There’s a moment in every project review meeting where someone points to a rectangle on the P&ID labeled “Scrubber” or “Baghouse” and asks, “This one will work, right?” And everyone nods, because the equipment vendor said it would, and the vendor knows more about it than anyone in the room.
Then the system starts up, and six months later the stack test shows emissions 30% over the permit limit — and nobody knows why.
I’ve been in that meeting. This article is about how to avoid it.
Start With the Pollutant, Not the Equipment
The most common mistake in air pollution control is selecting equipment before you’ve characterized the emission. You know there’s dust, so you buy a baghouse. But not all dust is the same, and the wrong baghouse makes things worse.
Before you look at any equipment catalog, answer these four questions:
What’s the particle size distribution? If the D50 is above 10 μm, a cyclone might be enough. Between 1-10 μm, you need a baghouse or wet scrubber. Below 1 μm, you’re looking at high-efficiency filtration or electrostatic precipitators. I’ve seen plants spend $200,000 on baghouses for submicron fume that passed right through the filter media.
What’s the gas temperature? Baghouses have temperature limits — typically 200-260°C for standard polyester bags, up to 260°C for Nomex. Above that, you need ceramic or metal filter elements, or you need to cool the gas first. Cooling adds cost and complexity.
Is there moisture? Wet gas and baghouses don’t mix. The moisture condenses on the filter cake and turns it into mud. The bags blind off, the pressure drop spikes, and production stops. If your gas stream is wet, go with a wet scrubber or a dry scrubber with temperature control to stay above the dew point.
Is the pollutant a particulate, a gas, or both? Particulates get captured by mechanical means (cyclones, baghouses, ESPs). Gases need chemical treatment (scrubbers, adsorption, thermal oxidation). If you have both — which many industrial processes do — you need a combination system.
Baghouses: Simple But Sensitive
A baghouse is basically a giant vacuum cleaner. Dirty gas comes in, passes through fabric bags, and clean gas goes out. Dust collects on the outside of the bags and gets knocked off periodically by compressed air pulses.
The system works beautifully — when the conditions stay within design range. Here’s what goes wrong:
Air-to-cloth ratio. This is the single number that determines baghouse performance. It’s the gas flow rate divided by total filter area, expressed as m³/min per m². For most industrial applications, the target is 0.8-1.2 m/min. Go above 1.5 and you’ll be replacing bags every six months. Below 0.6 and you’ve bought more filter area than you need. The right number depends on the dust loading, particle size, and whether you have sticky or abrasive dust.
Pressure drop tells you everything. A new baghouse starts at 50-80 mm water column pressure drop. When it hits 150-180 mm, it’s time to pulse or replace bags. Trend this number daily. A sudden drop means a torn bag (gas bypassing filtration). A gradual increase means blinding (moisture or sticky dust). If you track nothing else, track pressure drop.
Bag material matters. Polyester is the default because it’s cheap. But it hydrolyzes above 70°C in the presence of moisture. Nomex handles higher temperatures. PTFE membrane coatings resist sticky dust. Fiberglass handles very high temperatures but is fragile. The extra cost of the right bag material is always cheaper than unplanned downtime.
Wet Scrubbers: Great for Gases, Complicated to Operate
If you need to remove acid gases (HCl, SO2, HF) or water-soluble VOCs, a wet scrubber is your best option. Exhaust gas contacts a scrubbing liquid — usually water with chemicals added — and the pollutants transfer from gas to liquid phase.
What the equipment vendor won’t emphasize enough:
Liquid-to-gas ratio. Too little liquid, and the pollutant doesn’t transfer. Too much, and you’re pumping water for no reason. The typical range is 2-5 L/m³ of gas. But the optimal ratio depends on the Henry’s Law constant of your pollutant — gases with low solubility in water need higher L/G ratios or chemical additives.
pH control is not optional. If you’re scrubbing acid gases, the scrubbing liquid becomes acidic over time. At pH below 6, absorption efficiency drops dramatically. You need continuous pH monitoring and automatic caustic or lime addition. A manual pH check once per shift isn’t enough — the pH can crash in minutes if the process upstream has an excursion.
The mist eliminator is the most neglected component. After the gas passes through the scrubbing section, it carries entrained liquid droplets. A mist eliminator — basically a mesh pad or chevron vanes — catches these droplets before they exit the stack. If the mist eliminator plugs, pressure drops builds and gas backs up into the process. If it fails, you get visible plume and liquid carryover. Inspect it quarterly.
The Stack Test: Pass It Before You Get the Official One
Every air permit requires periodic stack testing. The testing company comes with their equipment, sets up on your stack, and measures emissions for several hours. If you fail, you report the failure to the regulator and develop a corrective action plan. It’s a public record.
Before the official test, hire the same testing company to do a pre-test — or at minimum, do internal monitoring during representative operating conditions. I’ve seen plants schedule their official stack test during a maintenance shutdown week when production was at 50% of normal. The test passed. The regulator noticed the production data and asked questions. Don’t game the test. Test at worst-case conditions — if you pass at maximum production, you pass at everything.
—
Air pollution control equipment works when it’s selected for the actual pollutant at the actual operating conditions and maintained according to the actual operating data. It fails when someone picks a standard package from a catalog and assumes it’ll work because the vendor said so.
The difference between a system that works for 20 years and one that causes recurring problems is usually two hours of upfront characterization work that nobody bothered to do.