STUDIES ON THE OPERATION AND PERFORMANCE OF UASB REACTOR FOR THE TREATMENT OF DISTILLERY SPENT WASH

Abstract

In India alcohol distilleries generate a large volume of waste water (spent wash) during the production of alcohol from molasses fermentation process. Around 15-20 1 of spent wash is generated per litre of alcohol produced. Based on the recent statistics about 18-24 billion litres of spent wash is produced per year in India. This distillery waste water/distillery spent wash (DWW/DSW) is highly acidic (pH:4.5-4.2) and is characterised by high BOD (40-65 g/1) and COD (80-120 g/1). The discharge of this spent wash in to water bodies without proper treatment can lead to serious pollution and health problems. According to the Central Pollution Control Board (CPCB), Delhi (India), the statutory standards for the discharge of DWW in to water bodies are : BOD: < 30 mg/1, COD: < 100 mg/1 and pH:~7.0. Among the alternatives available for the treatment of DSW, physico-chemical processes such as reverse osmosis, ultra filtration, incineration after concentration, animal feed application and spraying on to soils with a sprinkler irrigation system etc., are not feasible either because of the high cost or inadequate technical development. Among the biological treatment methods, traditional aerobic methods are unattractive because of the high energy-cost for aeration and the cost of disposal of the high amount of sludge generated, which could be as high as 50% of the total cost of treatment. The anaerobic treatment of DSW has the advantage of the recovery of inherent energy in the form of methane gas - a high calorie fuel - with much less sludge formation. Thus, the treatment of DSW by anaerobic processing becomes an operation in profit. However, the conventional anaerobic processing alone is not adequate enough to treat DSW as to allow the treated DSW to be discharged into water bodies or on soil. Thus, anaerobic digestion followed by aerobic treatment has gained wide acceptability in DSW treatment. However, efforts are being made to reduce the cost of aerobic processing by increasing the efficiency of anaerobic treatment. With this in view, several novel bio-reactors have been developed. Some of these are Down flow fixed film reactor (DFFR), Upflow Anaerobic filter (AF), Anaerobic fluidized bed reactor (AFBR), and Upflow anaerobic sludge blanket reactor (UASBR). All the above types of reactors work at very low hydraulic retention time (HRT) in comparison to solids retention time (SRT). The UASB reactor, conceived and developed by Lettinga and co-workers (1980) has received wide popularity in the treatment of high and medium strength organic wastewaters. Although UASB process has been used for the treatment of a variety of industrial and domestic waste waters, only a few reports are available regarding DSW treatment. Even these studies present a contradictory and confusing scenario on the treatment of DSW in UASB reactors with respect to the operational details and results. The present study is an attempt to present a cogent and a clearer picture of the UASB operation - its start-up and steady state operation for DSW treatment. The study was conducted in a UASB reactor, having a large settling chamber at the top with gas-disengagement and gas recovery provisions. The UASB reactor was fabricated in transparent plexiglass. It consisted of 1200 mm high reactor portion of 100 mmdiameter and a 600 mmhigh settler of 150 mmdiameter, attached on the top of the reactor. The reactor had a working volume of 9.75 litres. Along the height of the reactor, 6 ports were provided for sampling and other measurements. The inlet system consisted of 4 ports at the bottom which in turn was attached to a common manifold for uniform distribution of the feed. Inside the settler, a gas-liquid-solid separation system using a glass funnel and a gas pressuriser plexiglass iv tube was provided. The gas collection system consisted of two aluminium drums, the top drum being inverted in the other to float-up in water with gas collection. The reactor was fed with the help of a peristaltic pump. The present study consisted of three phases. During the first phase, the reactor was seeded prior to the substrate feeding, with digested sewage sludge from a conventional anaerobic sewage sludge digester after sieving it to remove debris. The reactor was started with the glucose + molasses solution (50% each) solution as the substrate at an organic loading rate (OLR) of 1.5 g COD/ l.day. Thereafter, the OLR was increased in steps. After 10 days of start-up the reactor was fed with molasses solution only. The aim was to granulate the sludge and acclimating it to take high organic loading. The feed was supplemented with a COD:N:P ratio of 300:10:1 by adding urea for nitrogen and KH2PO4 for phosphorous. The HRT was fixed to be 8 hours with an upflow velocity of 0.15 m/hr. The temperature of the reactor during start-up and steady state operation was maintained at 35 ±2°C. This was made possible by housing the reactor in a wooden environmental chamber and controlling the temperature with the help of a thermostat attached to a heat convector. In about 55 days, the seed sludge got granulated at an OLR of 8 g COD/ l.day. The sludge granules were angular in shape and were of sizes varying between 0.5-2 mm were found. The sludge VSS increased from the seed sludge concentration of 43.5 g VSS/ 1to 63.8 g VSS/ 1 at an OLR of 18 g COD/1.day. The setting velocity of the granules varied between 0.6 to 1.1 ml min with density varying 1060-1400 kg/m1. The SEM of the granules exhibited a mixed morphology on the surface of the granules while filamentous Methanothrix proliferation could be clearly seen in the interior of the granule. Methanosarcina and Methanothrix are the two dominant species found in the granules. After the formation of granules, the reactor was further loaded in steps. The reactor reached steady-state at every increment of load in 3-4 days indicating the stability and adaptability of the reactor. The reactor showed good performance and it could sustain an OLR of 18 g COD/ l.day with a COD removal efficiency of 88% and a gas production of 6.81 1/ 1reactor volume.day with a methane content of 75%. The sludge loading rate (SLR) at this stage was 0.92 g COD/ g VSS.day. The influent pH was maintained at 6.65 ±0.15 using sodiumJncarbonate in order to control the VFA concentration in the reactor to be within 500 mg/1, the operational limiting value for anaerobic process. In addition whenever the VFA increased beyond 500 mg/1, it was reduced by stopping the feed intermittently, for periods ranging from 2 to 6 hours. The reactor feeding was resumed once the VFA level got reduced below 500 mg/1. However, the maximum VFA level in this phase was 660 mg/1. During the second phase of operation, the substrate was switched over from molasses solution to distillery spent wash diluted to the required concentration with tap water. The starting OLR was 10 g COD/ l.day with a SLR of 0.54g COD/g VSS.dayThe HRT was continued to be maintained at 8 hours and the temperature at 35 ± 2°C. The COD:N:P ratio was maintained at 300:10:1 as before and the influent pH at 6.65 ± 0.15 by using sodium bi carbonate. The reactor sustained the change in the substrate very well without any drastic effect such as sludge washout or total stop of gas production. However, to begin with, the COD removal efficiency got reduced from the previous 88% to 54% on the first day of switching over to DSW. The gas production was about 2.58 1/1 reactor volume.day with a methane content of 54%. With increase in OLR, the COD removal efficiency increased and vl touched a maximum of 78% with an average of 75% at an OLR of 15 g COD/ l.day. The methane content got increased to 75% with a specific methane yield of 0.32-0.35 1/ g COD removed. The SLR at this OLR was found to be 0.651 g COD/ g VSS.day. On further increase of OLR in steps of 3 g COD/ l.day, the efficiency reduced and reached a minimum value of 58% at an OLR of 30 g COD/ l.day. The rate of increase in gas production decreased although overall gas production increased. The VFA level reached a value of 1080 mg/1. The SLR increased with OLR at a faster rate in the beginning, but it decreased after an OLR of 15 g/1.day. The final rate of gas production had reached 9.94 1/ 1 reactor volume.day but the methane content got reduced to 57%. The VSS got decreased from the initial 63.8% g VSS/1 to 53.87 g VSS/1 at an OLRof 12 g COD/ l.day. However, after acclimation to DSW it started increasing slowly and reached a maximum of 68.25 g VSS/1 at an OLRof 30 g COD/ l.day. Both, in the first phase as well as in the second, the sludge washout was observed to be insignificant. An OLR range of 15-18 g COD/ l.day was found to be the optimal range for the UASB operation with DSW as feed. However, an attempt was made to reduce the addition of this sodium bicarbonate, to economise the process by decreasing the influent pH stepwise, day by day. However, results showed that the influent pH could not be reduced below 6.5 since it started decreasing the performance efficiency. There was reduction of methane content in the biogas, indicating that the methanogenesis was being affected by pH reduction due to VFA build-up. The chromatographic analysis for the components of VFA namely, acetic-, propionic-, butyric- and caproic acids confirmed the fact that, for substrates like DSW, propionic acid is the most accumulated component. Hence the degradation of propionic acid could be the rate voi limiting step in the anaerobic stabilisation of DSW. The results of the study indicate that DSW can be treated effectively by adopting the UASB process; granulation of the raw seed sludge is possible with molasses solution as the substrate. The provision of a settler of any kind (increased cross section or same cross section) is necessary for the settling of the sludge as well as for preventing sludge washout. The UASB reactor can be operated with reasonably high load of 15 g COD/1.day with a good gas yield of 6.63 1/1 reactor volume.day having a methane content of about 75 %. In the third phase the reactor was operated continuously at an OLR of 15 g COD/ l.day for a period of 45 days. During this phase of operation the profiles of the working parameters namely temperature, pH, COD and VFA along the height of the reactor were determined. These profiles indicated a common feature that except in the inlet zone of the reactor, the values of these parameters were in the same range. This observation confirmed the fact that the reactor contents are well-mixed. This is due to the agitation caused by the rising gas bubbles and internal circulation in the reactor. Due to this mixing, the separation and demarcation of acidogenesis and methanogenesis inside the reactor is not possible. During this steady-state period, the COD removal efficiency varied between 75-78% while the BOD removal efficiency was between 85-91%. The total gas production was found to be 6.5-7.0 1/ 1 reactor volume.day. The influent DSW contained sodium, potassium, calcium and phosphorous (555 mg/1 as Na+, 367 mg/1 as K+, 35 mg/1 as Ca2+ and 19 mg/1 as P). The bacterial species could consume up to 74.8% sodium, 72.8% potassium, 65.7% calcium and 73.7% phosphorous. The effect of temperature on the performance of the UASB reactor was studied by increasing the temperatures to three different values-42°C, 45°C and 50°C. The results VXLX exhibited the foci that at 42nC there was no change in the performance efficiency of the reactor. At 45°C, there was a slight reduction in the COD removal efficiency (from 78 to 70%). At 50°C, the reactor showed a drastic reduction of COD removal efficiency by 50 %with indications of atotal shut down of gas production. With aview to study the effect of reduction in temperature on the performance efficiency, the reactor was kept at ambient temperature (27- 32°C). No significant effect in the performance of the reactor was observed during this change. This clearly indicated the wide mesophilic range of temperature (27 to 42°C) for UASB reactor operation. When, the reactor was exposed to the ambient temperature (6-10°C) showed an interesting, although, anegative change in the performance. The complete height of the sludge bed floated to the top of the reactor. After increasing the chamber temperature back to the 35 ± 2°C, the sludge bed started settling in parts to the bottom and attained the original level. Thus, the UASB process could treat high strength DSW with a good efficiency with minimum pre-treatment producing good quantity of biogas. Thus, the UASB process offers a viable treatment method for DSW with pollution abatement and energy recovery simultaneously.