Energy and emission properties of burley tobacco stalk briquettes and its combinations with other biomass as promising replacement for coal

For this study we used 11 different biomass samples from Serbia in the form of fuel briquettes. Six were made of either burley tobacco stalks from Šabac tobacco fields (44° 45′ N, 19° 42′ E, Mačva district in Western Serbia), wheat straw, corn cob, both from Stara Pazova fields (44° 59′ N, 20° 10′ E, Srem district, Vojvodina province, Serbia), soybean straw and sunflower head remains, both from the Golubinci fields (44° 59′ N, 20° 04′ E, Srem District, Vojvodina province, Serbia), or beech sawdust from a wood processing company (origin unknown). Five samples were mixtures (in a 50:50 ratio) of tobacco stalks and one of the above mentioned biomass material. We opted for burley tobacco because stalks are mowed and dried together with leaves, which facilitates collection and transport and enables energy saving in drying. After drying, leaves are removed from and used for cigarette manufacture. Beech sawdust was selected as a material of the longest tradition in the production of energy briquettes in Serbia.

Sawdust and tobacco stalks were pre-dried as described elsewhere (23), whereas all other biomass was heaped and dried in a room for 30 days, during which time it was turned regularly to improve evaporation (24). The humidity of all biomass after drying was 9.66–10.85 % and complies with the European Standard of ≤12 % (25).

Dry biomass was then ground and homogenised in a mill, which is an integral part of the briquetting machine (Macinatore, MAC 500, CO.MA.FER, Collebeato, Italy), and subsequently by manual mixing to make the briquettes.

Briquettes (of 6 cm diameter, 5–11 cm long) were made with no binding materials using a briquetting machine (Macinatore, MAC 500) operating at a pressure of 1000 kPa and 50–150 kg/h output. They were then stored in large impregnated paper bags in an isolated room for 40 days until heat value and gaseous combustion product analysis, because Obernberger and Thek (24) found that four to six weeks of storing ensures pellet quality that is otherwise achieved only with biological additives. For combustion we only used briquettes of a uniform length of 8 cm.

The obtained briquettes were ground in a mill and sieved repeatedly to separate 0.5–1.0 mm particle fractions, which were taken for analysis. Samples were stored in paper bags in a cool and dry place for up to 15 days.

Moisture content was determined by ovendrying the samples (laboratory oven Digitheat 80 L, JP Selecta, Abrera, Spain) at 105 °C as described elsewhere (26).

Ash content obtained by combustion in the electric muffle furnace at 810 °C was determined following the EN 14775:2011 procedure (27).

Carbon (C), hydrogen (H), and nitrogen (N) content was determined as specified by the European Standard EN 15104:2011 (28). Sulphur (S) content was determined as specified by the European Standard EN 15289:2011 (29). Oxygen (O) content was calculated using the following equation:

1 O (%)=100-C (%)-H (%)-N (%)-S (%)-Ash (%) ${\rm{O}}\,{\rm{(\% ) = 100 - C}}\,{\rm{(\% ) - H}}\,{\rm{(\% ) - N}}\,{\rm{(\% ) - S}}\,{\rm{(\% ) - Ash}}\,{\rm{(\% )}}$

Nicotine (≥99 % purity) standard was purchased from Merck (KGaA, Darmstadt, Germany) and other solvents were HPLC-grade from Fisher Scientific (Pittsburgh, PA, USA).

Briquettes were ground in a mill (Wiley Mill, Model 4, Thomas Scientific, New Jersey, USA), passed through a 2-mm mesh, oven-dried to a constant dry weight at 60 °C for 24 h, distributed into 0.5 g lots, and extracted with 10 mL of 25 mmol/L sodium phosphate buffer (pH 7.8) at 30 °C for 24 h with constant agitation. The aqueous extract was filtered through a filter paper, diluted ten times with water, filtered again through a 0.45-µm pore mesh, and sealed in a screw-capped septum vial to permit automatic injection of a 20-µL aliquot. Samples were then eluted with an isocratic mobile phase containing 40 % (v/v) methanol and a 0.2 % (v/v) phosphoric acid buffer (pH 7.25) at a flow-rate of 0.5 mL/min. Nicotine was identified with a Waters high-performance liquid chromatograph (HPLC) (Waters Breeze, Binary Pump systems, Milford, MA, USA) at UV wavelengths between 210 and 400 nm and quantified at 254 nm as described elsewhere (30, 31).

Flue gas nicotine content was investigated in the smoke created by burning 1 kg of briquettes. At the top of the furnace flue we placed a Cambridge filter pad (92 mm in diameter; Bogwaldt, Germany) to collect smoke gases (32). The filter was then removed using laboratory tweezers, and nicotine particles extracted and quantified using the HPLC method described above (30, 31).

To determine heat values of each briquette type, samples were burned in an oxygen bomb calorimeter (IKA C400 Adiabatisch) according to the EN 14918 standard (33). The upper thermal power was calculated based on the amount of heat released minus the heat released by the formation of sulphuric and nitric acid during combustion in the calorimetric bomb. From the amount of available hydrogen and moisture we then calculated the lower thermal power as a realistic parameter for evaluating the heat value of the biomass.

The combustion products (CO₂, CO, NO, NO₂, NO_x, and SO₂) were analysed in gas released from 3 kg of briquettes burnt in a 65 kW burning chamber at 1000 °C. The chamber had a fixed grid, and biomass was inserted manually.

Gases were determined following the protocol described elsewhere (34), with a probe of a flue gas analyser (VARIO plus industrial, MRU MessgerätefürRauchgase und Umweltschutz GmbH, Wiener Neustadt, Austria) inserted at the mouth of the furnace pipe exhaust, and read from the device’s display. Three measurements were performed for each sample at minute 1, 5, and 9 of combustion that lasted ten minutes. Measurement characteristics (range and accuracy) of the MRU flue gas analyser are shown in Table 1.

All parameters met standard requirements for lab testing (35), measurements were done in triplicate, and the results are presented as means ± standard deviations (SD).

The CO₂ emission factor is the average amount of greenhouse gas emissions in relation to the emission source, assuming that all carbon from the biomass is completely oxidised (burnt) and converted into CO₂ (36). It is calculated based on the amount of carbon in the fuel at the lower heat value of the fuel, as follows:

2 EF=m(C)100x3.664LHWx1000(tCO2/TJ), ${\rm{EF = }}{{m(C)} \over {{\rm{100}}}}x{{3.664} \over {LHW}}x1000\,{\rm{(tC}}{{\rm{O}}_2}{\rm{/TJ),}}$

where EF is the emission factor, m (C) – carbon mass in the fuel (with total moisture expressed in %), LHW the lower heat value of the fuel (with total moisture expressed in MJ/kg), and 3.664 the stoichiometric coefficient.

Data obtained from the experiments are expressed as means ± standard deviations (SD). We used one-way analysis of variance (ANOVA) to compare mean differences between samples. Correlations between samples were tested with Pearson’s correlation coefficient. All analyses were run on the SPSS version 23.0 (37).

Our findings should be taken with some reserve, as the briquettes used in the experiments were not completely homogeneous, mixtures in particular. However, this relative heterogeneity in briquette composition reflects real-life situations and still has some informative value, especially in practical terms.