Mineral and Heavy Metal Composition of Oil Shale Ash from Oxyfuel Combustion.

Oxyfuel combustion can reduce CO2 emissions from fossil fuels. Hence, it is currently being investigated for potential use in oil shale-fired power plants, which currently produce most of Estonia's electricity. Here, experiments were performed with kukersite oil shale for both oxyfuel and conventional combustion in a 60 kWth circulating fluidized bed combustor. In this paper, we provide data on the ash composition including mineral compositions and heavy metal concentrations. Oxyfuel conditions did not noticeably influence the concentrations of heavy metals in the ash but did have significantly lower amounts of free lime because of inhibition of the carbonate decomposition reactions. The results suggest that oxyfuel combustion would produce no significant problems in terms of the behavior of the ash or the fate of heavy metals contained in the ash.

Introduction

To avoid negative climatic effects caused by global warming, the amount of anthropogenic CO2 emitted to the atmosphere must be reduced.[1,2] However, currently the majority of the world’s energy needs are still supplied by fossil fuels. In Estonia, for instance, most of its electricity is produced by oil shale-fired power plants.[3] One way to reduce CO2 emissions from such power plants is to use pure oxygen along with recirculated flue gas in the combustion instead of air. This type of combustion produces flue gas with a high concentration of CO2, which can then be more easily captured and stored to reduce emissions.[4]

Additional advantages of such technology are that it offers the possibility of cofiring biomass if circulating fluidized bed combustion (CFBC) technology is used, further reducing anthropogenic CO2 production, without adversely affecting heavy metal emissions[5] or even the possibility of net reductions of anthropogenic CO2 via the so-called bioenergy with carbon capture and storage.[6] However, for oxyfuel combustion to be used industrially, its feasibility must be demonstrated. Many of the current obstacles are economic. For example, the price of CO2 (specifically European CO2 emission allowances) is too low and there is a lack of government support and investment.[6,7] This paper focuses on the various technological and environmental challenges that must first be addressed.

Earlier research has already investigated many aspects of oxyfuel combustion with oil shale.[8−10] Thermogravimetric analysis experiment has also been performed to better understand the kinetics of oil shale combustion in the CO2-rich oxyfuel atmosphere.[11] Small laboratory reactor experiments have also been performed,[12−14] and in some of these experiments, it was observed that oxyfuel combustion reduces the amount of free lime in the ash.[12] Yörük et al.[15] have also used a process simulator to model an oxyfuel combustion process with oil shale. In addition, our research group has performed oxyfuel combustion experiments with a larger 60 kWth CFB combustor,[10] and here, we present new data from these CFB experiments to investigate the effect of oxyfuel combustion on the resulting ash composition.

Oil shale contains an organic material, called kerogen, dispersed in a mineral matrix,[11] and thus, oil shale produces large amounts of ash upon combustion. For instance, Estonian kukersite oil shale contains around 65–70% inorganic matter and also contains various trace elements (Table ). The carbonates in the oil shale generally decompose on combustion, but at least 50 wt % of the oil shale remains as ash (see analysis in the Methods section). Oil shale ash was considered for a long time to be a hazardous waste; therefore, almost all of the produced ash was deposited in nearby ash fields.[16,17] Some fly ash is also emitted to the atmosphere with the flue gas.[18]

Table 1

Typical Heavy Metal Content of Oil Shale and Comparison with Coal and Limits Given in Estonian Law for Residential Soila,b

	kukersite oil shale	kukersite oil shale	kukersite oil shale (OS1)	coal (average)	coal (range)	suggested conc. limitsc	permitted conc. limitsc
refs	(19)	(20)	this work	(19)	(19)	(21)	(21)
As	7.6–21	21	6.5	10	0.5–80	20	30
Ba	1150–1500			200	20–1000	500	750
Be			1.7
Cd	0.4–4		ND	0.5	0.1–3	1	5
Co	2.9–3.0	3	3.1	5	0.5–30	20	50
Cr	17–38	17	22.1	20	0.5–60	100	300
Cu	17–55	7.5	10.6	15	0.5–50	100	150
Hg	0.08–0.3	0.22		0.1	0.02–1	0.5	2
Mn	310–387	340	368.4	70	5–300
Mo	3		1.1	3	0.1–10
Ni	12–21	15	15.4	20	0.5–50	50	150
Pb	20–30	30	31.1	40	2–80	50	300
Sb	0.5–0.6	<0.4	ND	1	0.05–10	10	20
Se			ND
Sr			160.4
Th	2.3–3.4			4	0.5–10
Ti			1162.5
Tl		0.5	<0.5			1	5
U	3			2	0.5–10	1	5
V		24	20.5			50	300
Zn			49.4

Unit: μg g–1.

ND = not detected.

Limits for residential soil prescribed in Estonian law.

Oil shale ash can potentially produce toxic effects. Blinova et al.[22] noted that aqueous eluates from fly ash were toxic to any aquatic species. They further noted that the toxicity was mainly related to the high alkalinity caused by the ash high free lime content (which hydrates in aqueous solution and is then neutralized by atmospheric CO2); however, “the toxic impact of trace elements such as As and Pb cannot be excluded”.[22] The fly ash emitted to the atmosphere can increase the concentration of heavy metals in the soil where it settles. However, Blinova et al.[22] did not observe any long-term soil contamination in the region near the power plants, at least not for As, Cd, Cu, Pb, or Zn. Additionally, Velts et al.[23] concluded that heavy metals in the oil shale ash have low leachability, and experimental results from Blinova et al.[22] showed that some trace elements were not bioavailable. Finally, it is worth noting that the concentration of most of the trace elements is still of the same order of magnitude as Estonian soil. Nevertheless, the ash does have a potential environmental impact, and if oxyfuel combustion affects the composition of the ash, the resulting environmental effects would weigh on the decision of whether or not to use the technology industrially.

An example of this can be seen when the power plants in Estonia started to replace older pulverized combustion (PC) boilers with newer CFBC boilers.[24,25] Both Blinova et al.[22] and Reinik et al.[26] compared ashes from the two current combustion technologies and concluded that fly ash from the CFBC boilers was less toxic than PC fly ash. Moreover, CFBC ash had lower concentrations of heavy metals, on average presumably because of the fact that the carbonate content was higher. Here, we investigate what effect oxyfuel combustion in a CFBC boiler has on the composition of the ash, with emphasis on the content of heavy metals in the ash.

Results and Discussion

Changes in Composition

From the chemical composition data in Table we can conclude that the amount of calcium and magnesium found in overhead zones decreased in the ash, and the amount of sulfate in the ash correspondingly decreased. However, the concentration of other elements, such as silicon, and metals, such as iron, aluminum, and titanium, increased. In addition, ashes from the overhead zones in the boiler had lower losses on ignition (LOIs).

Table 2

Chemical Composition of the Ashes from the Oxyfuel Experimentsa

	BA	EHE	C1	C2	FA
SiO2	10.61	20.79	31.09	30.76	36.21
Al2O3	2.50	4.92	7.48	7.41	9.38
TiO2	0.13	0.28	0.45	0.45	0.58
Fe2O3	2.23	3.75	4.68	4.72	5.53
MnO	0.069	0.073	0.052	0.053	0.046
CaO	39.00	36.77	27.39	27.98	24.86
MgO	10.37	9.56	6.74	6.82	5.53
Na2O	0.03	0.05	0.17	0.20	0.65
K2O	0.64	1.70	3.18	3.02	3.42
P2O5	0.093	0.121	0.130	0.137	0.158
SO3	8.71	7.29	6.00	5.12	4.69
LOI	25.59	14.65	12.59	13.28	8.87

Unit: wt %.

Ash minerals were also determined from the X-ray diffraction (XRD) data, and the results are given in Table . The oxyfuel ashes still contain large quantities of calcite. At higher temperatures, such as those in the boiler, calcite generally decomposes to form CO2 and lime. However, because oxyfuel combustion takes place in a CO2 atmosphere, the reaction equilibrium is shifted, and decomposition occurs at a higher temperature.[12] This explains the high concentrations of calcite in the oxyfuel ash. Our earlier experiments in a batch reactor showed a similar difference between oxyfuel and normal combustion ashes.[12] Dolomite generally decomposes into periclase and lime and reacts to form other Ca–Mg silicate phases.[12] However, interestingly, unlike the batch experiments, the CFBC oxyfuel ashes still contained some dolomite. Anthony et al.[27] have observed that dolomite can still be present in the ash of large demonstrations of pressurized FB systems, even though the temperature should be high enough that it should decompose.

Table 3

Mineral Composition of the Ashes from the 20 April 2017 Oxyfuel Experimenta

	BA	EHE	C1	C2	FA
quartz	8.4	16.0	25.4	25.3	28.6
K-feldspar	3.0	7.9	18.6	18.7	21.9
K-mica	0.5	0.8	0.7	0.6	0.7
calcite	44.2	24.6	20.1	20.4	16.6
dolomite	9.8	10.0	2.9	2.2	1.0
lime	4.7	8.1	4.3	4.6	4.9
periclase	6.4	6.8	5.2	5.1	4.3
anhydrite	11.5	11.0	8.6	7.4	7.2
C2S β	4.3	4.2	4.6	4.9	3.9
merwinite	3.2	4.0	3.6	4.0	4.3
akermanite	2.6	4.4	2.5	3.0	2.0
hematite	1.3	1.8	2.9	3.0	3.6
wollastonite		trace	0.5	0.7	0.9

Unit: wt %.

We can compare the results to the mineral composition of industrial oil shale ashes. Table gives data we measured, using the same XRD method, for various oil shale ashes from the Auvere power plant in Estonia. This power plant has a CFB boiler and uses kukersite oil shale as the fuel. The industrial ashes have higher lime contents, which again shows that the CO2-rich atmosphere in oxyfuel combustion effectively inhibits calcite decomposition.[12] Further, dolomite was only identified in significant quantities in the bottom ash, whereas in the oxyfuel ashes, the dolomite concentration was always higher. The lower content of free lime would help reduce expansive behavior in landfills[28] and could help decrease the environmental impact by reducing the pH of the ash leachate.[23,29]

Table 4

Mineral Composition of Ashes from the Industrial Auvere Power Planta

	bottom ash	cyclone	economizer	air preheater	ESP1	ESP2	ESP3	ESP4	ESP5	baghouse
quartz	3.4	5.1	14.9	12.4	12.3	10.4	9.8	8.2	7.8	1.4
K-feldspar	2.0	5.4	18.2	11.4	16.9	15.6	14.6	13.2	14.2	6.2
K-mica	2.8		5.8	2.9	6.2	6.2	6.4	5.0	6.6	3.8
calcite	26.9	0.8	12.6	10.4	8.5	9.2	9.7	10.0	10.2	20.7
dolomite	6.7			0.5
lime	23.5	32.0	17.4	21.3	20.4	21.5	21.2	23.1	21.8	23.8
periclase	8.5	8.4	6.0	8.5	6.8	6.5	6.6	6.7	6.5	6.0
anhydrite	12.2	22.7	7.3	12.2	7.7	7.5	7.7	8.2	7.8	10.8
C2S β	5.0	11.6	7.0	8.1	8.6	9.8	10.3	10.6	10.1	11.8
merwinite	4.1	5.7	6.2	6.1	7.1	8.3	8.6	9.3	9.4	11.4
akermanite	4.4	7.4	2.7	4.7	3.7	3.5	3.2	3.5	3.4	2.5
hematite	0.5	0.5	1.5	1.1	1.3	1.0	1.1	1.3	1.4	trace

Unit: wt %.

Heavy Metals in the Ash

The ashes from the normal and oxyfuel experiments can be compared to see if oxyfuel combustion might cause any differences in the ash composition. The experimental data are given in Table , and Figures and 2 provide a visual comparison for the bottom ash and external heat exchanger ash samples.

Table 5

Concentration of Heavy Metals and Other Trace Elements in the Oil Shale Ash from the Experimental 60 kW CFBCa,b

ash type	date	atmos.	As	Ba	Be	Cd	Co	Cr	Cu	Mn	Mo	Ni	Pb	Rb	Sb	Se	Sr	Ti	Tl	V	Zn	Zr
BA	12.07.2016	air	8.2		1.2	ND	2.3	12.3	4.2	677.1	0.5	9.9	24.8		ND	<0.5	236.6	664.8	ND	12.6	78.7
EHE	12.07.2016	air	8.8		1.3	ND	3.0	16.7	3.6	642.0	1.1	12.3	31.9		ND	<0.5	237.2	959.0	ND	17.5	90.8
BA	12.07.2016	oxyfuel 21%	7.2		1.2	ND	2.0	9.7	14.2	662.6	0.2	7.7	14.2		ND	ND	211.6	560.7	ND	10.5	37.1
EHE	12.07.2016	oxyfuel 21%	8.9		1.3	ND	3.0	18.0	3.8	631.1	1.2	12.9	23.4		ND	ND	231.8	1004.3	ND	17.7	45.5
C1	12.07.2016	oxyfuel 21%	12.4		2.5	<3.0	6.6	50.3	8.4	481.2	4.1	31.2	67.5		ND	0.7	291.4	2668.4	1.2	44.4	43.4
C2	12.07.2016	oxyfuel 21%	14.3		2.0	<3.0	7.8	59.0	9.8	413.8	5.1	38.5	87.6		<0.5	0.6	313.9	3092.6	1.8	54.1	88.9
FA	12.07.2016	oxyfuel 21%	14.5		1.6	<3.1	8.2	63.2	9.6	392.7	5.1	40.0	89.8		<0.5	<0.5	315.1	3163.9	1.9	55.2	43.2
BA	20.04.2017	oxyfuel 30%	11.5	264	1.3	ND	4.1	26.5	4.9	714.4	2.5	18.6	35.3	33	ND	<0.5	297.0	1511.2	ND	26.5	48.7	46
EHE	20.04.2017	oxyfuel 30%	12.4	254	1.6	<2.8	5.9	37.5	9.8	643.1	3.9	26.9	47.8	71	<0.5	0.6	292.8	2188.9	ND	36.6	45.4	77
C1	20.04.2017	oxyfuel 30%	14.4	124	2.4	ND	8.1	62.4	9.9	465.4	5.6	42.3	74.4	88	ND	0.7	334.6	3280.2	1.6	53.3	34.2	96
C2	20.04.2017	oxyfuel 30%	13.5	162	2.2	<2.9	8.4	75.4	10.7	495.2	6.5	48.8	82.7	98	<0.5	0.8	330.2	3351.8	1.6	54.6	48.5	106
FA	20.04.2017	oxyfuel 30%	17.4	180	2.7	<2.9	10.1	76.9	12.7	425.0	7.2	51.1	113.2	129	<0.5	1.2	385.7	4100.1	3.1	70.8	47.4	121
EHE	11.01.2017	air	12.1		1.4	ND	5.2	33.5	20.9	755.0	3.8	25.5	44.9		ND	0.6	310.3	1910.9	ND	33.0	47.3
BA	11.01.2017	air	13.4		1.3	ND	3.9	19.5	4.3	996.5	2.3	18.3	41.1		ND	<0.5	338.5	1149.7	ND	20.5	22.1
BA	12.01.2017	air	8.0		1.7	ND	2.5	12.8	3.4	753.8	0.7	10.2	24.0		ND	ND	290.8	763.1	ND	14.3	14.9

Unit: ppm.

ND = not detected. Ba, Rb, and Zr were measured using XRF.

Figure 1

Comparison of the composition of bottom ash from normal and oxyfuel experiments.

Figure 2

Comparison of the composition of ash from the external heat exchanger from normal and oxyfuel experiments.

Although some samples did have higher concentrations of heavy metals, our data indicate that the differences are not related to the oxyfuel conditions. At the very least, we would expect that the increased carbonate content in the oxyfuel ash would lead to somewhat lower heavy metal concentrations. As shown in the mineralogical data, oxyfuel combustion often reduces the amount of calcite and dolomite that decompose, and this would lead to an overall increase in the mass of the ash. The larger total mass would be expected to make the relative mass of heavy metals lower, even if the total amount of heavy metals remains the same. However, such an influence was not observed.

Indeed, one of the two samples with higher metal concentrations was from an oxyfuel experiment (oxyfuel with 30% O2), but the other was from combustion with air (experiment on 11 January 2017). The other oxyfuel sample (21% O2) had some of the lowest metal concentrations. Therefore, the differences must have been related to differences in how each CFBC experiment progressed. Differences in fuel composition affect ash composition, but that does not seem to be a major factor here because even experiments that used the same oil shale (those on 11 January 2017 and 12 January 2017) produced bottom ashes with different compositions.

If oxyfuel combustion does influence the concentration of heavy metals in the ash, then the effect is obscured by other factors. This suggests that oxyfuel conditions are not as significant as other parameters in the operation of the boiler in determining the concentration of heavy metals in the ash. Aunela-Tapola et al. noted that the ash composition can vary significantly, and they took multiple samples from the same locations in the Eesti power plant, and the standard deviations of the heavy metal concentrations were relatively large.

The heavy metal concentrations in the ash from our experimental CFB lie in the same range as oil shale ash from industrial power plants in Estonia (see the experimental and literature data in Tables and 6, respectively).[20,22,23,26] Thus, it seems reasonable that switching to oxyfuel in an industrial boiler would have little effect on the distribution of heavy metals in the ash.

Table 6

Literature Data on the Concentration of Heavy Metals and Other Trace Elements in Kukersite Oil Shale Asha

ash type	plants	refs	As	Ba	Cd	Co	Cr	Cu	Hg	Mn	Mo	Ni	Pb	Rb	Sb	Sn	Sr	Th	Tl	U	V	Zn
fly ash	Eesti–PC	(22)	48.9		0.94		51	20.18					193.79									179.1
fly ash	Eesti–CFBC	(22)	25.8		0.13		48	18.56					67.23									50.6
fly ash	Balti–PC	(22)	28.2		0.26		42	12.37					112.62									45.9
fly ash	Balti–CFBC	(22)	17.2		0.17		49	15.29					75.49									61.7
bottom ash	Eesti–PC	(20)	16			4.5	19	9.9		700		27	24		<0.4				<0.1		33
superheater	Eesti–PC	(20)	18			4.9	23	11		700		29	44		<0.8				<0.2		40
economizer	Eesti–PC	(20)	14			4.5	19	9.5		690		26	34		<0.6				<0.02		35
cyclone	Eesti–PC	(20)	16			4.5	21	9.3		650		26	45		0.6				0.3		38
ESP I and II	Eesti–PC	(20)	42			5.3	33	9.6		470		31	130		0.9				1.3		52
ESP III and IV	Eesti–PC	(20)	59			6.6	49	12		440		38	210		1.1				2.3		73
fly ash > 4–6 μm	Eesti–PC	(20)	68			6.6	58	13	0.1	350		38	200		2				2.7		81
fly ash < 4–6 μm	Eesti–PC	(20)	92			7.4	48	18	0.3	340		45	380		2.4				7.5		210
bottom ash	Balti–PC	(26)		107	0.1	3	17	7.9	<0.02	573	2.7	17		16	<0.04	<0.2	252	2.8		2.5	24	63
super-heater	Balti–PC	(26)		143	0.05	3.5	21	9.8	<0.02	463	3.1	18		30	<0.04	<0.2	228	3.3		2.7	29	56
pre-heater	Balti–PC	(26)		132	0.08	5.6	38	9.2	<0.02	373	2.3	28		66	<0.04	<0.2	245	4.8		3	45	38
cyclone	Balti–PC	(26)		154	0.2	3.5	19	36	<0.02	545	2.7	20		18	<0.04	<0.2	265	3.1		2.9	29	46
ESP I	Balti–PC	(26)		179	0.2	3.7	37	7.8	<0.02	373	6.4	22		50	<0.04	0.2	208	3.2		3.1	38	82
ESP II	Balti–PC	(26)		280	0.6	5.8	59	15	<0.02	420	8.4	33		98	<0.04	0.7	204	5.2		4.4	60	142
ESP III	Balti–PC	(26)		242	0.9	5	51	12	0.03	310	8.3	28		100	<0.04	0.3	168	4.2		4.6	80	143
bottom ash	Balti–CFBC	(26)		59	0.06	3.7	20	8	<0.02	826	3.5	21		22	<0.04	0.2	346	3.6		2.4	24	41
INTREX	Balti–CFBC	(26)		87	0.07	4.3	29	7.4	<0.02	875	3.6	25		28	<0.04	0.2	279	5.3		2.8	32	51
ESP I	Balti–CFBC	(26)		199	0.07	8	54	12	0.08	524	3.3	40		96	<0.04	0.2	354	6.5		4.3	67	56
ESP II	Balti–CFBC	(26)		125	0.06	3.4	20	7.7	<0.02	606	2.8	20		21	<0.04	<0.2	270	3.2		3	28	38
ESP III	Balti–CFBC	(26)		207	0.07	8.9	60	14	0.15	521	3.3	44		91	<0.04	<0.2	401	7.4		4.9	71	55
ESP IV	Balti–CFBC	(26)		180	0.06	7.5	52	14	0.12	429	3.6	41		69	<0.04	<0.2	371	6.1		4.3	62	53
bottom ash	2nd block–PC	(23)					25			545			13								20	155
mix	PC	(23)					35			600			38								30	120
ESP I	8th block–CFBC	(23)					45			385			30								38	∼30
mix	CFBC	(23)					40			445			20								30	∼30

Unit: ppm.

Other studies have observed that oil shale ash obtained further from the boiler itself, is enriched in heavy metals.[20,30] Our data from the oxyfuel experiments show the same trend (see Table ). This increase correlates with the increase in other elements, such as iron, aluminum, and silicon, that is shown in Table . One exception, both in our data and the literature data,[20] was for manganese, for which the concentration was actually lower in the filter ash than in the bottom ash.

Because the concentration of heavy metals generally increases in ashes further from the boiler, we also measured the concentrations in the fly ash. The fly ash samples were collected from the flue gas after the fabric filter. Two samples were collected during the 20 April 2017 experiments: one during normal combustion with air and a second during oxyfuel combustion. However, only a few milligrams of the ash could be obtained during the course of the experiments, and hence, our inductively coupled plasma mass spectrometry (ICP–MS) method could not provide reliable results. We were, however, able to measure the size distributions of the same fly ashes. These data were presented earlier,[31] but here, we show the size distribution again for convenience in Figure . Figure shows that the oxyfuel fly ash had a distribution that was shifted toward larger particle sizes. The median particle diameter for air combustion was 34 nm, and for oxyfuel, it was 42 nm. However, perhaps even more importantly, the oxyfuel experiment produced less fly ash in the flue gas: 1.0 mg m–3 (dry basis) compared to 1.6 mg m–3 during normal combustion. Even if the concentration of heavy metals in the oxyfuel fly ash is around the same as with normal combustion, if the amount of fly ash emitted is lower overall, then the environmental impact would most likely be lower.

Figure 3

Distribution of particles by size in the fly ash during normal and oxyfuel combustion.

Conclusions

Oxyfuel combustion conditions did not produce a noticeable effect on the concentrations of heavy metals in the ash. As the concentrations of heavy metals in our ash samples aligned with those of ashes from industrial boilers, we expect that switching to an oxyfuel atmosphere would not significantly affect the concentrations of heavy metals in ashes from industrial power plants. Additionally, in our experiments, flue gas during oxyfuel combustion contained fewer and larger ash particles. If a similar reduction occurs in an industrial boiler, then this would decrease the health impact of the plant.

The high concentration of CO2 in the oxyfuel atmosphere reduced the extent of decomposition of calcite and dolomite. As a result, ashes from oxyfuel combustion can be expected to have less free lime and this will reduce expansive behavior in landfills. The reduction in free lime will help lower the pH of ash leachate, which in turn ensures that the oxyfuel ash will have a lower environmental impact.

Methods

Oil Shale Samples

Kukersite oil shale from Estonia was used as the fuel. The fuel was dried, crushed, and separated to obtain oil shale with a size less than 3 mm. Table gives the proximate and ultimate analysis of the oil shale samples used. Proximate analysis was performed on the sample as it was received, and the ultimate analysis is given in terms of the dry sample. The data for sample OS3 was published earlier by Loo et al.[9]

Table 7

Proximate and Ultimate Analysis of the Oil Shale Samples Used

	OS1	OS2	OS3
LHVa (MJ kg–1)	8.82	9.83	8.56
moisture	9.0	0.2	0.5
volatile matter		49.2	47.5
fixed carbon		1.6	1.3
ash	50.7	49.0	50.7
C	27.3	28.6	27.4
N	0.07	0.07	0.07
S	1.7	1.6	1.6
H	2.7	2.8	2.7
total organic carbon	21.3	23.0	21.8
CO2 (mineral)	21.7	20.2	20.6

Lower heating value.

CFBC Experiments

The ashes were obtained from combustion experiments in a 60 kWth CFB combustor. Several experiments were carried out using either normal air-fired combustion or oxyfuel combustion with a mixture of CO2 and O2. The CO2 and O2 were taken from commercially available gas cylinders. The experimental CFB setup is shown in Figure . Table gives an overview of the experiments performed, and Figure shows the locations from which ash samples were obtained. Here, Table gives the distribution of the ash by mass fraction at different locations in the CFB. The experimental CFB and the combustion experiments are described in more detail elsewhere.[9]

Figure 4

Schematic of the 60 kW circulating fluidized bed combustor used for the experiments.

Table 8

Overview of the CFBC Experiments Performed

date	atmosphere	oil shale
12 July 2016	oxyfuel 21% O2	OS1
12 July 2016	air	OS1
11 January 2017	air	OS2
12 January 2017	air	OS2
20 April 2017	oxyfuel 30% O2	OS3
20 Apri; 2017	air	OS3

Table 9

Distribution of Ash among the Different Locations in the Experimental CFBC (See Figure )

	description of ash	mass fraction of total ash
BA	bottom ash	37
EHE	from external heat exchanger	7
C1	from 1st cooler	2
C2	from 2nd cooler	47
FA	from fabric filter	7

ICP–MS Measurements

Trace elements were quantified using Thermo iCAP Qc Quadrupole ICP–MS. Digestion of the samples (along with the filters) was done using concentrated HNO3 and HF. The samples were digested in sealable Teflon containers in a microwave oven. The analysis was run in kinetic energy discrimination mode with high-purity helium in order to eliminate any interferences.

XRF–XRD Measurements

The mineral composition of some of the oxyfuel ashes was investigated using XRD. For comparison, some oil shale ashes from the industrial Auvere power plant were examined using the same technique. A Bruker D8 diffractometer was used, which was fitted with a Lynx-Eye linear detector. Experiments were performed using Cu Kα radiation in the 2θ range of 3–72°, with a step size of 0.02° 2θ and a counting time of 0.1 s per step. The X-ray tube was operated at 40 kV and 40 mA. For analysis, the Siroquant 3.0 code was used, which is based on the Rietveld algorithm.[32] The presence of portlandite indicates that the samples had come into contact with moisture in the atmosphere, which caused hydration of some of the lime. Therefore, the compositions were recalculated to include the portlandite in the lime concentration.

X-ray fluorescence (XRF) spectrometry was used to measure the chemical composition of these ashes. Here, a Rigaku Primus II XRF spectrometer was used.

Size Distribution Measurements

The particle size measurements are already described in our earlier work,[31] but we describe the method here as well because we rely on some of these results in the discussion section. The fly ash analyzed by the ELPI device came from both air and oxyfuel experiments conducted on 20 April 2017 (Figure ).

Figure 5

Setup used for the ELPI measurements.

Particle sizes were measured using an electrical low-pressure impactor (ELPI+) which was produced by Dekati. The ELPI measurement technique was developed by Keskinen et al.,[33] and it measures the number concentration as a function of the aerodynamic particle diameter. In the device, the aerosol particles are first charged with a corona charger and then, they enter a low-pressure cascade impactor. The electric current produced by the charge carried by the impacted particles is measured by electrometers as the particles impact electrically isolated collection stages, and this signal can then be used to calculate the number of particles with different sizes. The measurements were carried out during the combustion experiments, so the ELPI provided near-real time data.

The sampling location for the ELPI device was after the fabric filter. A portion of the flue gas was removed and sent to a diluter. This gas was then sent to both the CO2 analyzer and the ELPI+ device. The measurement setup is shown in Figure .