get decided together. Our shop guides and technical articles cover the setup decisions that surface once airflow is finally under control.