SO₂ Emissions from Oil Shale Oxyfuel Combustion in a 60 kWth Circulating Fluidized Bed

Abstract

Carbon capture, utilization, and storage (CCUS) have emerged as pivotal technologies for curtailing emissions while maintaining fossil fuel. Estonia faces a challenge due to its dependence on carbon-intensive oil shale, but the need for energy security, highlighted by the war in Ukraine, makes reducing CO₂ emissions a priority while maintaining energy independence. In this context, the presented study determines the environmental impact of combustion of the Estonian oil shale from the release of SO₂ emission and compares sulfur retention in the ash between different oxyfuel combustion campaigns in a 60 kWth CFB test facility. The pilot was operated under air, O₂/CO₂, and with recycled flue gas (RFG), and we tested the application of extremely high inlet O₂ up to 87%_vol. The key objective of this study is to examine how different combustion atmospheres, operating temperatures, and excess oxygen ratios influence SO₂ formation. Additionally, the research focuses on analyzing anhydrite (CaSO₄), calcite (CaCO₃), and lime (CaO) in ash samples collected from the dense bed region (bottom ash) and the external heat exchanger (circulating ash). The results indicate that increased inlet O₂% does not significantly affect SO₂ emissions. Compared to air-firing, SO₂ emissions were higher than 40 mg/MJ under a 21/79%_vol O₂/CO₂ environment but were significantly reduced, approaching zero, as the inlet O₂% increased to 50%. Under O₂/RFG conditions, higher SO₂ concentrations led to increased sulfur retention in both the bottom and circulating ash. The optimal temperature for sulfur retention in air and oxyfuel combustions is below 850 °C. This study for the first time provides a technical model and discusses the effects of operating parameters on sulfur emissions of the Estonian oil shale CFB oxyfuel combustion.

1. Introduction

Combustion of fossil fuels remains a major means of satisfying the world’s rising energy needs. The fossil fuel combustion process is associated with the release of greenhouse gases (GHGs) and air pollutants (such as CO₂, SO₂, NO_x, etc.). In response to the growing demand for strategies to reduce CO₂ emissions, carbon capture, utilization, and storage (CCUS) has gained significant attention [1]. Unlike the direct replacement of fossil fuels, CCUS is the only technology that enables the continued use of fossil fuels while significantly lowering greenhouse gas emissions. Oxyfuel circulating fluidized bed (CFB) combustion is among the most promising CO₂ abatement technologies for cleaner combustion [2,3,4,5,6]. By substituting air with a mixture of high-purity oxygen and flue gas, the process becomes nitrogen-lean, resulting in flue gas that is CO₂-rich and ready for carbon capture and storage. This affects the rate of combustion, combustion of organic matter, reactions of mineral matter, heat transfer, boiler hydrodynamics, and the behavior of ash (i.e., CaCO₃–CaO decomposition ratios [2]). These differences are primarily caused by the dissimilar properties of CO₂ and N₂, which are the primary diluting gases in the furnace during oxyfuel and air combustion, respectively. For example, the molecular weight of CO₂ (44 g/mol) is higher than that of N₂ (28 g/mol); consequently, the density of CO₂ is higher. In addition, CO₂ exhibits a higher heat capacity compared to N₂ and possesses a lower O₂ diffusivity than N₂; therefore, an elevated O₂ concentration in the combustion atmosphere is necessary to maintain the same combustion characteristics as in air firing.

Under oxyfuel conditions, the concentration of CO₂ is high (exceeding 70%) which may prevent the calcination reaction. Calcination occurs only if the partial pressure of CO₂ is less than the equilibrium partial pressure of CO₂ at calcination temperature. There are several empirical correlations published for the calculation of the chemical equilibrium CO₂ concentration dependent on the prevailing temperature [7,8,9]. In Equation (1), an example of the correlation from Garcia-Labiano et al. [7] is shown:

$$y_{CO2,eq}=4.137·{10}^7{e}^{\left({\displaystyle \frac{-20474 }{T}}\right)}$$

(1)

where $y_{CO2,eq}$—the CO₂ concentration at chemical equilibrium, vol.%. T—the temperature in Kelvin.

Figure 1 shows the resulting curve at chemical equilibrium, where $y_{CO2,eq}$ is the CO₂ concentration at equilibrium conditions at a given temperature that divides the non-calcination and calcination reaction regions. According to Figure 1, calcination is inhibited at 850 °C when CO₂ concentration is over 50%_vol, and instead, if there is CaO present, re-carbonation may occur [10,11].

Sulfur dioxide (SO₂) that is generated during the combustion of fossil fuels containing combustible sulfur and is considered a harmful common air pollutant causes serious environmental problems [12] such as acidic rain and the formation of fine particulate matter. In addition, it can endanger the safety of process operation by causing fouling and corrosion issues [13,14].

The release behavior of sulfur depends notably on the form of sulfur. In general, the sulfur in coal or, i.e., oil shale, consists of three forms: pyritic, organic, and sulfate [15,16]. When fuel burns, sulfur oxidizes predominantly to sulfur dioxide according to the exothermic reaction:

$$\mathrm{S}+{\mathrm{O}}_2\to {\mathrm{SO}}_2$$

(2)

Part of the SO₂ may be converted into SO₃:

$${\mathrm{SO}}_2+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2\to {\mathrm{SO}}_3$$

(3)

The equilibrium conversion of SO₂ to SO₃ is low under high temperatures and the reaction rate decreases as the gas temperature cools; as a result, the conversion of SO₂ to SO₃ is small and thermodynamically limited under typical CFB conditions [11,17].

In practice, the actual reaction process of sulfur species is highly complex, involving the simultaneous interactions of various gases, such as SO₂, H₂S, COS, CS₂, O₂, CO, H₂, H₂O, CO₂, and others [18]. Figure 2 outlines the sulfur evolution process during coal combustion under oxyfuel conditions, which was proposed by Tian L. et al. [15]. During devolatilization and volatile combustion processes, organic sulfur compounds initiate combustion, while inorganic sulfur compounds are typically liberated during char combustion. Organic sulfur compounds, including H₂S, COS, CS₂, and other gaseous species, are emitted during the devolatilization phase, with H₂S being the predominant species [15,19]. Inorganic sulfur primarily consists of sulfides and sulfates, with pyrite (FeS₂) being the most prevalent species. Part of pyrite sulfur undergoes devolatilization in the combustion process to generate sulfide and elemental sulfur (S), which can further react with H and CO to produce H₂S and COS, or directly polymerize to Sn. Other pyrite and sulfide intermediates can directly react with H₂, CO, and C to generate H₂S, COS, and CS₂ at various temperature zones. Subsequently, char combustion releases large quantities of SO₂ at temperatures typical for CFB combustion (750–950 °C).

In CFB combustion, capture of SO₂ can be executed by injecting calcium-based sorbents during the combustion of solid fuels. In situ desulfurization is an economically attractive process when using a cheap sorbent (limestone or dolomite). In this in situ desulfurization reaction mechanism process, if it occurs in an air combustion environment, the sorbent first calcines according to the following chemical equation:

$${\mathrm{CaCO}}_3\leftrightarrow \mathrm{CaO}+{\mathrm{CO}}_2$$

(4)

Following the dissociation, the indirect sulfation take place in the CFB combustor, as limestone reacts with sulfur dioxide according to the following chemical equation:

$$\mathrm{CaO}+{\mathrm{SO}}_2+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2\to {\mathrm{CaSO}}_4$$

(5)

In oxyfuel combustion, the fuel burns under an O₂/CO₂ environment affecting the CaCO₃–CaO thermodynamic equilibrium as suggested by Baker [8], so the calcination process does not take place and direct sulfation occurs:

$${\mathrm{CaCO}}_3+{\mathrm{SO}}_2+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2\to {\mathrm{CaSO}}_4+{\mathrm{CO}}_2$$

(6)

Bed temperature and CO₂ partial pressure can define the route of in situ desulfurization, whether it will be a direct or indirect reaction, based on the CO₂ equilibrium curve as mentioned earlier. Many studies have been devoted to understanding sulfur capture efficiency and the obtained SO₂ emission studies were reported based on different aspects; some researchers [20] found that combustion of coal in an oxyfuel combustion environment can reduce the SO₂ emissions due to the increased sulfur retention on ash deposits, by enhancing SO₂ concentration in the combustor with the introduction of RFG. The formation of sulfates in oxyfuel cases was also improved by the longer gas residence time in the furnace compared to the air-fired combustion mode. The effect of direct and indirect sulfation on in situ desulfurization has also been studied and concluded in two ways; direct sulfation has an effect of enhancing capture efficiency [21], while some studies found that the delay in CaCO₃ dissociation and the increased CO₂ partial pressure could reduce desulfurization efficiency by transitioning from indirect sulfation to direct sulfation [22,23]. In principle, the rate of direct sulfation does not decline as significantly with increasing sulfation degree compared to CaO-SO₂ sulfation, due to the reduced impact of sintering during the direct sulfation of limestone. The diffusivity within the product layer exhibits a strong dependence on the temperature and remains relatively constant despite changes in the sulfation degree. This constancy is attributed to the porosity of the CaSO₄ product layer, resulting from CO₂ formation. In oxyfuel combustion, an increasing O₂% has the effect of increasing the bed temperature, and it was found that the influence of O₂/CO₂ combustion modes on the coal desulfurization increased, by increasing the oxygen concentration from 21% to 40% [23], as a result of the temperature increment. A higher excess oxygen ratio improves combustion and makes SO₂ emission decrease while it has an opposite effect on the NO formation [24], as the oxygen excess ratio has the effect of lowering the conversion values of SO₂ that correspond to higher excess oxygen which results from oxidation of sulfur directly to SO₃ [25]. The Ca/S molar ratio is another basic parameter directly affecting the removal of SO₂ emissions in both air and oxy-firing processes. In limestone sulfation, the degree of sulfation is found to be controlled by a chemical reaction and solid-state diffusion [26]. During the reaction phase and the formation of a CaSO₄ layer with a larger molar volume than CaO [27], the rate of limestone sulfation decreases, leading to the need to add more stoichiometric sorbent to remove sufficient SO₂. Ultimately, this will lead to increased solid waste generation. One advantage of using oil shale fuel is that no additional sorbent material is required to be fed to the CFB boiler for the in situ desulfurization process, due to the high content of carbonate as calcite and dolomite. Sorbent particle size also has an effect on the system desulfurization efficiency. It was observed that, as the overall particle size of the limestone samples decreases, the direct desulfurization conversion increases due to the increased surface area. This finding suggests that using finer sorbent particles can effectively enhance the overall desulfurization efficiency [28].

Dolomite, composed of calcium carbonate and magnesium carbonate, is another type of sorbent that has been found to have higher sulfur retention compared to limestone [29]. The half-calcination of the dolomite contributes to the enhancement of sulfur retention, as, when heated, it forms [30] the following:

$$\mathrm{CaMg}{({\mathrm{CO}}_3)}_2\to {\mathrm{CaCO}}_3·\mathrm{MgO}+{\mathrm{CO}}_2$$

(7)

$${\mathrm{CaCO}}_3+\mathrm{MgO}+{\mathrm{SO}}_2+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2\to {\mathrm{CaSO}}_4·\mathrm{MgO}+{\mathrm{CO}}_2$$

(8)

MgO reacts so slowly with SO₂ at temperatures between 540 and 980 °C that it hardly contributes to sulfur capture [30]. However, the CO₂ released from within the sorbent, as described in Equation (6), creates pores in the sorbent particles. This porosity enables SO₂ to penetrate the interior of the sorbent during the sulfation reaction, allowing CaCO₃ to react more efficiently with SO₂.

The fuel used in this study is Estonian oil shale, a low-grade Ca-rich fuel with high volatile organic compound and ash content, primarily calcite and dolomite, exceeding 60% [31]. The widespread utilization of Estonian oil shale necessitates the exploration of techniques to mitigate CO₂ emissions. Oil shale oxyfuel CFB combustion studies are very limited [32,33], and there has been no detailed evaluation of the impact of oxyfuel combustion conditions on oil shale emissions. It can be expected that, among all the available options, oxyfuel CFB combustion for the Estonian power sector represents the most promising carbon capture technology in terms of techno-economic feasibility [34], reduction potential of emissions, energy efficiency, and ease of retrofitting into currently operated power plants. Consequently, the aim of the present work is to study the effect of different inlet oxygen concentrations, bed temperatures, and recycled flue gases on the produced SO₂ emission along with limestone sulfation behavior. The work presents a series of air and oxyfuel combustion campaigns performed in a 60 kWth CFB pilot facility to investigate SO₂ formation and ash sulfur retention efficiency. This study marks the first investigation of oil shale oxyfuel combustion with higher inlet O₂% and the use of RFG.

2. Materials and Methods

2.1. Oil Shale

Estonian oil shale (kukersite) was used in this study. It is a low-grade Ca-rich fuel with high volatile organic compound and ash contents, primarily carbonate minerals [35,36,37]. The fuel mainly comprises carbonates and sandy clay parts; therefore, the molar ratio of Ca/S in the fuel used was higher than 8 (calculated based on the CaO available in the oil shale carbonates to the total sulfur measured by the elemental test). By using oil shale fuel, no additional sorbent material was required to be fed into the CFB boiler for SO₂ capture because carbonate minerals (calcite and dolomite greater than 60%) acted as sorbents for controlling SO₂ emissions. However, the decomposition of carbonate minerals negatively affected the amount of CO₂ emitted. This implies that the application of oxyfuel combustion provides several advantages. Kukersite-type oil shale samples, provided by the Viru Keemia Grupp, were used for the 60 kWth CFB pilot tests, and the fuel was sieved to pass through 3 mm openings.

Table 1. Proximate and ultimate analysis of the oil shale.

Proximate Analysis d (wt.%)				Ultimate Analysis d (wt.%)
Moisture	Ash	LHV (MJ/kg)		C	H	aO	N	Stotal	(CO2)Mineral
0.18	52.05	9.76		27.38	2.68	0.36	0.07	1.57	18.9

d = dry analysis, ^aO% = 100 − (W + A + CO₂ + S + TOC + H + N)%.

presents the ultimate and proximate analyses of the oil shale fuel.

Table 2. Chemical composition of the mineral part of the oil shale.

Components	SiO2	Al2O3	Fe2O3	CaO	MgO	Na2O	K2O	SO3	LOI950°C
Content, wt.%	15.39	3.64	2.03	22.52	3	0.09	1.6	3.91	47.12

LOI_950°C = loss of ignition at 950 °C.

and

Table 3. Mineralogical composition of the mineral part of the oil shale.

Components	Quartz SiO2	K-Feldspar KAISi3O8	Calcite CaCO3	Dolomite CaMg(CO3)2	Chlorite ClO2	Pyrite FeS2	Illite
Content, wt.%	11.6	6.5	43.1	20.3	3.5	1.9	13

list the chemical and mineralogical compositions of the mineral components of the oil shale, respectively. The oil shale particle size distribution (PSD), shown in Figure 3, was determined using a sieve shaker process for mesh sizes ranging between 45 µm and 4 mm (45, 63, 125, 250, 500, 1000, 2000, 4000 µm).

Fuel sample standard characterization (in compliance with the following standards: EVS-ISO 29541:2015 [38], EVS-EN ISO21654:2021 [39]) included proximate analysis (ash, moisture, and lower heating value (LHV)), and ultimate analyses including elemental analysis using an Elementar Vario MACRO CHNS and a TIC to measure C, H, N, and S and total inorganic carbon as (CO_2mineral), respectively. Total organic carbon was measured by calculating the difference between total carbon and inorganic carbon. The chemical composition of ash samples was analyzed using a Rigaku Primus II XRF spectrometer in accordance with ISO 29581-2:2010(E) [40]. Loss on ignition (LOI) represents the samples’ weight loss at 950 °C. Quantitative X-ray diffraction analysis of oil shale was conducted using a Bruker D8 diffractometer equipped with a Lynx-Eye linear detector (in accordance with ISO 29581-2:2010) [40]. The analysis utilized Cu Kα radiation in the 2θ range of 3° to 72°, with a step size of 0.02° 2θ and a counting time of 0.1 s per step, with the X-ray tube operating at 40 kV and 40 mA. Scanning electron microscopy (SEM) from an analytical Zeiss EVO MA15 low-vacuum device was used for sample morphology studies.

2.2. CFB Pilot Facility

A schematic of the pilot setup is shown in Figure 4, featuring a furnace with a height of 4.90 m and an inner diameter of 0.12 m. Fuel was fed into the system via a screw conveyor at a height of 1.17 m, while recirculated solids were reintroduced at a height of 1.5 m along with secondary air. The CFB test facility could be operated using regular air, preset gas mixtures, or RFG. The combustion process was monitored at 16 points using thermal sensors, and the test unit’s data control system was fully automated through LabVIEW. Temperature measurement taps were placed along the riser and heat exchanger. The locations for ash sample collection are indicated in Figure 4, including bottom ash (BA) and the external heat exchanger (EHE). Solid particles were separated from the flue gas in the cyclone and returned to the bottom of the fluidized bed via a return leg and an external heat exchanger. The flue gas composition was simultaneously measured using a Fourier transform infrared (FTIR) analyzer (GASMET4000), positioned between coolers 1 and 2, to measure emissions of CO₂, O₂, CO, H₂O, NO_x, N₂O, and SO₂.

To investigate the effect of O₂ concentration, bed temperature, and RFG on SO₂ emissions and sulfation behavior under CFB combustion, experiments were conducted in air and oxyfuel combustion environments. Tests under the oxy regimes were carried out using O₂/CO₂ mixtures in the required proportions either directly from cylinders or using flue gas recirculation. During the experiments, the concentration of O₂ in the O₂/CO₂ mixture supplied for combustion varied in quite wide ranges up to 59 vol% in the experiments when running directly from the cylinders. One of the important factors influencing the fate of SO₂ is the temperature in the combustion chamber, and in order to study its influence, the tests were carried out at different temperatures of 770–950 °C.

Table 4. Average operating parameters of 60 kWth CFB unit during air and oxy-combustion modes.

Working Parameter	Combustion Mode
	Air	O2/CO2	O2/RFG
Inlet O2 ratio (%)	-	21–59	50, 56, 87
CO2 in flue gas, dry (%)	13–18	76–93	42–73
TMAX (°C)	738–892	695–947	810–875
Oil shale feed rate (kg/h)	9.5	5–12	5–9
Ash sampling point	BA, EHE	BA, EHE	BA, EHE

shows the main average working parameters, under air, O₂/CO₂ (21–59 vol%), and O₂/RFG (50 and 75 vol%) combustion regimes. Combustion temperature (T_MAX) represents the maximum temperature in the riser measured during the combustion process, which was mainly observed at a height of (2.15) m or (4.37) m, depending on the combustion condition. The average oil shale feed rate under air combustion was 9.5 kg/h, while the feed rate showed a wider range under oxy-firing and with RFG experiments, as shown in Table 4. The detailed selection of the pilot test and combustion atmospheres can be found in [31,37].

Since oxygen environment experiments were conducted under different inlet oxygen concentrations, there are differences in the oil shale feeding rate. Thus, the SO₂ emissions in the flue gas should be recalculated to take into consideration this difference, by defining SO₂ emission rate as mass pollutant emitted per primary (oil shale) energy input (mg(SO₂)/MJ) [41,42].

The CFB pilot facility consistently initiated in air combustion mode. After the combustion process achieved the desired experimental temperature, a baseline test for air combustion was conducted. Upon successfully completing the air test under stable conditions, the transition to O₂/CO₂ or O₂/RFG combustion began. As shown in Figure 5, this transition was carried out smoothly and within a short time frame. Under oil shale oxy-firing conditions, as the combustion process proceeded, the increasing temperature promoted fuel combustion in the oxygen-rich environments (O₂/CO₂) with the CO₂ concentration in the flue gas exceeding 90%.

3. Results and Discussion

3.1. SO₂ Emissions under Air and Oxyfuel Combustion Atmospheres

The SO₂ emissions produced in the flue gas streams were extremely low due to the high molar ratio of Ca/S. Figure 6 shows the SO₂ emissions and their normalized values as mass pollutants emitted per primary (oil shale) energy input (mg (SO₂)/MJ). Under oxy-firing conditions, as the combustion process proceeded, the elevated partial pressure of CO₂ that causes the oxy-firing atmosphere differed from that in air mode. The sulfur capture mechanisms may shift from the normal sulfation path (CaCO₃ → CaO → CaSO₄) to the direct sulfation (CaCO₃ → CaSO₄). A literature study has shown that the formation of SO₂, when fuel is burned in an oxygen-rich environment is analogous to that when it is burned in air [43]. Figure 6 shows that the increased inlet O₂% does not affect the release of SO₂ emission; compared to air-firing, SO₂ emission were increased under O₂/CO₂ of 21/79%_vol and strongly reduced to near zero with the increment of inlet O₂. Similar behavior was reported previously by Duan et al. [44], who showed that desulfurization efficiency increased from 40.6% to 53.0% as O₂ concentration increased from 21 to 40%. The low combustion operating temperature (direct sulfation), combined with a low inlet O₂ concentration, led to increased SO₂ emissions. This is due to reduced combustion efficiency and enhanced CO formation in the combustor. The elevated CO levels act as a catalyst, accelerating the formation of SO₂, as indicated in the introduction, Figure 2. The slight variations that occurred during different air and 30/70% of O₂/CO₂%_vol (OXY30) experiments are affected by testing under different combustion temperatures; further details will be explained in Section 3.2. Previous experiments [45,46] of oil shale thermal gravimetric analysis coupled with a mass spectrometer (TGA–MS) showed an enhancement of SO₂ release when switching from air to oxyfuel combustion. Yet, under real test conditions, these emissions were reduced owing to the improvements of capturing efficiency (when the in-furnace sulfur capture is applied) due to higher SO₂ concentrations in furnaces operating in oxyfuel combustion atmospheres [2,33].

Under O₂/RFG, three ratios were used to recycle back the flue gas to the boiler. SO₂ emissions were unaffected by the application of RFG and remained at the lower level. It was reported that SO₂ emissions are increased with RFG [47], as accumulated SO₂ concentrations inside the boiler rise, with the recycled gas providing an additional sulfur source. As a result, an increase in SO₂ concentration is observed in O₂/RFG mode due to the substantial accumulation of SO₂ in the furnace during flue gas recycling [48,49]. Yet, the obtained results indicated no increase in SO₂ emissions, whereas higher SO₂ concentration in the combustor leads to higher sulfur retention in the ash [22,50]. Overall, the change in combustion atmosphere did not affect the release of SO₂ emissions in an oxyfuel combustion environment, and the sulfur capture efficiency reached 99% for all combustion modes. Notably, in the case of oil shale combustion, the thermal decomposition of dolomite contributes to sulfur capture. Moreover, the high Ca/S molar ratio in the oil shale fuel ensures that there is a sufficient excess of calcium available to bind with SO₂.

3.2. Effect of Temperature

The combustion temperature is a very important parameter in the release of SO₂ emissions; along with the CO₂ concentration, the sulfation process will proceed, and the limestone will operate under calcination or non-calcination conditions. Figure 7a shows the dependence of normalized SO₂ emissions on the maximum temperature in the riser during air and OXY30 combustion regimes. Under air combustion, the average maximum temperature ranges from 774 to 892 °C, this range corresponds to the typical operational temperature of fluidized bed combustion systems [26]; the differences in the excess O₂ (λ) were very small between all tests, ~1.115; and the minimum SO₂ emissions were obtained for T_Max temperatures < 850 °C. A literature study [25] has found that the maximum sulfur retention in fluidized bed boilers occurs at a temperature of approximately 850 °C in an air-firing CFB environment. This is because the reaction between CaO and SO₂ reaches its peak at this temperature. It suggests that a higher temperature in the combustor promotes sulfur emissions, that the formation of SO₂ is favored with combustion temperatures exceeding 850 °C, and that with increasing temperatures, the sulfation rate decreases due to the sintering effect. Sintering of Estonian oil shale ashes became a problem for combustion temperatures exceeding 900 °C [51]; however, further studies and tests are needed to identify the reasons.

Under an O₂/CO₂ environment with a 30% inlet O₂, the excess air ratios were between 1.31 and 1.45, the effect of combustion temperature on SO₂ emission started to occur after 900 °C, and SO₂ capture efficiency shifted to a higher temperature compared to air combustion. Studies show the same effect of oxyfuel combustion environments under 25% of oxygen concentration; there, sulfur capture efficiency was increased with increasing bed temperatures up to 880–890 °C, and with further increases in bed temperature, sulfur retention decreased [52]. Moreover, compared to air-firing, the effect of combustion temperature is shifted, as the environment in which the oil shale burns differs from that of air combustion. The temperature of the CFB combustor and CO₂ partial pressure is changed, affecting the CaCO₃–CaO thermodynamic equilibrium resulting in inhibition of carbonate decomposition. In this sense, the in situ desulfurization occurred via both direct and indirect sulfation, Equations (5) and (6). Obras-Loscertales et al. [29] reported direct sulfation at temperatures below 860 °C, and indirect sulfation of CaO-SO₂ above that temperature when investigating the sulfation reaction between 800 and 975 °C. The temperature association exhibits maximum SO₂ values at about 890 °C for air combustion and at about 945 °C for oxyfuel combustion. The oxyfuel’s maximum SO₂ release was shifted up by around 50 °C, which corresponds to the sulfation reaction mechanism below the CO₂–CaCO₃ equilibrium curve indicated in Figure 7b, which shows the conditions of the experimental points in the CO₂–CaCO₃ equilibrium diagram. The equilibrium temperatures for tests 1 and 2 are below 880 °C and lie in the non-calcination region where direct sulfation will probably occur. At higher temperatures (exceeding 900 °C), the increase in SO₂ emission occurred during the indirect sulfation reaction, and the higher the temperature, the more SO₂ released in the flue gas. This can be attributed to both indirect sulfation efficiency and reactivity of CaO due to the sintering effect at elevated temperatures exceeding 950 °C.

Figure 8 compares ash contents of lime (CaO), calcite (CaCO₃), and anhydrite (CaSO₄) for the extracted ash samples from bottom ash (BA) and external heat exchanger-circulating ash (EHE). The switch from air to oxyfuel combustion slightly increased the average sulfate content (as anhydrite) in the ashes from BA and EHE. The differences in temperature under the OXY30 regime mainly affected BA ash sulfate content, and for EHE ashes, anhydrite concentrations were almost similar despite the raise in temperature, because the capture efficiency is mainly enhanced due to the higher particle residence time in circulating ash compared to BA ashes. The effect of temperature on sulfation at higher combustion temperatures under OXY30 can be seen in the anhydrite concentration in the BA and EHE ashes for test 3, as the anhydrite content is close to test 2 despite the high concentration of SO₂ in the combustor. A possible explanation is that particles subjected to sulfation above 900 °C developed a dense outer layer of CaSO₄, exhibiting high diffusion resistance. This increase in diffusion resistance could elucidate the observed decrease in SO₂ removal rate as temperatures surpassed 925 °C. A parallel trend was reported by García-Labiano et al. [22], who noted sintering and diminished CaSO₄ formation above 900 °C. SO₂ emissions in test 2 within the direct sulfation route showed a reduction in emission with temperatures below 850 °C. However, when combustion temperatures exceeded a certain threshold (corresponding to the equilibrium curve of carbonation and calcination reactions), the sulfation process transitioned to an indirect route; at further increases in temperature particle sintering cannot be avoided resulting in low sulfur retention.

Overall, the results suggest that the increased formation of SO₂ emission under oxyfuel CFB combustion was shifted to higher combustion temperature. Lower temperatures, where the direct sulfation under oxyfuel combustion environment occurs, show lower SO₂ emissions compared to the indirect sulfation at higher operating temperatures which eventually effect capturing efficiency. It can be concluded that the optimum temperatures for sulfur retention for both air and OXY30 are at operating temperatures below 850 °C.

3.3. Effect of Excess Oxygen Ratio (λ)

The excess oxygen ratio (λ) refers to the amount of oxygen supplied to the combustion process compared to the theoretically required amount for complete combustion. Excess oxygen is often necessary to ensure full combustion and minimize the formation of harmful emissions such as carbon monoxide (CO) and unburned hydrocarbons. The value is calculated by comparing the actual oxygen concentration in the flue gas to the stoichiometric oxygen concentration needed to react with the fuel completely. Figure 9 shows different excess oxygen ratios ranging from 1.09 to 1.96 and compares total SO₂ emissions as mg/MJ between air, O₂/CO₂, and O₂/RFG combustion regimes. The results are similar for the air and O₂/CO₂ combustion modes, the SO₂ emission was reduced by increasing λ. With RFG application, SO₂ emissions are more affected by the SO₂ that is recycled back to the combustor. Air staging and excess air combustion are good methods to control emissions; in our previous study [53] we discussed the same experimental tests on the NO_x emissions and the results showed that reducing excess oxygen is the most dominant parameter to control NO_x formation. When it comes to SO₂ emissions, the opposite behavior is noticed, and the emissions increase under reducing environments. This has been discussed in several literature studies using other fuels, i.e., coal [24,42,54]: by lowering excess oxygen, the sulfur in the fuel is converted to H₂S instead of SO₂ because of the reducing environment in the dense bed region. H₂S might react with CaCO₃ to form CaS, Equation (9). The rest of the H₂S that is not captured by CaCO₃ would be oxidized to SO₂, Equation (10), and CaSO₄, Equations (5) and (6), by excess oxygen in the dilute phase zone. Yet, formation of CaSO₄ is limited due to the low SO₂ concentration in the dense bed zone and therefore its low ability to react with CaO/CaCO₃.

$${\mathrm{CaCO}}_3+{\mathrm{H}}_2\mathrm{S}\to \mathrm{CaS}+{\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{O}$$

(9)

$$2{\mathrm{H}}_2\mathrm{S}+3{\mathrm{O}}_2\to 2{\mathrm{SO}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(10)

Overall, SO₂ emissions from oil shale oxyfuel combustion are following an analogue route of other fuels, i.e., coal. The flue gas oxygen concentration is an important parameter during the combustion process; a reasonable excess oxygen ratio for CFB combustion operation can enhance sulfur capture efficiency. However, a specific limit needs to be followed to avoid NO_x formation. Oil shale possesses a substantial calcium content, resulting in an ash with a high Ca/S ratio. Consequently, the ash’s ability to self-desulfurize is significantly high. This process can slightly be impeded when oxygen levels are inadequate within the dense phase zone, irrespective of whether sulfation occurs indirectly or directly.

3.4. Sulfation and Ash Behavior

The SO₂ produced by oil shale combustion reacts with CaO produced by the thermal decomposition of calcite and dolomite in the combustor riser, or with the available undecomposed CaCO₃, or the dolomitic calcite produced from dolomite decomposition. These reactions culminate in the formation of anhydrite (CaSO₄). The presence of anhydrite in ash samples indicates ash sulfur retention, and anhydrite is mainly derived from the ash generated in the dense bed region and with higher content in the circulating ash. Figure 10 shows the content of anhydrite in BA and EHE ashes. It can be seen that ash sulfur retention is higher in oxyfuel combustion than in air; these results were also discussed in detail in our previous study [31]. The majority of SO₂ under the O₂/CO₂ regime is bound to circulating ash. Under O₂/RFG, a higher anhydrite content is found in mainly BA ash. The results are consistent with our explanation that higher SO₂ content is found under an O₂/RFG atmosphere, resulting from the enhanced sulfation reaction due to the increased particle reaction time as SO₂ was recycled back into the combustor, and the higher partial pressure of SO₂ in the combustion atmosphere, which ultimately leads to greater sulfur capture efficiency [25,31].

The effect of temperature and CO₂ partial pressure is shown more clearly on the SEM photographs of the same BA and EHE ashes under air, OXY30, OXY40, and OXY87+RFG combustion atmospheres, Figure 11. The variations in BA and EHE ash contents of lime, calcite, and anhydrite are shown in Figure 12. The particle surface of the porous decomposed carbonate is obvious in the air and OXY30 combustion modes, Figure 11a–d, which resulted from the high temperature exceeding the carbonate decomposition temperature, Figure 7b. There, the product layer became more and more porous with temperatures exceeding 900 °C (under OXY30), and the effect is shown clearly on BA ash compared to EHE ash. This can be explained by increasing the temperature beyond the equilibrium curve limit under oxy-conditions. In that case the decomposition of CaCO₃ occurred and the release of CO₂ during the calcination process caused the porous shape of the particles [55]; therefore, the surface of ash particles is much more porous. Literature experiments [55] have proved that the transition from direct to indirect sulfation can enhance sulfur capture efficiency, as SO₂ can enter the inner part of the CaO particle easily, which increases sulfur retention. The results indicate that SO₂ emissions increased under OXY30, whereas ash sulfur retention remained unchanged compared to OXY40, Figure 10. Consequently, a decrease in sulfation capacity at higher temperatures was observed, which may be attributed to the sintering phenomenon [56,57]. However, additional ash analyses are needed to confirm these findings.

Under the OXY40 combustion mode, Figure 11e,f, the indirect sulfation route is obtained within the operating temperature, and the particle surface has fewer pores, indicating limited decomposition of limestone observed in BA ashes. Circulating ash is more affected by riser temperature as more porous can be noticed on the surface of particles.

Under O₂/RFG, Figure 11g,h, and with a high inlet O₂ concentration, the ash distribution is linked to a pronounced grain size separation compared to the air combustion atmosphere [31]. This results in the production of finer particles as the O₂ concentration increases, as the higher oxygen levels enhance mineral vaporization and nucleation [58]. The effects of low operating temperature under the application of RFG were extracting ash of coarse mode from the dense bed region (BA), which is strongly dominated by the undecomposed large particles of calcite (CaCO₃), formation of anhydrite (CaSO₄), and fine particles from the circulating ashes (EHE).

It is suggested that both temperature and gas composition play crucial roles in shaping the pore structures of the product layer during oxyfuel combustion processes.

4. Conclusions

This work presented a series of air and oxyfuel combustion campaigns that were performed in a 60 kWth CFB pilot facility to investigate SO₂ formation and ash sulfur retention efficiency. This study marks the first investigation of oil shale oxyfuel combustion with higher inlet O₂% and with the use of RFG. The 60 kWth CFB operated under three main combustion environment cases: case 1: air combustion, case 2: oxygen enriched combustion O₂/CO₂, and case 3: oxyfuel combustion with RFG (O₂/RFG). The inlet oxygen concentrations under O₂/CO₂ oxyfuel combustion environment ranged from 21 to 50%, and under O₂/RFG the combustion modes were 50%, 56%, and 87% of inlet O₂.

The objective of the present work is to investigate the optimal parameters including combustion atmosphere and oxygen concentration, operating temperature, and excess oxygen ratio (λ) on the production of SO₂ pollutant emissions along with ash sulfur behavior. The results will provide a reference model for the industrial application of oxyfuel combustion in oil shale power plants.

The overall outcomes indicate that the adiabatic oxyfuel combustion reduced SO₂ formation compared to air-firing. The release of SO₂ emissions was enhanced under a low operating combustion regime. Eventually, with the increased excess oxygen ratio (λ), SO₂ emissions were reduced for all combustion atmospheres. The optimum combustion temperature for sulfur retention for both air and OXY30 was achieved at operating temperatures below 850 °C. Under O₂/RFG, higher SO₂ concentration in the combustor led to higher sulfur retention in the bottom and circulating ash. Under an O₂/CO₂ atmosphere, the ash anhydrite (CaSO₄) contents were majorly bound to circulating ash. Moreover, the temperature and gas composition are major factors in identifying the pore structures of the product layer during the oxyfuel combustion process. Yet, further studies are needed to identify ash sintering effects on sulfur retention.