Z. Liu, M. Hughes, Y. Tong et al.
Energy 239 (2022) 121925
5) were analyzed. P-sludge biochar altered the bio-oil chemical composition (GC-MS and GC-FID results are shown in Fig. 5 and Table 2, respectively). Only a few of organic constituents were detected and quanti fi ed by GC-FID within fi ve cycles, including 3,5- Dimethoxy-4-hydroxybenzeldehyde, ethylbenzene, and styrene. Additionally, GC-MS analysis showed that many major peaks such as toluene, ethylbenzene, styrene, phenol, cresol, indole, and dimethylthiohydantoin were detected in the 0B bio-oil but were not detected or had much lower peak height in the catalyzed bio-oil within fi ve cycles, indicating lower concentrations of organic con- stituents after catalysis. (0B denotes control without catalyst and without sand at 500 C downstream.) In addition to the chemical composition change, the optical properties changed due to p-sludge biochar catalyst (Figure S2). In particular, the dark brown, light non-aqueous phase was completely eliminated by p-sludge biochar and only an aqueous phase was produced. In contrast, the non-catalyzed control (0B) produced condensate composed of a light non-aqueous phase as well as an aqueous phase containing soluble organics, while the color of catalyzed bio-oil was clear. Even after three cycle tests, the bio-oil color was still semitransparent. However, due to the deac- tivation of p-sludge biochar, the dark phase formed again after four cycles. Dark bio-oil color is associated with the presence of unsat- urated hydrocarbons [35]. Therefore, the optical changes also indicated that p-sludge biochar resulted in fewer unsaturated hy- drocarbons in the bio-oil compared to the control test 0B. The used p-sludge biochar (cycle 1) still had a porous structure, while the surface morphology became much less porous after the fi fth use (cycle 5) (SEM-EDS images are shown in Fig. 6), indicating coke deposited on the surface. Microstructure analysis by BET veri fi ed that the surface area and pore volume decreased from 10.21 m 2 /g and 0.025 cc/g (cycle 1) to 7.12 m 2 /g and 0.018 cc/g (cycle 5), respectively. Though the surface area and pore volume of used p-sludge chars were lower than those of new p-sludge biochar (52.3 m 2 /g and 0.033 cc/g), the catalytic ability was not affected substantially within four cycles (Fig. 3). The stable catalytic performance of p-sludge biochar was most likely due to the exposed Ca on the p-sludge biochar during cyclic use. Lime (CaO) is widely used in the pulp and paper industry in the recausticizing cycle [36]. The chemically-formed CaCO 3 is also used asa fi ller to improve paper quality [37]. Therefore, the settled paper mill sludge during paper mill wastewater treatment contains CaCO 3 [38]. XRD analysis con fi rmed the raw p-sludge and the new p- sludge biochar used in this study were mainly composed of CaCO 3 (XRD images are shown in Fig. 7). With the increasing use of p- sludge biochar (i.e. cycle 1 to cycle 5), the CaO peaks became more predominant, indicating some CaCO 3 was continuously decom- posed at 800 C in each cycle. SEM-EDS analysis showed that Ca was still detectable after fi ve cycles on the p-sludge biochar surface. The surface with exposed Ca ostensibly provided catalytic sites for pyrolysis vapor upgrading. CaO has been extensively studied on tar decomposition and is capable of upgrading pyrolysis vapor by destroying high molecular weight hydrocarbons [39,40]. Moreover, biochar that contains Ca proved to be very effective in reducing biosolids-derived bio-oil in our previous work [23]. Since surface Ca was not completely covered by coke, p-sludge biochar retained some catalytic activity after the fi fthuse. The TG (thermogravimetry) and DTG (differential
decrease was ostensibly due to the relative increase of H 2 and the biochar enhanced dry reforming as proposed by previous re- searchers [33]. The CO concentration was stable at 20 vol% to 25 vol % when using catalyst at high temperatures. The CH 4 , C 2 H 4 , C 2 H 6 , andC 3 H 8 concentrations were lower, but they still accounted for a signi fi cant percentage of the HHVs compared to H 2 and CO. These gases are typically generated by tar cracking at higher temperatures [34].
3.2. Impact of catalyst loadings on mass yields and py-gas composition
Py-gas yield increased as a function of biochar catalyst loading at 800 C (Fig. 2a). Wood and biosolids biochars were selected for subsequent comparison with p-sludge biochar since they produced high py-gas yield. At the catalyst/biosolids mass ratio of 0.33, p- sludge biochar and the other two biochar catalysts increased the py-gas yield to approximately 37%; the yield in the 0B control test was only 22%. With the increase of catalyst/biosolids mass ratio to 1, py-gas yield increased further. In particular, p-sludge biochar resulted in the highest py-gas mass fraction of 42% at a catalyst/ biosolids mass ratio of 1. The bio-oil yield concomitantly decreased to the lowest value of 18%. Py-gas composition was also affected by catalyst loading (Fig. 2b). The H 2 concentration did not change when using p-sludge biochar but the H 2 concentration increased when using wood and biosolids biochars. P-sludge biochar and the other two biochar catalysts increased CO concentration when the catalyst/biosolids mass ratio increased from 0.2 to 1. As aforementioned, the increase in CO concentration was most likely due to the steam methane reforming, dry reforming, and tar reforming processes and the decrease in CO 2 concentration was a result of the biochar enhanced dry reforming and the relative increase of other gas components [33].
3.3. Impact of catalyst cycle numbers on mass yields and product/ catalyst properties
Catalytic activity of the tested wood, p-sludge, and biosolids biochars decreased over fi ve cycles (Fig. 3), but p-sludge biochar possessed very stable catalytic capability within four cycles. There was no obvious drop in the py-gas yield, and py-gas fraction was constant at 39%. The corresponding bio-oil yield remained at 21%. After the fi fth cycle, the py-gas yield decreased to 33%, indicating that the p-sludge biochar catalytic activity decreased. The cyclic use trend of wood and biosolids chars were similar through four cycles. However, wood biochar fi nally lost all observable catalytic activity after fi ve cycles as indicated by the similar py-gas yield at 800 C0S control test (29%). The py-gas mass fraction catalyzed by biosolids biochar was slightly higher than that of the 0S control test after the fi fth use, implying almost no further catalytic effect. P-sludge bio- char could be used to process four times as much feedstock compared to wood and biosolids biochars. As for py-gas composition, the H 2 concentration was relatively stable using p-sludge biochar, ranging from 42 vol% to 45 vol% (Fig. 4). In contrast, the H 2 concentration decreased from 45 vol% to 30 vol% after fi ve cycles using wood and biosolids chars, approaching the results of the 800 C 0S control test. Since p-sludge biochar had very good catalytic capability and stability, the resulting bio-oils during cyclic tests (cycles 1 through
5
Made with FlippingBook - Online magazine maker