The emissions from co-firing of biomass and torrefied biomass with coal in a chain-grate steam boiler

ABSTRACT

In this study, biomass of rice straw (RS) and wood (WD) and their torrefied biomass (RS_T and WD_T) were used as solid biofuel (SBF) for co-firing individually with coal in a commercial continuous chain-grate steam boiler system, which was conducted at fixed input rate of heating value of mixture of SBF and coal and at fixed airflow rate. The effects of key system parameters on the gaseous and particulate pollutions and ash were examined. These include SBF type and blending ratio (R_BL) of biomass (i.e., SBF) in the mixture of coal and biomass based on heating values for co-firing.

The results indicated that wood, which possesses high heating value while less amount of ash, is more suitable for co-firing with coal than rice straw. Torrefaction can increase the heating value of biomass and homogenize its property, being beneficial to co-firing. Also, torrefaction can decompose the hydroxyl group of biomass, which makes biomass tending to possess hydrophobicity. This, in turn, helps the storage and transportation of biomass. Generally, adding the RS (with R_BL = 5-10%), WD (2-15%), RS_T (2-10%) and WD_T (2-20%), respectively, with coal decreases the emissions of NO_x and SO₂, but increases that of CO (except RS_T). The emission of HCl is little. The addition of biomass also increases the emission of fine particulate matters (PM) especially PM_2.5 in the flue gases, raising PM_2.5/PM₁₀₀ from 34.87 to 78.35 wt.% (Case 50%WD_T). These emissions for the Cases tested satisfy with Taiwanese emission standards of stationary sources which set limitations of NOx, SO₂, CO and HCl < 350, 300, 2000 and 80 ppmv, while PM < 50 mg/m³, respectively. The results support the use of RS, WD, RS_T and WD_T for co-firing with coal.

Implications: This study examined the suitability of using solid bio-fuels to co-fire with coal in an industrial chain-grate steam boiler system with a capacity of 100 kW, in order to achieve carbon-free emissions. Both biomass and torrefied biomass of solid bio-fuel were tested. The findings would be useful for proper design and rational operation of solid bio-fuel/coal co-firing combustion matching the appeal of sustainable material management and circular economy of biomass, and of adaptation of global warming induced by greenhouse gases. It also provides information for policy-makers to promote the co-firing application of biomass and related bio-waste materials.

Introduction

Nowadays, the considerable use of fossil fuel causes many negative effects on the earth, such as global warming, climatic change and energy depletion. As a result, the development of renewable energy is an essential goal in most country. Among those renewable energies, bio-energy has an extensive application around the world. Biomass goes through photosynthesis combining carbon dioxide and water to grow up. Finally, biomass is combusted to generate heat while emitting CO₂ returning to the atmosphere. Because of this carbon cycle, biomass can be regarded as a carbon-neutral fuel. Therefore, the CO₂ produced from the combustion of biomass is not counted as CO₂ emission. The more biomass the power plants substitute for coal, the less CO₂ will be released. However, as a solid fuel, biomass has a few defects comparing with coal, such as high hydrophilicity, low heating value, low energy density and low grindabilty. Despite of these shortcomings while in order to mitigate the negative impacts of fossil fuel usage, the biomass co-firing technology has been considered for the increasing biomass usage in combustion and power-generation sectors (Al-Mansour and Zuwala 2010). There are three main methods to conduct co-firing including direct co-firing, parallel co-firing and indirect co-firing (Al-Mansour and Zuwala 2010; IEA, 2013; Maciejewska et al. 2006). Due to the low cost of firing systems, direct co-firing is the primary form among these methods.

Direct co-firing still faces some challenges. The different blending ratios of biomass to coal as well as the type of biomass used can easily affect the performance of combustion (Hupa 2005). Heating values of biomasses, fuel feeding rates and combustion efficiencies of boilers all have influences on the co-firing efficiency and energy production (Hupa 2005; Maciejewska et al. 2006).

Whether the existing facilities for controlling pollutants can also be applied for combustion of biomass meeting with the pollutant emission standards is now an important issue regarding the combustion of biomass, which has no homogeneous property (Veijonen et al. 2003). Main emissions released from co-firing of biomass with coal contain CO, SO₂, NO_X and fine particles. CO formation is highly related to combustion efficiency. Biomass, which has a high content of volatiles, will fire and decompose quickly leading to a decrease of the retention time of combustion. This also makes biomass co-firing encountering low combustion efficiency with more CO emission (Gani et al. 2005; Limousy et al. 2013; Wan et al. 2008). Several researches indicated that the N content in fuel, char loading in combustion chamber, and combustion temperature influence the NO_X emission (Baxter 2005; Dayton 2002; Gani et al. 2005; Hupa 2005; Johnsson 1994; Leckner and Karlsson 1993; Limousy et al. 2013; Veijonen et al. 2003; Wan et al. 2008). Thus, NOx emitted from biomass co-firing is case-dependent. According to other researches, high content of volatiles in biomass can effectively form a fuel-rich zone in fire during combustion so that the organic nitrogen materials from the fuel can easily converse to harmless N₂ (Beř, Jacques, and Farmayan 1981; Dayton 2002; Niksa et al. 2003). By contrast, SO₂ emissions and blending ratios of biomass in co-firing are presented as a linear relationship as reported by Hupa (Hupa 2005). The more biomass is added during co-firing, the less SO₂ is emitted. Because of low sulfur content in biomass while with some calcium mixtures existing in ashes produced from combustion of biomass, more SO₂ can be reduced (Baxter 2005; Leckner and Karlsson 1993; Maciejewska et al. 2006; Veijonen et al. 2003; Wan et al. 2008). Due to a higher content of Cl in biomass than that in coal, the exit gas from biomass co-firing will also release more HCl than sole coal firing (Desroches-Ducarne et al. 1998; Wan et al. 2008; Wei et al. 2009). As for particle issue, several studies pointed out that combustion of biomass resulted in the release of fine particles. Concentrations of PM₁₀, PM_2.5, and PM₁ (particulate matters with sizes ≦ 10, 2.5, and 1 μm, respectively) emitted from combustion of biomass have a much wider variability than these from combustion of coals (Jiménez and Ballester 2005; Maenhaut et al. 1999; Wang et al. 2007; Zhang et al. 2011), while combustion of biomass preferentially releases PM₁ (Ruscio, Kazanc, and Levendis 2016).

This study analyzed these primary pollutants and ash formed from co-firing of rice straw (RS), wood (cryptomeria) (WD), and their torrefied biomass (RS_T and WD_T) noted as solid bio-fuel (SBF) with coal in order to examine the feasibility of co-firing of these SBF in Taiwan. Effects of types and torrefaction of SBF, and the blending ratio of SBF in SBF/coal mixture based on heating values (R_BL) on the performance of co-firing of SBF with coal were elucidated.

Material and methods

Materials and procedures of manufacturing RS_T and WD_T

This study used bituminous coal (BC) mixing with other granulated solid bio-fuel SBF including rice straw RS, cryptomeria wood WD, and torrefied rice straw RS_T, and wood WB_T as fuels. The sizes of these SBF pellets are listed in Table 1. The RS_T and WD_T were made by torrefaction of RS and WD, respectively, employing a rotary furnace with the size of 12 m in length, 1 m in inside diameter, and an inclination of 3% (Ozone Engineering Technology Co., Yi-Lan, Taiwan). The torrefaction was conducted with the carrier gas of nitrogen at temperatures of 533 and 563 K for RS and WD, respectively, at time of 60 min and rotating speed of 1 rpm.

Table 1. Sizes of various carbonaceous materials used in this study.

	BC	RS	WD	RST	WDT
Pellet diameter (mm)	1-30	8	12	7.5 ± 0.5	11 ± 0.5
Thickness^a or length^b (cm)	1-3^a	1.5–2.5^b	1.5–3.0^b	1-2^b	1.5–3.0^b

Notes. BC, RS, WD, RS_T, WD_T: Bituminous coal, rice straw, wood, torrefied RS, torrefied WD;

^{a, b}Thickness, length.

BC firing and BC co-firing with SBF

The commercial chain-grate steam boiler system was housed in Li An Don Chemical Factory in Miao-Li, Taiwan. The system was composed of a water-tube steam boiler with designed feeding capacity of 14,000 kg/hr and pressure capacity of 16 kg/cm² G, a feed-water heater for heating cold water from room temperature to 333 K (60°C), an air pre-heater, a fuel dryer, a bag filter which can remove the particles with diameters above 100 μm, a fuel hauling equipment and a bottom ash conveying device. The combustion was performed with continuous coal/biomass feeding of 1,500 kg/hr for 4 hr. The bottom and fly ashes were collected at ash discharge ports of a combustion chamber and bag filter, respectively, after 3-hr operation, while the flue gases including particulates were sampled for analysis after 1-hr operation as the boiler was ensured to reach stable combustion.

This work conducted 17 combinations (100% BC firing, denoted as 100%BC (100 – y) kcal % BC co-firing with y kcal % SBF, symbolized as y%SBF for SBF = RS, WD, RS_T, WD_T) of co-firing (listed in Table 2) applying fuels from four biomasses and one bituminous coal with continuous feeding. For example, 2%RS stands for 2 kcal % RS mixed with 98 kcal % coal. The feeding rate was set at 1500 kg/hr which can be adjusted to maintain stable steam generation. The experimental schematic diagram of co-firing process is illustrated in Figure 1.

Table 2. The amount of fuels used for co-firing in full-scale tests.

Test	Coal (BC): Percent to heat load (kcal %)	Biomass: Percent to heat load (kcal %)	Total heat demand (kcal/hr)	Coal weight (kg)	Biomass weight (kg)
100%BC	100	0	8,250,000	4288	0
2%RS	98	2	8,250,000	4202	125
5%RS	95	5	8,250,000	4074	313
10%RS	90	10	8,250,000	3859	626
2%WD	98	2	8,250,000	4202	107
5%WD	95	5	8,250,000	4074	266
10%WD	90	10	8,250,000	3859	533
15%WD	85	15	8,250,000	3645	799
2%RST	98	2	8,250,000	4202	93
5%RST	95	5	8,250,000	4074	233
10%RST	90	10	8,250,000	3859	466
2%WDT	98	2	8,250,000	4202	90
5%WDT	95	5	8,250,000	4074	224
10%WDT	90	10	8,250,000	3859	448
15%WDT	85	15	8,250,000	3645	672
20%WDT	80	20	8,250,000	3430	896
50%WDT	50	50	8,250,000	2144	2241

Figure 1. Schematic diagram of chain-grate steam boiler system for co-firing process.

Solid analysis

Proximate, ultimate, metal and heating value analyses were performed in accordance with the standard methods from American Society for Testing and Materials (ASTM), National Institute of Environmental Analysis (NIEA), Chinese National Standards (CNS) and Taiwan Bureau of Standards, Metrology & Inspection (TBSMI). The detailed descriptions of the methods of analyses and the instruments employed can be found in the study of Chen et al. (2017a).

Gas and particle analyses

In this study, flue gas and particle detections were authorized to SGS Taiwan Ltd. (Taipei, Taiwan). The methods and applied instruments of these analyses are listed in Table 3. Concentrations of CO (range of 0–5000 ppmv), CO₂ (range of 0–20 vol.%), O₂ (range of 0–25 vol.%), SO₂ (range of 0–3000 ppmv) and NOx (range of 0–2500 ppmv) in flue gas were measured by a PG250 portable gas analyzer (Horiba, Kyoto, Japan). Non-dispersive infrared (NDIR) absorption method was used to quantify CO, CO₂ and SO₂, while cross-modulation ordinary pressure chemiluminescence and a galvanized Zr cell for NOx and O₂ measurements, respectively. The sampling unit comprises a filter, a mist catcher, a pump, an electronic cooling unit, and an NO₂ to NO converter. HCl was measured by a mercuric thiocyanate method (NIEA A412.73A). The HCl in flue gas was absorbed by sodium hydroxide solution, followed with the addition of mercuric thiocyanate and ammonium iron(III) sulfate solutions and the measurement of absorbance of ferric thiocyanate. Particle size distribution in exhaust gas was monitored by Apex cascade impactor CI-700-K (Apex Instruments, Inc., Fuquay-Varina, N.C., U.S.A.), which includes seven-stage impactor, a PRA-K pre-cutter, nozzles, interconnecting tube, glass fiber filters (45 mm, 47 mm and GFDN filters) and foil collection plate. The particle size distribution of stack particulate matters measured from the Cascade Impactor was analyzed by windows-based Cascade Impactor Data Reduction System software.

Table 3. Measurements of flue gas and instruments.

Item	Measure method	Instrument
Particle size distribution	USEPA M201A	Apex cascade impactor CI-700-K
CO	NIEA A704.04C	HORIBA PG250
CO2	NIEA A415.72A	HORIBA PG250
O2	NIEA A432.73C	HORIBA PG250
SO2	NIEA A413.74A	HORIBA PG250
NOx	NIEA A411.74A	HORIBA PG250
HCl	NIEA A412.73A	Spectrophotometer

Results and discussion

Basic properties of fuels

Tables 4–5 and Figure 2 show the results of proximate, ultimate and dry-basis high heating value (H_HD) analyses of bituminous coal BC, rice straw RS, wood WD, RS_T and WD_T, respectively. These materials can be classified into three groups: coal, biomass and torrefied biomass. In this study, the coal used was bituminous coal with a relatively higher water content of 30.84 wt.% compared with reference due to wet weather in Taiwan. The coal has the highest fractions of fixed carbon and carbon element standing for 30.5 wt.% (wet basis) and 58.77 wt.% (dry basis), respectively. Therefore, the H_HD of coal is the highest of 5772 kcal/kg among these fuels. The biomass group includes rice straw and wood, which have relatively high oxygen and hydrogen contents and majorly consist of volatiles. Heating values of these biomasses are related to their carbon contents (Table 5) leading to a comparatively higher H_HD of wood (4645 kcal/kg) than that of RS (3954 kcal/kg). After torrefaction, the biomass can be carbonated, and the energy intensity of biomass can also be enhanced (Chen et al. 2017a). Thus, the H_HD of RS and WD increase from 3954 and 4645 kcal/kg to reach 5308 and 5522 kcal/kg for RS_T and WD_T, respectively. The quality of torrefied biomass can also be close to coal (Chen et al. 2017a). Moreover, less oxygen and hydrogen retained in RS_T and WD_T than RS and WD indicate that they are more hydrophobic.

Table 4. Proximate analyses of fuels.

Fuel	Moisture^a (%)	Volatile matter^a (%)	Fixed carbon^a (%)	Ash^a (%)	Ash^b (%)
Coal (BC)	30.84 ± 0.07	31.42 ± 0.23	30.5 ± 0.13	7.24 ± 0.17	10.47
RS	10.87 ± 0.03	58.26	18.71	12.16	13.64
WD	6.72 ± 0	77.96 ± 0.41	14.08 ± 0.52	1.24 ± 0.11	1.33
RST	11.84 ± 0.08	50.61 ± 0.02	20.37 ± 0.53	17.18 ± 0.44	19.49
WDT	5.28 ± 0.14	61.41 ± 0.03	16.14 ± 0.3	17.17 ± 0.18	18.13

Notes. ^awt.% on wet basis; b: wt.% on dry basis.

Table 5. Ultimate analyses of fuels.

Fuel	C (%)	H (%)	O (%)	N (%)	S (%)
Coal (BC)	58.769 ± 2.063	4.361 ± 0.198	21.634 ± 0.376	0.812 ± 0.047	0.708 ± 0.066
RS	40.68 ± 0.949	5.95 ± 0.275	38.204 ± 0.453	0.781 ± 0.059	0.685 ± 0.099
WD	48.167 ± 0.095	6.809 ± 0.004	43.246 ± 0.019	0.246 ± 0.037	0.469 ± 0.023
RST	51.196 ± 0.033	4.501 ± 0.187	22.332 ± 0.788	1.017 ± 0.051	0.356 ± 0.238
WDT	51.709 ± 0.123	6.411 ± 0.018	39.257 ± 0.257	0.273 ± 0.002	0.434 ± 0.049

Notes. wt.% on dry basis.

Figure 2. Heating values in dry basis of fuels.

The ash contents of fuels are different withRS of 13.64 wt.% higher than WD of only 1.33 wt. % on dry basis. After long combustion in the boiler system, the high ash content of fuel causes scale deposition (fouling), which needs more frequent maintenance in order to keep the boiler system in good condition.

Chemical compositions of fuels in Table 6 indicate that coal has comparatively higher contents of Fe and Al than RS and WD. However, the RS and WD, which take nutrients to grow themselves, possess higher contents of K, Na, Ca and Mg than coal. It is worth to mention that Cl content is also an important indicator for the usage of biofuel as chlorides formed after the combustion of fuel can lead to corrosion of boiler. The Cl content of bituminous coal is below 0.002 wt.% much lower than those of biomasses, especially of RS and RS_T of 0.359 and 0.313 wt.%, respectively.

Table 6. Chemical compositions of fuels.

Composition	Coal (BC)	RS	WD	RST	WDT
K (%)	0.0242	1.69	0.23	1.83	0.167
Na (%)	0.00907	0.0397	0.0254	0.0502	0.0375
Ca (%)	0.151	0.552	0.338	0.73	0.333
Mg (%)	0.0776	0.256	0.136	0.312	0.125
Al (%)	0.226	0.0498	0.0118	0.0989	0.00444
Fe (%)	0.632	0.126	0.0223	0.246	0.0117
Ti (%)	0.00928	<0.00025	0.000357	0.000464	<0.00025
Si (%)	0.0204	0.0211	0.00573	0.0104	0.00268
Hg (%)	<0.0000015
Ni (%)	<0.0005
Cl (%)	<0.002	0.359	0.113	0.313	0.195

Notes. wt.% on dry basis.

Compositions of gases from co-firing of biomass with coal

The compositions of gases from co-firing tests are shown in Table 7. The concentrations of emissions were analyzed after the boiler reached stable combustion in 1 hr. The data of Table 7 were the average concentrations computed by the results from per minute measurement during 1 hr combustion. As noticed in Section 2.2, “%name of fuel” represents for the kcal proportion of the specified fuel in mixture, for example, 2%RS denotes that there is 2 kcal % RS mixed with 98 kcal % coal. The accuracy of data can be reflected by the standard deviation along with average value, as also displayed in Table 7.

Table 7. Components of flue gas in full-scale co-firing tests.

Test	NOx (ppmv)	SO2 (ppmv)	CO (ppmv)	CO2 (vol.%)	O2 (vol.%)	HCl (kg/hr)	HCl (ppmv)
Coal (BC)	52.0 ± 1.4	42.1 ± 3.6	228.2 ± 108.3	5.47 ± 0.24	13.67 ± 0.28	0.12	ND<2.85
2%RS	62.3 ± 5.8	37.9 ± 3.2	281.9 ± 132.9	5.48 ± 0.29	14.30 ± 0.36	0.12	3.00
5%RS	41.7 ± 7.3	22.9 ± 4.0	394.4 ± 97.2	3.21 ± 0.56	17.00 ± 0.65	0.25	6.48
10%RS	31.9 ± 15.1	21.7 ± 4.6	652.7 ± 189.7	1.79 ± 0.60	18.55 ± 0.74	0.25	6.35
2%WD	41.9 ± 7.3	27.7 ± 5.2	246.6 ± 142.2	5.11 ± 0.96	15.29 ± 1.00	0.12	ND<2.66
5%WD	48.3 ± 1.4	48.5 ± 5.8	521.6 ± 88.5	6.56 ± 0.25	13.68 ± 0.29	0.12	ND<2.66
15%WD	40.3 ± 2.2	32.7 ± 12.4	676.1 ± 493.6	6.60 ± 0.74	13.72 ± 0.84	0.10	ND<2.61
2%RST	47.3 ± 8.5	6.4 ± 2.1	141.9 ± 61.6	5.36 ± 1.08	15.14 ± 1.26	0.11	ND<2.69
5%RST	42.9 ± 1.5	7.6 ± 1.4	80.6 ± 14.1	4.90 ± 0.34	15.69 ± 0.37	0.11	ND<2.69
10%RST	44.4 ± 3.1	10.7 ± 2.5	163.1 ± 115.7	5.56 ± 0.35	14.91 ± 0.36	0.22	ND<5.36
2%WDT	46.4 ± 2.0	33.9 ± 4.0	429.4 ± 297.3	6.42 ± 0.47	13.91 ± 0.51	0.11	ND<2.69
5%WDT	39.2 ± 2.4	34.1 ± 4.5	532.8 ± 222.5	6.03 ± 0.29	15.02 ± 0.31	0.19	4.84
10%WDT	43.4 ± 1.9	35.8 ± 1.6	231.3 ± 77.4	6.19 ± 0.17	14.87 ± 0.18	0.12	2.94
15%WDT	39.4 ± 2.1	27.1 ± 2.4	238.6 ± 121.2	6.02 ± 0.35	15.04 ± 0.37	0.12	ND<2.72
20%WDT	35.7 ± 3.0	21.4 ± 2.9	227.8 ± 87.9	5.50 ± 0.44	15.62 ± 0.48	0.12	ND<2.72
50%WDT	99.0	97	461	15.00	15.24	0.10	ND<2.59
Emission standard^a	150^b 250^c 350^d	100^b 300^c,d	2000

Notes. ^aEmission standards of stationary sources in Taiwan (TEPA, 2013a);

^bGas fuels

^cLiquid fuels

^dSolid fuels

The volume concentrations of emitted gaseous pollutants were adjusted to the conditions at 273 K and 1 atm with 6% O₂.

Emission of CO

CO emission from combustion is related to the combustion efficiency of the boiler. When the fuel at high temperature does not have enough oxygen or the feeding rate of fuel is too fast such that the fuel has insufficient time for combustion in the boiler, more CO will exist in the flue gas. As shown in Table 7, un-torrefied biomass co-firing with coal generates a higher concentration of CO than sole-firing of coal. For RS and WD, a higher blending ratio of biomass also generally resulted in more emission of CO. It is known that biomass exhibits an instant decomposition via de-volatilization at mild temperature while coal offers smooth combustion performance. This causes the barrier of good combustion efficiency for co-firing because of the different combustion characteristics of two different fuels. For the practical need, the heat output must be maintained at specific value so that the supply of steam can be stable. When the biomass had a quick mass loss during combustion, while the feeding rate had to be adjusted according to the heat output, it resulted in low combustion efficiency and high CO concentration. This phenomenon was observed in the mixture of RS and BC.

Regarding the function of torrefaction, Chen et al. (2017a) reported that the torrefied biomass possesses property closer to coal than the un-torrefied biomass and, therefore, its de-volatilization during combustion causes less mass loss, thus exhibiting combustion trend more similar to coal. For the co-firing of torrefied biomass with coal, the results of Table 7 show that CO emissions for Case RS_T decrease comparatively lower than those for Case RS, indicating the beneficial effect of torrefaction on RS. The reduction of CO emission for Case WD_T is not apparent. However, the concentrations of CO emitted for the cases of sole-firing of coal, and co-firing of RS, WD, RS_T and WD_T with coal examined in Table 7 all meet the CO emission standard of stationary pollution source in Taiwan (lower than 2000 ppmv) (TEPA, 2013a).

It is noted that the unburned carbon in fuel could turn into the carbon contents in fly ash and bottom ash, so high carbon contents in both ashes may represent an inefficient combustion. Hence, Armesto et al. (2002) reported that the combustion efficiency of biomass in the boiler can be determined by the carbon contents in fly and bottom ashes. Figure 3 shows carbon contents in fly and bottom ashes and CO emissions of coal (BC) and SBF (2% and 10% RS and RS_T; 2% and 15% WD and WD_T). The results indicated that carbons in fly and bottom ashes showed different trends. However, bottom ash possessed higher carbon contents than fly ash. Carbons in fly ash for SBF were all higher than that of BC, while carbons in bottom ash of SBF would generally lower than that of BC. Comparing the findings of CO emission, only torrefied RS displays the lower CO concentration than that of BC. It is quite difficult to identify general trends for combustion efficiency in co-firing coal with SBF based on the carbon balance of fuel ashes.

Figure 3. Carbon contents in fly and bottom ashes and CO emission of coal and SBF.

Emission of NOx

In the light of the study of Nussbaumer (2003), NO and NO₂ (summarized as NO_X) can be formed in three different mechanisms during combustion. Thermal NO_X and prompt NO_X are originated from the nitrogen in the air at high temperatures and in the case of prompt NO_X in the presence of hydrocarbons. Further, fuel NO_X can be formed from nitrogen-containing fuels. For biomass combustion, fuel-bound nitrogen is the main source of NO_X emissions, while thermal and prompt NO_X are not relevant due to relatively low temperatures as has been shown by theoretical and experimental investigations. Fuel nitrogen is converted to intermediate components such as HCN and NHi with i = 0, 1, 2 and 3. These can be oxidized to NO_X if oxygen is available, which is the case in conventional combustion. If no oxygen is present, intermediates can interact in the reduction zone and form N₂ via reactions such as NO + NH₂ = N₂+ H₂O. In combustion, char content is also an important factor that can help the NOx reduction reaction. Consequently, the N content, oxygen supply and char content during combustion are the main factors for the formation of NOx.

Observing element compositions of all fuels in Table 5, the RS_T and WD_T possess nitrogen contents higher than RS and WD due to torrefaction. All except RS_T contain less nitrogen than coal. These results can explain why NO_X concentrations from co-firing of biomass containing lower nitrogen contents are lower than those of coal. Although torrefaction slightly increases the nitrogen content, there are no apparent effects on the NOx emissions between Cases 2-10%RS, 2-15%WD and 2-10%RS_T, 2-15%WD_T (Table 7). The case with high blending ratio R_BL of 50%WD_T contains biomass which has relatively low fixed carbon content (Table 4) in full scope of about 23.32 wt.%. The excess use of WD_T for substituting BC thus results in the NOx emission reaching 99 ppmv, which is about twice as much as that of BC. Nevertheless, regressions of NOx emissions detected from the combustion of fuels at all cases with their fixed carbons (R² = 0.4316) and nitrogen contents (R² = 0.4427) showed poor linear relationships (data not shown).

Emission of SO₂

Biomass generally has lower S content than coal so that coal co-fires with biomass forms less SO₂. In addition, according to the study of Veijonen et al. (2003), the components of fuels can affect the SO₂ formation when these fuels are combusted together. SO₂ formed from combustion can react with Ca in fly ash produced mainly from the combustion of biomass so that SO₂ emission can decrease. Sotiropoulos, Georgakopoulos, and Kolovos (2005) proved that increasing the free CaO in fly ash decreases the respective SO₂ emissions. The free CaO (often called the “free lime”) is the amount of calcium oxide present as CaO, that is, not bound into the cement minerals. The available free CaO facilitates the natural desulfurization (CaO + SO₂ + 1/2O₂ = CaSO₄), thus reducing SO₂ emissions to the atmosphere. Davis and Fiedler (1982) also indicated that the sulfur retention in fly ash emitted from coal-fired boilers is a function of CaO content of fly ash, suggesting that the presence of CaO enhances the adsorption of sulfur.

Based on Table 7, after mixing coal, respectively, with RS and WD, SO₂ emission from co-firing decreases slightly compared with that from combustion of pure BC of 42.1 ppmv because of their low S contents (RS of 0.69 wt.% and WD of 0.47 wt.% lower than BC of 0.71 wt.% as in Table 5). Figure 4 illustrates the relation between SO₂ emissions and Ca contents in fly ash from all Cases. The tendency indicates that the more Ca in fly ash is, the less SO₂ is emitted, which is in accordance with the findings of Veijonen et al. (2003).

Figure 4. Effect of calcium content in fly ash on SO₂ emissions.

The S content in biomass can be reduced after torrefaction. It is removed as part of the volatiles, which contains sulfide, during decomposition and gasification (Chen et al. 2017b). RS_T and WD_T have lower S contents of 0.36 wt.% and 0.43 wt.% than RS and WD, respectively. The SO₂ emissions of 2-10%RS_T are much lower than those of 2-10%RS. On the other hand, combustions of 2-20%WD_T emit close concentrations of SO₂ as of WD. The trends of RS and WD were similar to their Ca contents as shown in Table 6, highlighting the role of Ca contents in SBF for desulfurization. Moreover, on the base of Ca contents in fly ash of Cases 2-50%WD_T (9.13, 11.3, 11.1, 11.1, 11.5 and 5.65 wt.%), Case 50%WD_T has the lowest Ca content, which results in the highest SO₂ emission of 97 ppmv.

Emission of HCl

The results in Table 7 show that only Cases RS and 5-10%WD_T have detectable HCl concentrations in emissions. The HCl concentrations have no apparent difference between the combustion of BC and biomasses of other cases. The Cl in biomasses may be combined with alkaline such as Ca, K and Na and transferred into ashes during combustion (Ren et al. 2018). This thus results in low HCl emission in exhausted gas. High Cl content in ashes from biomass co-firing can be expected. The Cl contents in fly ashes of SBF/coal co-firing were 0.38–1.82 wt.% higher than that of sole coal firing of 0.08 wt.%. As for the Cl contents of bottom ashes, SBF/coal co-firing resulted in 0.035–0.305 wt.% also higher than sole coal firing with <0.01 wt.%.

Emission of particulate matters

The concentrations of PM emitted after bag filter for cases examined are listed in Table 8. For sole BC combustion, the PM concentration is at 0.61 mg/Nm³. After adding biomass to co-fire with BC, there is no obvious difference of PM concentrations of 0.47–0.95 mg/Nm³. Only Cases 2%RS_T and 2%WD_T have relatively high PM concentrations of 0.95 and 0.78 mg/Nm³, respectively. On account of a bag filter connected with the boiler, PM with the size larger than 100 μm was collected, cutting down the emissions of PM. Table 8 also presents the weight percent of PM₁/PM₁₀₀, PM_2.5/PM₁₀₀, PM₁₀/PM₁₀₀, PM_2.5/PM₁₀, PM₁/PM₁₀ and PM₁/PM_2.5. The PM₁₀/PM₁₀₀ values are as high as 99.10 wt.% or greater, indicating that the emitted PM₁₀₀ mostly consists of PM₁₀. This further reveals that the bag filter employed efficiently removed most of PM larger than 10 μm. After torrefaction, co-firing of RS_T has a higher emission of PM_2.5 (0.47–0.72 mg/Nm³) than that of RS (0.37–0.49 mg/Nm³), while co-firing of WD_T has no noticeable difference of emission of PM_2.5 (0.46–0.58 mg/Nm³) with that of WD (0.44–0.47 mg/Nm³) at the same level of blending ratio.

Table 8. Particulate matters (PM) concentration (mg/Nm³) and percentiles of PM₁, PM_2.5, and PM₁₀ (wt.%).

Test	PM concentration (mg/Nm^3)	PM1/PM100	PM2.5/PM100	PM10/PM100	PM2.5/PM10	PM1/PM10	PM1/PM2.5
100%BC	0.6105	21.92	34.87	99.10	35.19	22.12	62.87
2%RS	0.6441	53.91	73.89	99.97	73.91	53.93	72.96
5%RS	0.7893	44.36	61.99	99.99	62.00	44.36	71.55
10%RS	0.5707	45.47	65.70	99.92	65.75	45.51	69.21
2%WD	0.6210	57.27	75.36	99.98	75.38	57.28	75.98
5%WD	0.6503	56.43	73.20	99.98	73.22	56.44	77.09
10%WD	0.6091	53.23	73.87	100.00	73.87	53.23	72.05
15%WD	0.6294	53.35	71.68	99.94	71.73	53.38	74.42
2%RST	0.9461	56.94	76.17	100.00	76.17	56.94	74.75
5%RST	0.6253	58.34	76.80	100.00	76.80	58.34	75.96
10%RST	0.6211	56.52	76.31	99.98	76.32	56.53	74.06
2%WDT	0.7803	55.53	74.69	99.99	74.70	55.54	74.35
5%WDT	0.6293	54.12	73.67	100.00	73.70	54.13	73.47
10%WDT	0.6262	55.03	75.12	99.99	75.12	55.04	73.27
15%WDT	0.6117	57.22	75.90	100.00	75.90	57.22	75.39
20%WDT	0.6348	56.25	74.48	99.99	74.48	56.25	75.52
50%WDT	0.4717	59.75	78.35	99.96	78.37	59.78	76.27

The particle size distributions of emitted PM from all conditions of combustions tested are shown in Figure 5. There are four main peaks standing for particle sizes of 0.16, 0.4, 1 and 4 μm. Case 100%BC releases more coarse particles in emission. Its PM_2.5/PM₁₀₀ is 34.87% (Table 8). However, after BC mixing with biomass, most of the particles in emission turn to fine particles, resulting in PM_2.5/PM₁₀₀ of 61.99–78.35 wt.%. The above findings are consistent with those of other studies pointed out that combustion of biomass results in an emission of fine particles with diameters smaller than 10 μm, which are harmful to health and cause environmental pollution (Zhang et al. 2011). A lower pellet size of solid fuel would increase the combustion efficiency because of better heat and mass transfers for smaller particles. However, it should be noted that the pellet size of coal and SBF used in the study were quite uniform (shown in Table 1). Hence, further experiments would be helpful to understand the effects of pellet size on particle size distribution of PM.

Figure 5. Particle size distribution of PM in flue gas in full-scale co-firing tests.

Co-firing the mixed fuels exhibit high percentiles of PM₁/PM₁₀₀ and PM₁/PM_2.5 of 44.36–59.75 wt.% and 69.21–76.27 wt.%, respectively, which are all higher than the sole combustion of coal. After blending with biomass, the potassium content in fly ash rises to 4.76 ~ 13.7 wt.% from 2.21 wt.% of sole coal firing, and the chlorine content also increases from 0.0672% to 0.38 ~ 1.82%. As a result, K and Cl contents in fly ash of co-firing with biomass are much higher than that from the sole firing of coal because of the presence of high Cl and alkali metal in the biomass. These components are also related to the formation of fine particles. Literature data show that the contents of potassium, chlorine and sulfur in the fuel influence the composition of the emitted inorganic submicron particles to a large extent. Potassium chloride and potassium sulfate were the main components exited in the emitted submicron particles (Johansson et al. 2003). Gaseous K₂SO₄ was found to be the main potassium compound vaporized and condensed to form PM₁, with an average size of 0.5 μm. When it came to the high combustion temperatures, gaseous KOH became the main potassium compound vaporized, most of which may react with aluminosilicates to form coarse particulates, and the remaining vaporized K₂SO₄ and other sulfates/oxides condensed to form PM₁, with an average size of 0.2–0.3 μm. The higher the combustion temperature, the more volatile potassium was transferred to coarse particulates (Zhou et al. 2010). According to these research results, high potassium amount in biomass lead to the considerable formation of fine particles.

In this research, the relatively higher contents of the alkali group (especially K) and chlorine in biomass and torrefied biomass than those of coal give rise to the formation of KCl and K₂SO₄, which are described above to be volatilized during combustion and then condensed to submicron-size particles. Consequently, there are more fine particles released in emissions. Moreover, the alkali group and chlorine contained in the produced fly ash may also react with gas chemicals and subsequently induce the formation of fine particles. As a result, the formation of the small size of particles also highly dependent on the K and Cl content in fly ash. For example, Case 50%WD_T has high PM₁/PM_2.5 of 76.27 wt.%. Hence, when using a substantial amount of biomass for co-firing, stringent control of emission of fine PM is necessary.

Comparisons with emission standards

Recently, Taiwan is dedicated to promoting the application of sustainable energy. Preparation of new legislation on the control of emissions from combustion of biomass in existed non-utility boilers is undergoing. Notice that the combustion of biomass in Germany has been widely applied and well developed for a long time. The regulations of emissions in Germany are also adopted along with Taiwanese restrictions to assess the co-firing results of this study. In Germany, for large-scale combustion, the regulations of emissions can be referred to “13. Bundes-Immissionsschutzgesetz (13. Blmschv)-Die Verordnung über Großfeuerungs- und Gasturbinenanlagen” (Bundesministerium der Justiz und für Verbraucherschutz 2018a). The emissions are limited according to different fuels (coals, biomasses, diesel and gases) that are utilized in boilers and the heat capacities (above 50 MW) of boilers. In Taiwan, the application of fossil fuels and fuels from biomasses for non-utility boilers has the same restriction of emissions. Unified restriction limits emissions from all kinds of non-utility boilers and fuels. Table 9 lists Taiwan’s and Germany’s emission standards. As for Germany’s standard, by separating coals and biomasses, NO₂ and SO₂ emissions of combustion of biomasses can be asked stricter than those of coal. CO formation depends highly on the efficiency of combustion. As a result, asking low CO emission (150 ppmv) is also asking high energy conversion from fuels. Taiwan also has special regulations for special boilers such as steam and electricity symbiosis equipment boilers which are used by utility and power companies to generate electricity (noted as utility boilers). The utility boilers are restricted by stricter regulation than the conventional non-utility boilers (classified as stationary sources) also shown in Table 9.

Table 9. Comparison of air pollution standards of gas emissions.

Item	Unit	Taiwan 1 (ST1)^a	Taiwan 2 (ST2)^b	Germany 1 (SG1)^c (Solid coals, 50–100 MW)	Germany 2 (SG2)^c (Solid biomasses, 50–100 MW)
PM	mg/m^3	20-50	50	10	10
CO	ppmv	-	2000	150	150
NOX	ppmv	70-175	350	300	250
SO2	ppmv	60-150	300	400	200
HCl	ppmv	-	80	-	-

Notes. ^aEmission standards of steam and electricity symbiosis equipment boilers in Taiwan (TEPA, 2013b);

^bEmission standards of stationary sources in Taiwan (TEPA 2013a);

^c“13. Bundes-Immissionsschutzgesetz (13. Blmschv)-Die Verordnung über Großfeuerungs- und Gasturbinenanlagen” (Bundesministerium der Justiz und fürVerbraucherschutz, 2018a).

The utilization of different solid fuels in small-scale boilers is also prevalent in Germany. The restriction of emissions is presented in another regulation “1. Bundes-Immissionsschutzgesetz (1. Blmschv)-Die Verordnung über kleine und mittlere Feuerungsanlagen” (Bundesministerium der Justiz und für Verbraucherschutz 2018b). The small boilers which are adopted in this law have only kilowatts (kW) level of heat capacity. The fuels also have a specific content limitation in standard “DI.” Thus, the regulation limits not only amounts of emissions of pollutants but also the contents of fuels in order to reduce air pollutants effectively. Moreover, different conditions of fuels and boilers are given different limitations of emissions like “13. Blmschv”, which provides more appropriate limitations for the application of different fuels. These are good aspects to which Taiwan can try to refer.

Comparison of the results of Tables 7–8 with the emission standards of Table 9 indicates the follows. The emissions of CO meet the limitation of Taiwan standard ST2 for non-utility boilers while exceeding Germany regulations SG1 and SG2 for large-scale combustion. Other emissions of NOx, SO₂, HCl and PM are all satisfied with the corresponding regulation, and this thus elucidates the appropriation and applicability of the use of biomasses tested in the present study for co-firing with coal.

Conclusion

Addition of biomass of RS (with R_BL = 5-10%), WD (2-15%), RS_T (2-10%) and WD_T (2-20%), respectively, to co-fire with coal generally results in a decrease of NOx and SO₂ while increase of CO (except RS_T) emissions in exhausted gas. Fixed carbon contents in biomass-coal mixtures and Ca contents in fly ashes are highly related to NOx and SO₂ emissions, respectively. After torrefaction, the reductions of emissions of CO and SO₂ from co-firing of RS_T are significant, while those of WD_T are not apparent.

Adding biomass for co-firing gets rise to more fine particulate matters in emissions, especially PM_2.5 particles. After torrefaction, co-firing of RS_T has a higher emission of PM_2.5 than RS, while co-firing of WD_T has no noticeable difference of emission of PM_2.5 with WD. Nevertheless, all examined cases of coal co-firing, respectively, with RS, WD, RS_T and WD_T meet the emission standards of stationary sources in Taiwan.

Nomenclature

BC: bituminous coal

BF: bio-fiber (extracted from MSWs)

H_HD: dry-basis high heating value (kcal/kg)

MSWs: municipal solid wastes

PM₁, PM_2.5, PM10, PM100: particulate matters with sizes ≤ 1, 2.5, 10, 100 μm, respectively

R_BL: blending ratio of biomass in mixture of coal and biomass based on heating values (kcal/kcal)

RB: rice straw bio-char

RS: rice straw

RS_T: torrefied RS

SBF: solid bio-fuel

SF: solid fuel

SG1: Emission standards for large-scale combustion (solid coals, 50-100 MW) in Germany

SG2: Emission standards for large-scale combustion (solid biomasses, 50-100 MW) in Germany

ST1: Emission standards of steam and electricity symbiosis equipment boilers in Taiwan

ST2: Emission standards of stationary sources in Taiwan

WB: wood bio-char

WD: wood

WD_T: torrefied WD

y%SBF for SBF = RS, WD, RS_T, WD_T: (100-y) kcal % BC co-firing with y kcal % SBF

100% BC: 100%BC firing

Additional information

Funding

This work was supported by the Ministry of Science and Technology and Environmental Protection Administration, Taiwan.

Notes on contributors

Chia-Chi Chang

Chia-Chi Chang obtained his Ph.D. from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Taiwan. Currently, he is a research scientist in Taiwan Bio-energy Technology Development Association, Rm. 312, No. 130, Sec. 3, Keelung Rd., Taipei 106, Republic of China. He can be contacted at [email protected].

Yen-Hau Chen

Yen-Hau Chen obtained his Master degree from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. Currently, he is a Ph.D. student in Institute of Combustion and Power Plant Technology, University of Stuttgart, Pfaffenwaldring 23, Stuttgart 70569, Germany. He can be contacted at [email protected].

Wei-Ren Chang

Wei-Ren Chang obtained his Master degree from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Chao-Hsiung Wu

Chao-Hsiung Wu is a Professor at Department of Environmental Engineering, Da-Yeh University, No. 168, University Rd., Dacun, Changhua 515, Republic of China. He can be contacted at [email protected].

Yi-Hung Chen

Yi-Hung Chen is a Professor at Department of Chemical Engineering and Biotechnology, National Taipei University of Technology, No. 1, Sec. 3, Zhongxiao E. Rd., Taipei 106, Republic of China. He can be contacted at [email protected].

Ching-Yuan Chang

Ching-Yuan Chang is a Distinguished Professor at Graduate Institute of Environmental Engineering and Department of Chemical Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Min-Hao Yuan

Min-Hao Yuan is an Assistant Professor at Department of Occupational Safety and Health, China Medical University, No. 91 Hsueh-Shih Road, Taichung 404, Republic of China. He can be contacted at [email protected].

Je-Lueng Shie

Je-Lueng Shie is a Distinguished Professor at Department of Environmental Engineering, National I-Lan University, No.1, Sec. 1, Shennong Rd., Yilan City, Yilan County 260, Republic of China. He can be contacted at [email protected].

Yuan-Shen Li

Yuan-Shen Li is a Professor at Department of Environmental Engineering, National I-Lan University, No.1, Sec. 1, Shennong Rd., Yilan City, Yilan County 260, Republic of China. He can be contacted at [email protected].

Sheng-Wei Chiang

Sheng-Wei Chiang obtained a Ph.D. from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Tzu-Yi Yang

Tzu-Yi Yang obtained her Master degree from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. She can be contacted at [email protected].

Far-Ching Lin

Far-Ching Lin is an Associated Professor at School of Forestry and Resource Conservation, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Chun-Han Ko

Chun-Han Ko is a Professor at School of Forestry and Resource Conservation, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Bo-Liang Liu

Bo-Liang Liu obtained his Master degree from Graduate Institute of Environmental Engineering, National Taiwan University, No.1, Sec. 4, Roosevelt Road, Taipei 106, Republic of China. He can be contacted at [email protected].

Kuang-Wei Liu

Kuang-Wei Liu is working in Environmental Analysis Laboratory, Environmental Protection Administration, No. 260, Minzu Rd. Sec 3, Taoyuan 320, Republic of China. He can be contacted at [email protected].

Shi-Guan Wang

Shi-Guan Wang is working in Environmental Analysis Laboratory, Environmental Protection Administration, No. 260, Minzu Rd. Sec 3, Taoyuan 320, Republic of China. He can be contacted at [email protected].