Oxy-fuel combustion is regarded as one of the most promising carbon capture and storage (CCS) technologies to mitigate the climate change, which has been widely studied and demonstrated in academia and industry [1,2,3]. In the oxy-fuel process, a mixture of recycled flue gas and pure O2 obtained from an air separation unit (ASU) is introduced into the combustion chamber to replace air as oxidant gas. Therefore, high concentration of CO2 can be obtained in flue gas, which is suitable for the subsequent carbon storage and utilization, the schematic diagram of oxy-fuel combustion system is shown in Fig. 4.1. It also has other advantages, such as low NOx emission, easy scale-up, and applicability in existing power plant [4, 5].

Fig. 4.1
A schematic diagram illustrates the setup of oxy-fuel combustion. The components labeled are boiler, steam turbine, steam condenser, particle removal, cooler and condenser, Sulphur removal, and carbon dioxide compressor.

Schematic of the oxy-fuel combustion

Full size image

However, the intensive energy consumption associated with the ASU is the bottleneck limiting its further commercialization [2, 6]. The systematic analysis through process simulation has shown that the energy consumption of ASU is about 61.5 MWth, accounting for ~15.9% energy of a 388 MWth coal-firing oxy-fuel boiler (as shown in Fig. 4.2) [7]. As stated in the earlier sections, the OCAC process has a proven advantage of burning fuel at lower oxygen-fuel ratio compared to conventional FBCs. In addition, because the recirculation brings oxygen to the furnace adding to the new oxygen from outside, the oxy-fuel boiler using flue-gas recirculation may have a very low over-all oxygen excess. There is an overall oxygen stoichiometric factor and an internal oxygen stoichiometric factor, the latter is always larger than the overall factor [8]. Therefore, the combination of oxy-fuel combustion and the OCAC technology, i.e. oxygen-carrier-aided oxy-fuel combustion (oxy-fuel-OCAC), can be expected to improve the utilization of O2, leading to less energy consumption from the ASU.

Fig. 4.2
A 3-D pie chart of the oxy-fuel power cycle. The gross power equals 388.0 megawatts electric. The chart has the highest value of 264.3 megawatts electric for net power, and the lowest value of 7.2 megawatts electric for the feed water pump.

Net power and parasitic power demand of the oxy-fuel power cycle (reprinted from [7] with permission of Elsevier)

Full size image

At present, the only oxy-fuel-OCAC studies were carried out by the fluidized-bed conversion and gasification (FBC&G) group from CanmetENERGY in a 50 kWth CFB combustor (Fig. 4.3) [9,10,11]. Different factors including the type of fuels (coal, biomass and natural gas), the bed materials (quartz sand, olivine sand, ilmenite ore or their mixture), O2 concentration in flue gas (2, 5 and 8%), the bed temperature (800, 850 and 900 °C) and Ca/S ratio (0 and 2) have been systematically analyzed in both bubbling and circulating fluidizing modes (Tables 4.1 and 4.2).

Fig. 4.3
A schematic diagram of the 50 kilowatts thermal capacity oxy F B C. The parts labeled are the Pressurized hopper dry feed system, wind box, electric heaters, combustor, cyclone, baghouse, stack, condenser, condensate sampling, and recycle blower.

Schematic of the 50 kWth Oxy-FBC of CanmetENERGY (reprinted from [9] with permission of American Chemical Society)

Full size image
Table 4.1 Test matrix (adapted from [9])
Full size table
Table 4.2 Operating parameters (adapted from [9])
Full size table

4.1 Combustion Performance

The typical temperature profiles when burning coal with different bed are shown in Fig. 4.4. The research results of oxy-fuel-OCAC show that the fluctuation of reactor temperature profile using ilmenite ore as the bed material was less than that using quartz sand [9], suggesting that oxy-fuel-OCAC was beneficial to obtain an evenly distributed furnace temperature. The unburned species (CO and CxHy) emissions were significantly reduced by replacing the quartz sand with the ilmenite ore. Using ilmenite ore as bed material to replace the quartz sand, the CO concentration in the flue gas was reduced by 30% when burning Highvale coal and 13% for Poplar River coal, and the hydrocarbons, i.e. CH4 and C2H4, measured above the bed were also reduced by almost 50% for both coals [9]. By measuring the unburned species at the top of the bed and cyclone, it was found that the improvement on combustion was even more pronounced in the dense phase than that in the dilute phase using OCAC, in particular under conditions with low O2 in flue gas or low bed temperature (as shown in Fig. 4.5). When bed material was switched from quartz sand to ilmenite ore, at the case of 2.5% O2 in the flue gas, a 50% reduction of CO concentration occurred at the top of the bed. If the O2 in the flue gas was increased to 5%, the reduction in CO concentration was 40%, and finally, there was almost no observed reduction of CO concentration when O2 in the flue gas was 8%. This is because the reaction rate of CO with gaseous oxygen is much faster than that with the OCs. Therefore, gaseous O2 tends to be the control factor in the overall rate of CO conversion in the high O2 concentration case. Similar conclusion was also drawn from the oxy-fuel-OCAC experiments using natural gas and biomass as the fuels [11].

Fig. 4.4
3 line graphs. Graphs a and c plot bed temperature versus time. They display six curves that begin at 10 A M and rapidly fall after 3 P M. Graph b plots the riser temperature versus time. It displays four curves that decrease after 3 P M.

Temperature profiles when burning coal with different bed: a in the bed (sand bed, highvale coal), b in the riser (sand bed, Highvale coal) and c in the bed (ilmenite bed, Poplar River coal) (reprinted from [9] with permission of American Chemical Society)

Full size image
Fig. 4.5
A graph plots C O concentration versus O 2 concentration. It displays four declining slopes for sand and ilmenite. The emission of the gas sample at the top of the bed is greater than that of the gas sample at the top of the cyclone.

CO emission from burning Highvale coal (adapted from [9])

Full size image

4.2 Emissions

4.2.1 SO2 Emission and Desulfurization

Figure 4.6 shows the SO2 emissions in the different operation conditions of oxy-fuel-OCAC. When the quartz sand as bed material was replaced by ilmenite ore, the SO2 concentration at the top of the bed and cyclone both always became lower, regardless of the bed temperature, the exhaust oxygen concentration, and the use of a desulfurizer [10]. Under the condition of low exhaust oxygen concentration, the inhibition effect of ilmenite ore on the SO2 emission was more significant. When the bed materials were all ilmenite ores, the SO2 emission was reduced by more than 60% at different bed temperatures compared to the usage of quartz sand as bed material. Under the same mixing ratio (1:1) of quartz sand and ilmenite ore, the fresh ilmenite ore was more beneficial to the reduction of the SO2 emission than the used one. When desulfurizer was introduced during the OC aided oxy-fuel combustion process, the process with ilmenite ore as bed material also obtained a lower SO2 emission in comparison to the case of quartz sand.

Fig. 4.6
Two graphs plot S O 2 concentration versus O 2 concentration, and temperature. Graph 1 displays decreasing trends for H C and P C substances. Graph 2 displays fluctuating trends for P C sand, P C fresh mix, P C spent mix, and P C ilmenite ore. It also displays an increasing trend for P C sand at the top of the bed and a decreasing trend for P C ilmenite ore at the top of the cyclone.

SO2 emissions in different operating condition. HC and PC are Highvale coal and Poplar River coal (adapted from [10])

Full size image

The existing study [10] only gave the variation rules of SO2 emissions in OCAC conditions, but the potential mechanisms of SO2 emission and desulfurization in oxy-fuel-OCAC was not clear. Combining with the mechanisms of SO2 reduction proposed in Ref. [10], it may be speculated that the low SO2 emissions in oxy-fuel-OCAC process depend on three factors. Firstly, the ash-related components on the surface of and inside the OC particles can absorb sulfur [12], thereby reducing the SO2 emission. Secondly, the addition of OCs leads to more uniformly distributed bed temperature and oxygen concentration, which may potentially reduce the absorption of SO2. Finally, more uniform oxygen distribution can promote direct desulfurization and indirect desulfurization according to Reactions 3.3 and 3.4, thus improving the desulfurization efficiency and reducing the SO2 emission. Further investigations in this subject are highly desired to validate these hypotheses.

4.2.2 NOx and N2O Emission

Lu et al. [10] investigated the NOx and N2O emissions in oxy-fuel combustion in the presence of OCAC in the plant of Fig. 4.7. They found that the NOx concentration increased significantly at different measured locations at various temperatures and exhaust oxygen concentrations. They assumed that the CO and hydrocarbons concentrations in the boiler decreased in the OCAC condition, which will weaken the reduction of NOx through gas-phase reaction. Meanwhile, OC may reduce the concentration of ammonia (from the volatiles) in the dense-phase zone, thus reducing the reduction of NOx by NH3 in the recirculated flue gas (via Reactions 4.1–4.3), resulting in a higher NOx emission. Lu et al. [10] measured the composition of flue gas and found that the exhaust NH3 concentration was reduced for oxy-fuel-OCAC compared to the case of using quartz sands as bed materials [13, 14].

$$ 6{\text{NO}} + 4{\text{NH}}_3 = {\text{5N}}_{2} + 6{\text{H}}_2 {\text{O}} $$
(4.1)
$$ 4{\text{NO}} + 4{\text{NH}}_3 + {\text{O}}_2 = {\text{4N}}_2 + {\text{6H}}_2 {\text{O}} $$
(4.2)
$$ 8{\text{NH}}_3 + 6{\text{NO}}_2 = 7{\text{N}}_2 + 12{\text{H}}_2 {\text{O}} $$
(4.3)
Fig. 4.7
2 graphs. Graph a plots N O x versus O 2 concentration. It displays 4 positive slopes. Graph b plots N 2 O versus O 2 concentration and temperature. It displays 2 positive slopes for high vale coals and negative slopes for poplar river coals.

NOx and N2O emissions in different operating conditions. a NO emission from Highvale coal at constant temperature. b N2O from Highvale and Poplar River coals at constant O2 in the stack (adapted from [10])

Full size image

On the other hand, NH3 not only can participate in the Reactions 4.1–4.3 in CFB boilers, but it also may be oxidized by O2 and OCs, resulting in a high NOx emission. Taking iron-based OC as an example, the possible reactions are as suggested as follows [15,16,17,18,19]:

$$ 2{\text{NH}}_3 + 1.5{\text{O}}_2 = {\text{N}}_2 + 3{\text{H}}_2 {\text{O}} $$
(4.4)
$$ 2{\text{NH}}_3 + 2{\text{O}}_2 = {\text{N}}_2 {\text{O}} + 3{\text{H}}_2 {\text{O}} $$
(4.5)
$$ 2{\text{NH}}_3 + 2.5{\text{O}}_2 = 2{\text{NO}} + 3{\text{H}}_2 {\text{O}} $$
(4.6)
$$ {\text{NH}}_3 + 4.5{\text{Fe}}_2 {\text{O}}_3 = 0.5{\text{N}}_2 + 3{\text{Fe}}_3 {\text{O}}_4 + {\text{CO}}_2 + 1.5{\text{H}}_2 {\text{O}} $$
(4.7)
$$ {\text{NH}}_3 + 6{\text{Fe}}_2 {\text{O}}_3 = 0.5{\text{N}}_2 {\text{O}} + 4{\text{Fe}}_3 {\text{O}}_4 + {\text{CO}}_2 + 1.5{\text{H}}_2 {\text{O}} $$
(4.8)
$$ {\text{NH}}_{3} + {12}.{\text{5 Fe}}_{2} {\text{O}}_{3} = {\text{NO}} + 5{\text{Fe}}_3 {\text{O}}_4 + {\text{CO}}_2 + 1.5{\text{H}}_2 {\text{O}} $$
(4.9)
$$ {\text{NH}}_3 + 10.5{\text{Fe}}_2 {\text{O}}_3 = {\text{NO}}_2 + 7{\text{Fe}}_3 {\text{O}}_4 + {\text{CO}}_2 + 1.5{\text{H}}_2 {\text{O}} $$
(4.10)

Several studies [15,16,17,18,19] have also shown that OCs can oxidize HCN, another typical NOx precursor, into N2, N2O, NO and NO2 (4.4–4.10). Therefore, more uniform oxygen distribution under oxy-fuel-OCAC operation may reduce the formation of NH3 and HCN, which in turn promote the generation of NOx.

No matter whether OCs or inert silica sand were used as bed materials, the N2O emission always decreased with the increase of bed temperature, while it increased with the increase of exhaust oxygen concentration. At high exhaust O2 concentration and low bed temperature, N2O emission is slightly lower in the case of oxy-fuel-OCAC than that of using quartz sand, and vice versa [10]. It should be noted that the N2O concentration was measured and the difference in N2O emissions were very small between the two bed-material cases.

Since N2O can be decomposed at high temperature, the main parameter affecting N2O emission under oxy-fuel-OCAC process during FBC is still dominated by temperature [10]. Although higher temperature can reduce N2O emissions, it will also enhance the generation of NOx and reduce the desulfurization efficiency of the boiler. Thus, it is hard to simultaneously reduce N2O and NOx emissions via variation of the combustion temperature [20,21,22]. In addition, compared with quartz sand as bed material, the oxy-fuel-OCAC enables to operate the boiler at a lower excess air, i.e. exit O2 concentration, which is also expected to make the boiler achieve lower N2O emissions.

4.2.3 Condesate Liquid Analysis

Lu et al. [10] also analyzed the liquid discharged from the front cooling of the recycle blower, the analysis results are shown in Table 4.3. It was found that all the samples were acidic with pH values ranging from 2.32 to 2.62. The pH value of the aqueous condensate from the oxy-fuel-OCAC case was slightly higher than that from using silica sand, which may help to alleviate the corrosion of the heat-exchanger surface. The main components in the aqueous condensate were sulfate (180 mg/L to 270 mg/L) and chlorides (9.6–50.7 mg/L) accounting for approximately 90 wt% of dissolved species, with the coexistence of fluorides and bromines (<0.5 mg/L). There is no obvious change in the content of the sulfate condensate between conventional combustion and oxy-fuel-OCAC. However, the concentration of the chlorides in the condensate varies significantly between the two cases. The aqueous condensate generated from the oxy-fuel-OCAC case was reduced by more than 50% compared to that from conventional combustion. This might be due to the reaction between Cl2 and TiO2 following the reaction [23]:

$$ {\text{2TiO}}_{2} + {\text{3C}} + {\text{4Cl}}_{2} \to {\text{2TiCl}}_{4} + {\text{2CO}} + {\text{CO}}_{2} $$
(4.11)
Table 4.3 Condensate analysis (adapted from [9])
Full size table

Reaction 4.11 is a part of the process for producing TiO2 from ilmenite by using chlorine as the oxidant [24]. This process is typically performed in a fluidized bed at 900–1000 °C, and coke needs to be added to the bed inventory to generate a reducing atmosphere, which is very close to the dense-bed environment during fluidized-bed combustion. The product, TiCl4, is a hazardous volatile gas pollutant, whose boiling point is 136 °C. Whether the OCAC technology using ilmenite ore as bed material causes an emission of TiCl4 into the atmosphere or not needs to be further evaluated by specific experiments.

The effect of the chlorine content on the corrosion and ash deposition of the boiler’s heat exchanger tubes is particularly significant [25, 26]. In severe cases, it may cause an unplanned shutdown of the boiler. Although it is known that the oxy-fuel-OCAC technology using ilmenite ore as bed material can significantly reduce the emission of chlorides, many aspects need to be further clarified. The generation of potentially harmful substances should be investigated in detail. In addition, OC-based bed materials other than ilmenite ore should be screened to find the best option to achieve low chloride emission. Nevertheless, control of Cl emission could be a key feature of oxy-fuel-OCAC technology, especially for burning fuels with high chlorine content, such as MSW.

4.3 Bed Material Analysis

Hughes et al. [9] have analyzed the OC bed materials (ilmenite ore) by in-situ sampling the bed particles followed by XRD characterization, the analysis results are shown in Table 4.4. It was found that the crystallinity of the used ilmenite ore was about 30–36% after oxy-fuel-OCAC, far lower than that in the original material (65–75%). The used ilmenite ore may have better reactivity, as the amorphous-phase species in the ilmenite ore are usually more reactive [27, 28]. In addition, under the operating conditions of low exhaust O2 concentration and bed temperature, higher concentrations of FeTiO3 and FeO were found in the bed material, which were the reduced phases of Fe2TiO5 and Fe2O3, respectively. It suggests that the OC-based bed material indeed participates in the redox reactions during the oxy-fuel-OCAC process, especially at low temperature and exhaust O2 concentration.

Table 4.4 XRD analysis of bed samples (adapted from [9])
Full size table

The tested coal in the above-mentioned studies contains a high concentration of Na [9]. A high content of Na in the ash is likely to cause problems related to bed agglomeration and fouling downstream [29, 30]. However, no Na-related compounds were found in the solid samples from the bed, which indicates the potential for Na release to the downstream. During the experiments, no hot spots and defluidization were found in the bed. After SEM analysis, it was found that the bed material did not show any signs of agglomeration. This indicates that the OCAC technology has a good applicability in burning high Na-content coal, and that there is no agglomeration problem or Na-induced deactivation of the oxygen carrier. This is beneficial for burning high-sodium coal. The Na-based salts have an important influence on the formation of atmospheric particulate matter (PM) and the formation of ash deposits on the heat-exchanger surface. The effect of the OCAC technology on the PM formation, ash deposition and corrosion is still unclear, which is worthy a detailed investigation.