Management of filtrate from the hydrothermal pre-treatment of sewage sludge using anaerobic digestion followed by chemisorption

Hydrothermal Carbonisation (HTC) could be a promising method for the complete treatment of human excreta as a community based, packaged process. In the work described here HTC was combined with anaerobic digestion (AD) following the separation of the liquid and solid fractions after the thermal treatment. The liquid fraction was shown to contain a high concentration of both dissolved and particulate organics. The main objective of this study was to evaluate the anaerobic digestion (AD) of this liquid by-product (wastewater) following HTC treatment of synthetic faeces and real sewage sludge. Methane production, to provide energy for the system, was compared with previous work on AD and theoretical gas yields. The results demonstrated that the HTC filtrate of sewage sludge and synthetic faeces as could be anaerobically digested, based on the standard AD performance indicators (Ripley s ratio, COD conversion to gas), were satisfactory and stable (COD removal of 59 % and 81%respectively). However, the biogas produced was low; below 0.01L/Lreactor/day for the sludge thermally processed in the lab, and 0.03L/Lreactor/day as the average yield from the Cambi® processed sludge. The synthetic faeces filtrate behaved differently, as its high COD concentration caused shock loads. Thus, dilution by recycling was used and proved a suitable method to overcome inhibitions, providing buffering and recycling of nutrients.10:1 recycling dilution resulted in the best AD performance, with average COD removal of 66%, biogas yield of 0.12 L/L reactor/d and Total Solids and Volatile Solids removal 81 and 88% respectively. Digestion of the whole fraction (representing the usual practices until now, CAMBI, Veolia), gave similar methane yield as the filtrate digestion, thus the extra COD entering with the whole fraction did not benefit the process with extra methane production. The type of filter used after HTC process seems also important for the AD following; COD concentration of the effluent using 60 μm increases 3 times compared to 10 μm use. Focus was given to the influence of HTC reaction conditions (temperature and time) on liquid products characteristics and the following AD process. It was concluded that milder thermal conditions generated more gas. The data also shows that low AD hydraulic retention time (HRT), typical of high rate fixed biomass digesters can be used to treat the HTC filtrate. Halving the anaerobic digestion HRT to 0.9 days resulted in 1.8 to 6.8 times greater biogas yield (0.31-0.86 L/Lreactor/day). Taking into consideration also the char production effectiveness a case was made for running the HTC at 180oC for 30 min which also resulted in high biogas yield when the HRT was short (0.9 days). GC-MS analysis followed the observations of the AD experiments showing that HTC temperature but also residence time are crucial for the extent of HTC reactions occurring. The more intensive the HTC conditions, the more complex and greater the extent of Maillard reactions. Standard biodegradability tests, respirometry and BOD, also supported previous work that HTC conditions at higher temperatures and durations reduced biodegradability. Levels of COD in the digester effluent were still high, indicating further treatment would be required if release into the environment or recycling is considered. The work described here evaluated the further COD reduction using adsorption. The hydrochar produced during the experiments of sewage sludge and synthetic faeces were compared to activated carbon. The alkali activation was essential in order to stabilise the performance and reach a satisfactory efficiency. Specifically, KOH activation increased the specific surface area for the sewage sludge char achieving COD removal of 75.2%, while the KOH- HCl treatment did not improve the performance further (76.4%) despite a measured increase in surface area. In contrast adsorption by the char produced from synthetic faeces was improved by HCl activation to COD removal of 79.3% compared to 59.6% when only KOH activation was applied. The highest adsorption capacity was reported for PAC (>90%). The nature of this adsorption reaction was presumed to be an anion sorption, highly dependent on the pH, with higher removal at low pH values, for all the materials except the PAC. A pseudo-second order kinetic model described the results satisfactorily. Regarding the effect of the adsorbent dose, the BET isotherms type II and IV modelled the results suggesting an initial monolayer followed by a multilayer adsorption. The PAC results on the other hand followed the Freundlich isotherm. However, further investigation on these reactions is required to improve the understanding of the kinetic and adsorption models and their consequent application improve implementation.