Multi-source field deployment of a low-cost electrostatic PM1.0–PM2.5–PM10 continuous emission monitoring system: Fuel-dependent emission signatures across gaseous-, liquid-, and solid-fuel industrial stacks

Continuous, size-resolved data on stack particulate matter (PM) are scarce for the small- and medium-scale industries of Southeast Asia. This study tests whether one low-cost electrostatic continuous emission monitoring system (CEMS) can resolve fuel-related PM signatures at gaseous-, liquid-, and solid-fuel stacks. In a companion study, the instrument had been validated against U.S. EPA Method 201 A at a heavy-fuel-oil (HFO) boiler (R2 > 0.92). Here it logged PM1.0, PM2.5, and PM10 every minute over 53 monitoring days at five stacks in northern Thailand fired with liquefied petroleum gas (LPG), HFO, or biomass. Emission levels followed the fuel, not the industry: campaign-mean PM10 was 0.030 and 0.60 mg·m–³ at the two LPG stacks and 30–70 mg·m–³ at the three HFO and biomass stacks. The PM2.5/PM10 ratio separated condensed-fuel stacks (81–88%) from premixed gaseous stacks (67–72%), by a margin comparable to the effective-density uncertainty, but did not separate HFO from biomass. The PM1.0/PM10 ratio (56–70%) showed no fuel ordering. Daily-mean PM10 varied 11- to 18-fold within single deployments as the firing rate changed, tracking plant fuel consumption (rank correlations 0.997 and 0.999). Absolute concentrations at the four stacks without on-site reference validation are indicative rather than reference-traceable, although the fuel-related patterns are more robust than the absolute magnitudes. The system ran stably up to 170 °C and 100% relative humidity, with impactor-plate loading setting a service interval of about 150 hours. Such a CEMS can therefore supply continuous, fuel-specific PM data for emission inventories and control-technology selection.