TREATMENT OF LOW STRENGTH WASTEWATER USING AN ANAEROBIC BAFFLED REACTOR

Abstract

In developing countries an urgent need is being felt to develop low-cost largescale, centralized as well as small-scale, onsite waste treatment systems that could provide good sanitation with minimal utilization of natural resources. Domestic wastewater is categorized as low-strength, complex wastewater, characterized by low COD (about 500 mg/L), high fraction of suspended solids (around 40%), and fluctuations in hydraulic and organic loading rates. Anaerobic baffled reactor (ABR) is one of the treatment alternatives developed recently. It can be described as a reactor made up of series of UASB reactors. It uses a series of vertical baffles to force the wastewater under and over them as it passes from inlet to outlet of the reactor. The reactor is simple in design, with no moving parts or mechanical mixing and relatively inexpensive to construct. The objectives of the present study were to evaluate performance of ABR for the treatment of low strength (a) soluble and (b) complex wastewaters separately by integrating various aspects like (i) variations in effluent characteristics over long operating period, (ii) performance of reactors at different organic loading rates (OLRs), (iii) compartment-wise variation of various parameters, (iv) hydrodynamics of the reactors, (v) compartment-wise morphology and granulation of bacterial cultures, and (vi) ability to resist organic and hydraulic shock-loads, in one single study. Two laboratory scale ABRs, each of 10 L capacity and eight compartments were fabricated. Synthetic wastewaters of total COD (CODT) of 500 mg/L and having different compositions (a) soluble wastewater and (b) complex wastewater [having soluble COD (CODs) of 300 mg/L and particulate COD (CODP) of 200 mg/L] were used as feed separately in reactors Ri and R2 respectively. Both reactors were started with 20 hour HRT (OLR = 0.6 kg COD/m3d). The HRTs of the first and second reactors were changed in steps to 16, 12, 8, 6, 10, 8 and 6 hour, and to 15, 10, 8 and 6hours respectively. Accordingly, OLRs varied from 0.6 to 2kg COD/m3d in both Ri and R2. Pseudo steady-state (PSS) condition (at each HRT or OLR) was assumed to have been achieved when variations in effluent COD values were found to be insignificant. The reactors Ri and R2 were operated for 678 and 592 days respectively. The COD and BOD removals were found to be over 88% for both soluble and complex wastewaters at OLRs ranging from 0.6 to 2 kg COD/m3d. Once reactors attained PSS, effluent characteristics were not found to vary significantly at different OLRs studied for both soluble and complex wastewaters. Inspite of almost identical COD and BOD removals in R\ and R2, generation of biogas or methane were found to differ significantly. In case of soluble wastewater, 74-85% of consumed COD got converted into CH4 COD while percentage conversion was only 41-55% for complex wastewater. Presence of particulate COD affected conversion of COD into CH4 COD. Maximum spatial COD reduction took place in first compartment while later compartments, 4 to 8 appear to have played no role. Spatial variation of parameters particularly pH and VFA indicated that bacterial cultures differed in different compartments. Based on tracer studies the dead space ranged from 19-25% and 23-34% in reactors fed with soluble and complex wastewaters respectively. Presence of particulate COD in synthetic wastewater increased the dead space significantly. It was observed that for both the wastewaters the flow pattern within the ABR remained intermediate between plug and perfectly mixed flows (irrespective of HRT or OLR). However, it in was found closer to plug flow than completely mixed flow. Nature of the COD (i.e. particulate or soluble) did not affect dispersion number significantly. Biomass granulation was observed in both the reactors. In the reactor fed with soluble wastewater it was observed at an OLR of 1.5 kg COD/m3d while in case of reactor fed with complex wastewater it was observed at 1.2 kg COD/m d. Different bacterial morphologies were observed in different sections of the ABRs, Ri and R2. With time, the size ofthese granules increased, particularly in the earlier compartments. Compartment 1 in both cases had the greatest variety of hydrolyzing and acid producing micro-organisms, predominated by clusters of rod-shaped bacteria and cocci of varying sizes. Methanosarcina dominated at higher substrate concentrations in Compartment 1, while Methanosaeta grew at lower acetate concentrations in Compartments 2 and 3. Organic shock loads revealed that maximum disturbances in characteristics of wastewater occurred in compartment 1. In general, appreciable changes in different parameters were observed only in first three to four compartments. During and after the organic shock load, characteristics in compartments 5to 8almost remained unchanged. The amount of sludge washed out from the system during the six-fold hydraulic shock load was found to be less than 10% of the sludge mass inside the reactors. The total effluent COD varied up to 450% during the transient conditions due to the heavy sludge washout. Reactors, R, and R2 recovered to initial conditions (in terms of effluent quality) in less than 9 and 11 hours respectively. The ABR has potentials and can prove a good alternative for the treatment of domestic wastewater. IV