PHENOLIC WASTEWATER: BMP AND TREATMENT USING UASB REACTOR

Abstract

Phenols are the major organic constituents of effluents from coal conversion processes, coke ovens, petroleum refineries, phenolic resin manufacturing, herbicide manufacturing, fiberglass manufacturing and petrochemicals. •The concentration of phenols in. effluents varies from 10 mg/1 to 17 x 10 mg/1. In generaJL. @®p contributed by phenolic compounds ifi these effluents fanges from 40%.-»80% of the total COD. Coal conversion process effluents and coke oven effluents contain on ar» average $0% phenol arid 30% cresols (Cooper and Wheatstone, 1973; ^Fedorak and Hrudey, 1966; Nakhla and Suidan, 1995). • ' , •'':.'' '."• Phenols are toxic, carcinogenic, mutagenic and teratogenic {Autenrieth, et al., 1991). They are growth inhibitory to microorganisms in biological treatment processes and regarded as priority pollutants in the USEPA list. A phenol/Concentration of 1 mg/1 or greater affects aquatic life. Therefore, in most cases stringent effluent discharge limit of less than 0.5 mg/1 is imposed (Chang et al., 1995; Tay et al., 2001). Removal of phenols can be accomplished through physical, chemical and biological processes. Recovery of phenols is economical at high concentrations by physical processes (e.g., solvent extraction, activated carbon adsorption). At intermediate concentrations, i.e., phenol concentrations ranging from 5 - 500 mg/L (Patterson, 1975), biological oxidation techniques are feasible. In general, physicochemical processes are employed as pretreatment or post treatment units coupled with biological processes in order to reduce the concentration of phenols to effluent discharge limits. Prior to 1990, the treatment of phenolic wastewaters by aerobic processes received considerable attention (Table 2.2). The situation has reversed since 1990 and emphasis has been more on anaerobic treatment. Among highrate anaerobic processes, UASB has attracted a lot of attention worldwide. Anaerobic degradability of a substrate in a UASBR depends on the nature of the wastewater and operational parameters such as sludge bed height, type of the sludge, OLR, SLR, HRT, etc. A reactor treating toxic wastewater exhibit high sensitivity to changes in operating conditions. Phenol as a sole substrate up to a range of 500 - 750 mg/L is not inhibitory to UASB process. The toxic effects can be mitigated by using cosubstrate (Grady, 1990). Phenol concentration more than 500 mg/L can be effectively treated either with recirculation of the treated effluent or supplementing with cosubstrates such as glucose, VFA, etc (Fang et al., 1996; Tay et al., 2001;Kennes et al., 1997). Use of readily degradable wastewater as a co-substrate may have wider acceptability than pure substrates. Considering these, a necessity was felt to further investigate the concept of treating slowly degradable or toxic organics in the presence of a readily degradable wastewater. This can be done by either increasing the toxicant concentration in the readily degradable wastewater or vice versa. It has therefore led to the formulation of a project to investigate the anaerobic degradability of phenolic wastewater mixed with readily degradable molasses based wastewater. The feasibility of treating phenolics with or without molasses wastewater has been explored in bench scale batch reactors and continuous UASBR. The objective of study is to evaluate anaerobic degradability of a wastewater containing a wide range of phenolic fractions with granular sludge as seed inocula. The scope of the work is given as under: Bench Scale Batch Studies Set I - Fate of seed sludge with and without substrate Set II - BMP of Molasses based Wastewater (SWW) containing different fractions of (i) Phenol (ii) o-cresol (iii) m-cresol (iv) p-cresol Set III - Effect of F/Mon anaerobic degradability of phenol and molasses based SWW Continuous Reactor Studies • Start up ofUASBR with dilute molasses as a feed • Performance evaluation of UASBR at each step increase in phenol COD fraction in the feed • Process failure and restoration • SMA of reactor sludge at pseudo-steady state at different fraction of phenol COD. in The thesis has been organized into 5 chapters: Chapter 1: Introduction Chapter 2: Literature Review Chapter 3: Materials and Methods Chapter 4: Results and Discussion Chapter 5: Conclusions Chapter 1 deals with the introduction, identification, justification and scope of the problem as briefly mentioned above. Chapter 2 presents literature on the use of physical, chemical and biological methods for the treatment of phenolic wastewaters. With increasing recognition to the UASB process, the treatment of phenolic wastewaters using high-rate anaerobic processes in general and UASB in particular has been reviewed in detail. During the course of the present project, Tay et al. in 2001 reported the response of UASBR to phenolic wastewater with glucose as a cosubstrate. The glucose supplement was shown to increase the operational stability under the conditions of shock organic loading or temperature change. Chapter 3 deals with the materials and methods adopted during experimental investigations. Experiments were conducted concurrently in two phases: namely batch studies (Phase I) and Continuous reactor studies (Phase II). To ascertain the blank correction in batch BMP bioassay test, the net growth of microorganisms (sludge) was monitored in the presence and absence of substrate in batch reactors over a period of one month. Eight bottles charged with 200 mg of easily degradable substrate and inocula along with equal number of seed blanks were incubated at 30 ± 2°C. At regular interval of time (4 - 5 d) the contents of (i) a serum bottle with substrate and (ii) a serum bottle without substrate were analyzed for total COD (TCOD), soluble COD (SCOD), total suspended solids (TSS) and volatile suspended solids (VSS). Biogas and methane were regularly monitored. A series of batch experiments were conducted to determine the BMP of wastewater using serum bottle technique (Owen et al., 1979). A synthetic wastewater (SWW) was prepared by diluting molasses with tap water. Each serum bottle was iv charged with a feed containing SWW and a phenolic compound in varying amounts. A total of 200 mg COD (equivalent to 2000 mg/L) fed to each bottle contained 0%, 2%, 7%, 12%, 17%, 22%, 29%, and 32% (Run I) and, 0%, 40%, 50%, 60%, 70%, 80%, 90%, 100%(Run 2) of COD contributed by (i) phenol (ii) o-cresol (iii) m-cresol and (iv) p-cresol. Experiments were terminated after 30 days. Granular sludge from a UASBR treating SWW was used as inocula. Serum bottle experiments were carried out in duplicate at 30± 2°C and F/M = 1.0. A bottle without any phenolic or non-phenolic substrate (designated as seed blank) was also incubated for monitoring background CH4 production. Total of 144 i.e., {(9x2) x4} x2 serum bottles were monitored for cumulative methane production in eight batches of 30 days each with an interval of 5 days in between. An additional run of experiments with digested sludge (from a conventional digester of a sewage treatment plant) as inocula was conducted with substrate containing 0% - 32% of phenol COD (18 bottles). Residual phenol in the content of each serum bottlewas measured on termination of experiments. Effect of varying amount of inocula (sludge) on the biotransformation of (i) phenol to methane and (ii) dilute molasses to methane was evaluated at F/M (substrate to microorganism ratio) of 0.25, 0.75, 1.0, 1.5, 1.75 and 2.0. Total of (7x2) x2 serum bottles including seed blanks were monitored for biogas production at regular intervals. Monitoring was continued till the gas production ceased (around 12 days). Bench scale UASB reactor with a working volume of approximately 12 L was used to study the treatability aspects of SWW containing phenol. Seed to the reactor comprised of a mixture of granular and digested sludge at 50:50(v/v) ratio. Molasses diluted in tap water constituted the feed during start up of the reactor. Macronutrients, N and P were provided in the ratio of 300:10:1(COD: N: P). Sodium bicarbonate was used to maintain pH ~ 6.5 - 7.5 within the reactor. Phenol mixed with molasses based SWW in different proportions comprised the phenolic wastewater. Specific methanogenic activity (SMA) of the sludge was carried out with acetate as substrate. Sludge samples were drawn for SMA analysis at pseudo steady state corresponding to every step increase in phenol COD in the feed. Sludge in the serum bottle was fed with acetate to evaluate SMA. Substrate to sludge ratio varied from 0.5-1.0. Initial VSS concentration of 2000 mg/L and 4000 mg/L were provided. Observations made from experimental investigations are discussed in Chapter 4. Results are organized according to the sequence reported in Chapter 3. The substrate (SWW) did not alter the fate of sludge inocula. Reduction in TSS and VSS was of the same order of magnitude in reactors with and without substrate. Soluble COD was reduced to 50% in comparison to 30% reduction in solids over a period of one month. The rate constant corresponding to total COD and solids reduction was comparable with and without substrate. Degradation of solids and COD were in agreement with linear as well as exponential models. The exponential degradation constants, kj, for the sample and the blank were of the same order of magnitude. However, the reduction rates, (slope of the line (mg/L.d)) for the blank and the sample in the linear models are significantly different. While both linear and exponential models are acceptable for the observation made during incubation period, the exponential model has an acceptance over a wide range of concentration over a wider period. BMP of substrates containing phenol COD fraction up to 0.17 with digester sludge, as seed inocula did not vary significantly. The granular sludge was found to be more resistant to inhibition due to phenol COD. Cresols however are more inhibitory than phenol. Phenol COD > 70% was inhibitory to methane production. BMP5 with digester sludge was higher in comparison to granular sludge at respective phenol COD. The cumulative methane production from the substrate and digester seed inocula was in conformity to the first order rate equation: Yt = Ym(l-e'kt) where Ym and Yt are ultimate methane yield and cumulative methane yield at time t, respectively. However, the time response of methane generation from granular seed sludge did not follow the same trend. The rate of methane generation with digested sludge is initially fast and is reduced considerably with depletion in substrate whereas with granular seed as inocula, the rate of methane generation is initially slow and subsequently becomes fast. The gas generation from a reactor with granular sludge represents an accelerative reaction with time lag in methane collection. The data is empirically related to the following equation: ' r 2 VI Where Y, = CH4-COD at time t, mg/L; kr - Rate of change of CH4-COD (mg/L) production rate, mg/L.d2; t = time, d The mechanism of substrate utilization by digested seed sludge and granular seed sludge appears to be different. The diffusion of substrate across the compact granular sludge may be the rate-limiting step. A few results from different substrates are compiled in Table 1. The inhibitory effects of phenolics follow an order given as under: m-cresol>o-cresol~p-cresol>phenol Table 1. BMP30* of Molasses based SWW containing phenolic COD at varying proportions. Phenolic compound —• Phenol o-cresol m-cresol p-cresol % Phenolic COD in substrate J Digested Sludge Granular Sludge Granular Sludge Granular Sludge Granular Sludge 0.0 0.703 0.729 0.643 0.503 0.611 17.0 0.717 0.749 0.466 0.217 0.540 50.0 0.286 0.109 0.066 0.040 100.0 -0.329 -0.406 -0.454 -0.417 ¥ - g CH4-COD /g COD fed The effect of F/M was noticeable in the case of phenol as a sole substrate. The degradation of wastewater from molasses however, was not influenced by F/M. The inhibition depends on the toxicant to sludge ratio more than the ratio of toxicant to SWW. Sludge acclimatization is facilitated over an extended period of incubation. The specific rate of methane generation with adapted sludge is high. Results of continuous reactor studies are discussed under three headings: i) Start up ii) Evaluation of maximum OLR and iii) Treatment of phenolic wastewater. Reactor was operated over a period of 632 days. Start up operation lasted for 55 days. Subsequently OLR was gradually increased from 4g /L.d to a maximum of 18 g/L.d either by increasing the influent COD concentration and/or by reducing the HRT. SLR was maintained at 0.8 - 1.2 g/g.d. Sludge bed height was around 30±3% of the reactor height. Reactor performance was evaluated by regular monitoring of biogas production, % CH4 in biogas, COD removal efficiency and sludge washout. At OLR ~ 18.0 g/L.d Vll (i.e., Influent COD concentration ~ 6000mg/L and HRT- 8h), sludge washout (-9-12 g VSS/day) was noticed. OLR was reduced to 12 g/L.d by increasing the HRT from 8 h to 12 h. At OLR of 12 g/L.d, feed composition was changed from 100% molasses COD to 10% phenol COD and 90% molasses COD. Due to poor phenol removal efficiency (-20% - 40%) of the reactor, probably due to preferential uptake of easily degradable substrate, OLR was further reduced to 8 g/L.d (Influent COD - 4000 mg/L and HRT - 12 h) by maintaining the phenol COD at 10% of feed COD (i.e., phenol COD equivalent to 400 mg/L). Performance evaluation was made with respect to gas generation, methane %, COD reduction and phenol removal. Reactor exhibited phenol removal efficiency of-99.9% at steady state. Phenol COD was gradually changed step wise with an incremental increase of 10% of feed COD after the reactor attained pseudo steady state and exhibited at least 80% COD removal. Up to 70% of phenol COD (i.e., phenol COD - 2800 mg/L and molasses COD - 1200 mg/L) in the influent, reactor exhibited phenol removal efficiency of > 90% and COD removal efficiency of over 80%. SLR and sludge occupancy were maintained at 0.8 - 0.9 and - 30% respectively by regular desludging of the biomass from sludge bed. At 80% phenol COD, reactor performance deteriorated gradually. Due to desludging operation, the performance deteriorated abruptly and gas production ceased. Some amount of sludge was added into the reactor and operated by increasing theHRT from 12 h to 24 h. Dense nature of sludge, lowering of HLR and low biogas production rate probably resulted in piping conditions, which did not improve the performance of the reactor. Reactor was revived by lowering the phenol COD to 70% and reducing theHRT from 24 h to 12 h. Once the reactor attained pseudo steady state, HRTwas increased from 12h to 24 h. Once again, reactor performance deteriorated gradually as noticed from gas production. It appeared that HRT of 12 h was necessary to maintain the sludge bed under expanded condition and to facilitate uniform mixing of substrate and preventing piping condition. The reactor was revived by feeding SWW containing diluted molasses (COD - 4000mg/L) at HRT ~12h. After reaching pseudo steady state, phenol was introduced with phenol COD corresponding to 50% of the feed COD. Due to inhibitory nature of phenol at this concentration, phenol COD was gradually reduced to - 30% and acclimatized gradually to higher phenol COD. After attaining pseudo steady vin state at 50% phenol COD, OLR was increased step wise from 8g /L.d to a maximum of 13g/L.d by increasing the fded COD. However, phenol COD was maintained at 50% of feed COD. Reactor exhibited COD removal efficiency of50% - 60%. Phenol COD > 70% is inhibitory to biomass. This finding is further corroborated by the fact that the specific methanogenic activity (SMA) of granular sludge exhibited uniform activity up to phenol COD of 70%, at pseudo steady state conditions during step increase in phenol COD. Conclusions based on analysis are summarized below. • Decay rate of sludge in the presence or absence of substrate is same. High background CH4 from seed blanks with granular sludge as inocula may be attributed to the entrapped SCOD within the granular sludge. • Molasses based SWW containing phenol yield higher methane potential as compared to cresols. With granular sludge as inocula, BMP of SWW with phenol COD up to 32% was similar to SWW containing easily degradable substrate such as diluted molasses. Granular sludge had a higher resistance to phenol shock as compared to digested sludge and higher COD conversion capacity to methane at increasing phenol COD concentration. The extent of inhibition of phenol and cresols to CH4 generation may depend upon the ratio of phenolic and non-phenolic (easily degradable) COD. • The methane generation from granular sludge as inocula suggests the possibility ofusing a pre-reactor to convert phenol to readily degradable substrate. Arising out of this, it is suggested to compare the single stage treatment of phenolic wastewater with two-stage treatment. • Significant delay (lag) in gas collection was noticed with granular sludge. This lag increased with increase in phenol to biomass ratio. • In continuous UASB reactor, at influent COD concentration of 4000 mg/L and HRT - 12h, phenol COD up to 70% in the feed can be treated effectively with removal efficiency >90%. Step increase in 10% phenol COD would allow the sludge to easily acclimatize to the increased phenol concentration. SMA of granular sludge exhibited uniform activity up to phenol COD loading of 70% in the reactor. IX Present study deals with BMP of SWW containing phenolic compounds in molasses based SWW. Granular sludge fed on molasses based SWW is used as inocula. SWW may not give a true representation of the actual wastewater. BMP of phenolic waste also depend upon the culture history of the sludge. BMP of a given wastewater with biomass previously acclimatized to the particular wastewater would give higher CH4 yield as compared to unacclimatized biomass.