Alternate: Effects of Primary Sludge Blending and Alum Addition on Thermal Hydrolysis-Enhanced Anaerobic Digestion of Waste Activated Sludge and Recalcitrant Dissolved Nitrogen Formation

Introduction The waste activated sludge (WAS) produced in Washington Suburban Sanitary Commission's (WSSC Water) Water Resource Recovery Facilities (WRRFs) is historically known to have relatively poor digestibility, even with thermal hydrolysis pretreatment (THP). In agricultural waste management, it has been a common practice to co-digest a poorly biodegradable feedstock with an easily biodegradable feedstock to improve the overall digestibility of the former (Karki et al., 2021). Since primary sludge (PS) generated in WRRFs usually possesses much better digestibility over that of the WAS, this study tested the hypothesis whether blending PS with WAS can increase the biodegradability of WAS. Because WRRFs need to use more coagulants such as alum to precipitate more PS for the blending, the impact of alum on the WAS biodegradability was also taken into consideration. Therefore, this study investigated the effects of PS:WAS blending ratio and alum doses on the WAS digestibility in a THP-anaerobic digestion system. Meanwhile, this study also tested the hypothesis that blending PS with WAS may produce more recalcitrant dissolved organic nitrogen (rDON) as a result of Maillard reaction in THP (Zhang et al., 2020). Maillard reaction occurs between the carbonyl groups of reducing sugar and the amino groups of proteins (Hodge, 1953). Since PS and WAS are mainly composed of carbohydrates and proteins, respectively, and the typical THP temperature at 165 0C is ideal for Maillard reaction, it is a reasonable prediction that more rDON may form within THP fed with PS and WAS blending than without. Materials and Methods Experimental design Dewatered WAS and PS cake samples with around 20% total solids (TS) were collected from two WRRFs of WSSC Water. Depending on the Al that came with each type of cakes, additional Al was supplemented according to the Al doses designed in Table 1. Basically, more Al was supplemented when more PS was blended to simulate the scenario when more alum was consumed for more PS production. A jar test was performed to estimate the Al dose it took to produce enough PS for the blending ratio in Table 1. Then, the PS and WAS cake samples were either individually processed in THP-anaerobic digester (AD) systems as controls or blended according to the blending ratio as summarized in Table 1 to study the blending effects during this three-phase study. All cake feedstocks were diluted to 16% TS prior to feeding into the CAMBI pilot THP system operated at 165 oC for 30 min. The THP effluent was further diluted to 9% prior to being fed into the four pilot mesophilic AD systems. These ADs were inoculated with fresh effluent sludge from a full-scale THP-fed mesophilic AD at DC Water. The steady-state digestates were analyzed for understanding the impacts of PS:WAS blending ratio and alum doses on WAS digestibility. In general, four phases of studies described in Table 1 were performed with controls to understand the effects of 1:1 blended with low Al in phase I, 1:1 Blended with high Al in phase II, and 3:1 Blended with high Al in phase III. rDON determination The four AD effluent samples from Phase II study in Table 1 were subjected to a rDON test. This test was divided into two steps including air stripping and aerobic incubation. The goal of air stripping was to reduce the total ammonia nitrogen (TAN) of the AD effluent to lower than 200 mg/L. This is because the TAN concentration is significantly greater than that of dissolved organic nitrogen (DON) and thus may overshadow the DON level readings which are calculated as the difference between soluble Kjeldahl nitrogen and TAN. The aerobic incubation aimed to remove the biodegradable DON, and in turn the remaining DON in the system can be regarded as rDON. Results WAS digestibility The volatile solid reduction (VSR) has been used as an indicator of sludge digestibility (Luo et al., 2022). Hence, the steady-state VSR of all digesters were determined. Assuming the VSR of PS does not change with and without blending, the blended WAS VSR can be calculated and compared to the control WAS VSR without blending and Al supplementation to assess whether the WAS digestibility has been improved. As can be seen from Figure 1, the WAS VSR did not change with and without blending with PS at a 1:1 ratio without external Al addition. This means blending PS at this low ratio does not improve the WAS digestibility. When a low dose of Al was added to increase the blended sludge background Al level from 1.49 to 1.95%, it can be seen that WAS VSR decreased about 2%, indicating the inhibitory effect of Al on WAS digestion. The reduced hydrolysis rate as a result of binding between Al and protein in WAS has been regarded as the reason accounting for the WAS digestibility decrease (Luo et al., 2022). This conclusion was reinforced by the fact that WAS VSR substantially decreased to 15% when further increasing the background Al level from 1.95 to 2.85% without changing the blending ratio (Figure 1). It is interesting to see that tripling the PS blending ratios can bring WAS VSR back to 28% even when the same high Al dose (2.85%) was used in the sludge mixture. These observations suggest that co-digesting PS with WAS does not directly improve WAS digestibility; however, adding more PS does mitigate the Al inhibition on WAS digestion. It was reported in a previous study that Al is expected to play greater binding effects on WAS than on PS because the affinity of Al to proteins in WAS is higher than that to the polysaccharides in PS (Luo et al., 2022). Therefore, PS might be able to prevent the binding between Al and protein in WAS by acting as a barrier. Hence, the reduced binding potential between Al and protein in WAS could mitigate the Al inhibition on WAS digestion. rDON The centrates containing DON generated from all four digesters in phase II study (Table 1) were exposed to aerobic incubation for 14 days. Profiles in Figure 2a showed that centrates from PS only and blended sludge with high Al addition contained the least DON. PS does not contain much protein and thus is reasonable to produce less DON. Al ¬ is known to be capable of precipitating DON (Zhang et al., 2020), which may explain the low DON level in centrate with high Al addition. In contrast, the DON levels in both WAS only and blending sludge without external Al addition were two times greater. The same trends were also observed with the residual DON, namely rDON in Figure 2b. It should be pointed out that the rDON level in blended sludge was actually substantially greater than the average level from the two individual sludge without blending, indicating extensive Maillard reaction might have occurred when PS and WAS were blended. Conclusions It can be concluded from this study that blending PS does not directly improve the WAS digestibility. Instead, blending more PS did mitigate the inhibitory effect of Al on WAS digestion. However, blending PS with WAS promoted rDON formation due to Maillard reaction between the two types of sludges.