Research

Pollutant formation and control in wood stoves

Abstract

Wood stoves provide a low-cost, carbon-neutral and aesthetically appealing option for primary or supplementary heating in households. However, organic gaseous compounds and particulate matter (PM) emitted during the batch-wise combustion of wood logs in stoves are of strong concern because of their harmful effects on human health and the environment. Promising solutions to optimize combustion conditions and reduce pollutant emissions of wood stoves include the application of the sensor-based automatic air control system and an optimum air-staging strategy, but laboratory studies are still lacking. In addition, a better understanding of the formation mechanisms of the pollutants such as PM and organics is also needed for the development of efficient emission control strategies. The application of sensor-based air control systems requires a comprehensive evaluation of the performance of cost-effective sensors that can be used in small-scale combustion appliances such as wood stoves. In this PhD project, two low-cost absorption-based single-wavelength light-emitting diode (LED) sensors (280 nm and 417 nm wavelengths) aiming at the in situ and real time monitoring of the emission levels of soot and tars were tested on a modern wood stove. Optical measurements using the LED sensors were supplied by the measurements through wide-spectrum ultraviolet (UV at 188-360 nm wavelengths) spectroscopy, Flame Ionization Detector (FID), Scanning Mobility Particle Sizer (SMPS) and Electrical Low Pressure Impactor (ELPI+). Measuring signals of LED sensors generally correlated well with the PM and organic matter concentrations derived from other advanced instruments over the combustion cycle. The good correlations were validated under both the nominal combustion conditions and the intense combustion conditions characterized by significantly higher emissions. In addition, soot emission peaks indicated by the optical measurements were in good agreement with the combination of PM number and volume/mass concentration peaks identified by ELPI+. Finally, a preliminary evaluation of the long-time performance of the LED sensors shows the desirable measuring results over a period of several months. Therefore, given that the spectrum data are already useful for process control, and considering the advantages of non-intrusive measurements, low maintenance and fast response, the cost-effective optical sensors under investigation offer wide possibilities to be utilized for the process control and possibly also emission inventory in residential wood combustion. Air-staging is another important primary measure for combustion optimization and emission reduction of wood stoves, which generally consists of the primary air, secondary air and tertiary air that are directed from different positions to the stove chamber. While the position and pattern of the inlets for primary air and secondary air are rather fixed, the design of the tertiary air inlet is more flexible and needs further investigations. The way of the tertiary air fed directly to the space above the fuel-bed is studied to facilitate the volatile combustion in the initial combustion stage. Three types of tertiary air inlets (i.e., without injection holes, with small injection holes and with large injection holes) were tested on the automated wood stove with regard to their influences on emissions. With no quantitatively different overall combustion conditions observed in all cases, it was observed that the large-hole design results in a decrease of CO and total hydrocarbon concentrations and an increase of PM/soot emissions at the ignition phase. It is suspected that the use of large injection holes for tertiary air led to a higher degree of turbulence that may change the local flame conditions to higher temperatures and less but still fuel rich. However, no straightforward evidence was available due to the lack of localized in-flame measurements in the present study. The fundamental investigation on carbonaceous PM formation mechanisms is also an important topic of this PhD project, and a main focus was put on the interaction with fuel-bond inorganic constituents. Particularly, the influence of alkali metals is of importance due to their wide presence in biomass. In this context, PM emissions during the combustion of wood logs impregnated with potassium chloride (KCl) in the wood stove were characterized. Results show that the presence of KCl did not alter the particle size distribution (PSD) pattern but led to a lower number emission of ultrafine particles and a higher number emission of accumulation mode particles (size ~200-400 nm). In addition, the emission factor of condensed organic compounds increased with the increasing KCl content in the fuel, likely due to the facilitated heterogeneous condensation of semi-volatile hydrocarbons on pre-existing ash particles. A slight increase of soot emission factor was also observed, but the relevant influential mechanism was still unclear. In order to further understand the interaction between potassium and the soot formation process, and driven by the uncertainty in previous research regarding the influence of alkali metals on soot chemistry, a fundamental study using the flow reactor setup was conducted. Specifically, effects of KCl and potassium hydroxide (KOH) on the benzene and soot formation in fuel-rich oxidation of methane was investigated at atmospheric pressure and temperatures ranging from 1273 to 1673 K. Aromatic species are known to be important soot precursors. Particularly, the formation of the first aromatic ring is of interest as it is a critical step for soot formation during the combustion of aliphatic fuels as well as biomass fuels that initially contain few aromatic structures (note that lignin is the only biomass composition with aromatic rings). It was found that the addition of both potassium salts promoted the formation of acetylene and benzene below 1523 K. This promoting effect was most pronounced at a high C/O ration of 5 and in the presence of KOH, while it diminished at temperatures above 1573 K. A CHEMKIN-based kinetic model that could partially reproduce the promoting effect of KOH suggested that the buildup of OH radicals through the rate controlling reaction KOH + M ⇌ K + OH + M in the presence of KOH favors the production of active intermediates (e.g., methyl and vinyl radicals) that are critical to benzene formation. The particle measurement by SMPS indicated a net increase of soot volume concentration and a shift of PSD to larger diameters at C/O=5 in the presence of KOH, but there was no quantitative difference in the mass of carbonaceous particles collected on filters. To conclude, this work proved the good performance of low-cost optical sensors for the online monitoring of either low or high levels of soot and tar emissions from a modern wood stove. The observed results show a promising application of optical measurements in process and emission control of residential wood combustion. In addition, fundamental studies on the influence of potassium on PM formation indicated that interactions existed between inorganic species and the organic soot chemistry.

Info

Thesis PhD, 2020

UN SDG Classification
DK Main Research Area

    Science/Technology

Corresponding Author

Contributors

To navigate
Press Enter to select