Strategies for recovery of imbalanced full-scale biogas reactor feeding with palm oil mill effluent

Characteristics of inhibited AD sludge, POME, biogas effluent, and active methane-producing sludge

An inhibited sludge sample was collected from the mesophilic biogas plant (40 ± 2 °C) of a palm oil mill, Prasang Green Power Co., Ltd., Surat Thani Province, Thailand. The biogas reactor was operated in continuous mode, feeding with POME at a high OLR (4.5 g-COD/L/d), resulting in acidification of the AD reactor. Biogas effluent, POME, and active methane-producing sludge were collected from the biogas plant at Pitak Palm Oil Co., Ltd., Trang Province, Thailand, and analyzed for their characteristics, according to the procedure described in APHA (2012). The inhibited sludge had an acidic pH (3.9), 4.8 g/L of total volatile fatty acids (tVFA), 17.0 g/L of suspended solids (SS), and 14.5 g/L of volatile suspended solids (VSS). The active methane-producing sludge had neutral pH (7.5), very low tVFA (0.92 g/L), but a high SS and VSS content of 59.8 g/L and 52.2 g/L, respectively.

Recovery of AD process imbalance

Experiments carried out in a batch reactor, an AD process imbalance, indicated by inhibited sludge, were recovered using three strategies. Firstly, the inhibited sludge sample was diluted with tap water (TW) and biogas effluent (BE) at a ratio of 9:1, 8:2, 7:3, 6:4, and 5:5, respectively, as a dilution strategy. Secondly, the pH of inhibited sludge was adjusted using 0.85–1.50% w/v sodium hydrogen carbonate (NaHCO₃), 0.10–0.14% w/v sodium hydroxide (NaOH), 0.10–0.50% w/v calcium hydroxide Ca(OH)₂ and 6.0–10.0% w/v oil palm ash as pH adjustment strategy. Thirdly, the inhibited sludge was recovered by adding active methane-producing sludge at 5, 10, 15, 20, 25, 30, 35, 40, 45, and 50% v/v called re-inoculation or bioaugmentation strategy. All strategies were combined with 20% v/v POME addition as low flow rate feeding (40 m³-POME/d), instead of stop feeding, and self-recovery was used as a control. All strategies were tested at an inhibited sludge sample concentration of 1 g VSS. All experiments were flushed with N₂: CO₂ mixed at 80:20 ratios to create the anaerobic condition and secured tightly with a butyl-rubber septum and aluminum cap. The experiment was carried out in triplicate and at a temperature of (40 ± 2°C) for 45 days. The biogas production in the headspace was measured via the water displacement method, and the biogas content was analyzed by a gas chromatograph equipped with thermal conductivity detectors (GC-TCD). Microbial sludge from each treatment was analyzed for the microbial community structure using polymerase chain reaction denaturing gradient gel electrophoresis (PCR-DGGE) techniques. The specific methanogenic activity (SMA) of the inhibited sludge and recovery sludge was also determined (Hussain & Dubey, 2017). The recovery time is the incubation time for methane production reaches 90% of maximum methane production in the recovery experiment and simulation results (Wu et al., 2015).

The validity of lab-scale results in the full-scale recovery of AD process imbalance

The inhibited sludge from a 6,000 m³ full-scale hybrid cover lagoon reactor was tested in the lab-scale experiment for recovery strategies. The selected strategy (dilution with biogas effluent) was applied to a 6,000 m³ full-scale hybrid cover lagoon reactor (Prasang Green Power Co., Ltd., Surat Thani Province, Thailand) for recovery of the AD process imbalance and validated the lab-scale results. The biogas reactor was operated at hydraulic retention times (HRT) of 30 days. The organic loading rate (OLR) was reduced from 4.5 kg COD/m³/d to 1.25 kg COD/m³/d. The 50 m³ of inhibited sludge and 10 m³ of POME were diluted with biogas effluent at a ratio of 8:2 every day for two weeks before adding to the full-scale hybrid cover lagoon reactor. The reactor was operated normally when the pH of inhibited sludge was increased to 7.5. Biogas effluent with pH 7.8 was collected from another biogas reactor effluent with good performance at Prasang Green Power Co., Ltd., Surat Thani Province, Thailand. POME was collected from the palm oil mill at the biogas plant at Prasang Green Power Co., Ltd., Surat Thani Province, Thailand. The pH and methane production rate of full-scale biogas sludge were monitored daily. The SMA of full-scale biogas sludge was monitored every 3 days. Acetate was used as a substrate during SMA tests as representative acetoclastic bacteria to investigate the reactor performance during the recovery process.

Microbial activity and microbial community analysis

The inhibited and recovered sludge methanogenic activities were evaluated by the SMA test using avicel (cellulose), glucose, gelatin, and acetic acid as a representative microbial population group of hydrolytic bacteria, acidogenic bacteria, proteolytic bacteria and, methanogenic archaea, respectively. This evaluates the anaerobic sludge capability to convert an organic substrate into methane, escaping quickly to the gas phase, reducing the COD in a liquid phase. The SMA value was calculated by the slope of methane production (based on g COD of CH₄) against incubation time and divided with VSS added of sludge sample (Hussain & Dubey, 2017). The microbial community structure was analyzed by polymerase chain reaction denaturing gradient gel electrophoresis (PCR-DGGE) techniques, according to Prasertsan, O-Thong & Birkeland (2009). The 0.2 g of the sludge sample was extracted for genomic DNA using the Ultraclean Soil DNA Kit (MoBio Laboratory Inc., USA). The 16S rDNA gene of bacteria was amplified by the first PCR with universal primer 27f (GAGTTTGATCCTTGGCTCAG) and 1525r (AAGGAGGTGWTCCARCC). 16S rDNA gene for archaea was amplified using Arch21f primers (TTCCGGGTTGATCCYGCCGGA) and Arch958r (YCCGGCGTTGAMTCCAATT). The V3 region of bacteria was amplified in a second PCR by primer 357f (CTCCTACGGGAGGCAGCAG) with CG clamp and 518r (GTATTACCGCGGCTGCTGG), using the first bacteria PCR as a template. The V3 region of archaea was amplified in a second PCR by primer 340f (CCTACGGG-GYGCASCAG) with CG clamp and 519r (TTACCGCGGCKGCTG), using the product of first archaea PCR as a template. Second PCR products performed the denaturing gradient gel electrophoresis (DGGE) analysis with 6% polyacrylamide gel for bacteria and 8%polyacrylamide gel for archaea containing a linear of urea/formamide gradient with denaturant ranging from 40% to 70% in 0.5 TAE buffer at 20 volts for 20 min, and 70 volts for 15 h, at a constant temperature of 60 °C. Sybr-Gold was stained in the DGGE gels for 60 min and photographed on the Gel Doc XR system (Bio-Rad Laboratories). Predominant DGGE bands were excised with a sterile tip and suspended in 30 µL sterilized Milli-Q water. Excised DGGE band was incubated at 4 °C overnight and re-amplified by PCR using the same primers without the GC clamp. PCR products were purified and sequenced by Macrogen Inc. (Seoul, South Korea). Closest matches for partial 16S rRNA gene sequences were identified by database searches in Gene Bank using BLAST (Tatusova et al., 2016).

Analytical methods and calculation

The biogas composition (H₂, N₂, CH₄, and CO₂) was determined by gas chromatograph GC-8A (Shimadzu, Kyoto, Japan) with a 1-meter stainless steel column packed with Shin Carbon (60/80 mesh) equipped with thermal conductivity detectors (TCD). The argon at a flow rate of 14 mL/min was used as the carrier gas. The temperatures of the oven, detector, and injection port were at 40 °C, 100 °C, and 120 °C, respectively. The 0.5 mL of the biogas sample was injected in duplicate for each reactor. The daily biogas production for each reactor was counted using the water displacement method (Yan et al., 2015). The chemical and physical composition of POME, biogas effluent, active methane-producing sludge, and inhibited sludge were determined for pH, lipid content, total solids (TS), volatile solids (VS), volatile suspended solids (VSS), total nitrogen (TN), total volatile fatty acid (tVFA), and alkalinity according to Standard Methods for the Examination of Water and Wastewater (APHA, 2012). Determination of TS was performed at a temperature of 90 °C instead of 105 °C, till constant weight to avoid decreasing of VFAs (Angelidaki et al., 2009). The VFAs composition was determined through a gas chromatograph GC-17A (Shimadzu, Kyoto, Japan) with a Stabilwax^®-DA fused silica column (30 m of length, 0.53 mm of diameter, 85 °C) connected to a flame ionization detector (FID) at 240 °C. The helium at 30 mL/min was used as the carrier gas. The VFA samples were collected by syringe (one mL) and filtered through a nylon membrane (0.2 µm). The filtrated samples were acidified to pH 3.0–3.2 with 30% (v/v) phosphoric acid for VFAs analysis Raposo et al. (2015). Buswell’s equation was used to calculate theoretical methane yield, assuming the total stoichiometry conversion of the organic matter to methane and carbon dioxide (Buswell & Mueller, 1952). The hydrolysis constants (k_H) were determined by using the first-order kinetic model in Eq. (1) according to the protocol of (Raposo et al., 2006). Where B(t) is the cumulative methane yield (mL-CH₄/g-VS_added) at time t, and B_∞ is the maximum cumulative methane yield (mL-CH₄/g-VS_added). k_H is the hydrolysis constant (d⁻¹); t is the fermentation time; λ is the lag phase (day). (1)${\displaystyle \mathrm{B(t)}={\mathrm{B}}_{\infty }\left[1-exp\left(-{\mathrm{k}}_{\mathrm{H}}\left(t-\lambda \right)\right)\right].}$

The modified Gompertz model was used to predict the methane production, methane production rate, and lag phase (Nopharatana, Pullammanappallil & Clarke, 2007) as follows in Eq. (2). (2)${\displaystyle \mathrm{M}=\mathrm{P}.exp\left\{-exp\left[\frac{\mathrm{Rme}}{\mathrm{p}}\left(\lambda -\mathrm{t}\right)+1\right]\right\}}$ Where M is the cumulative methane yield at the time t (mL-CH₄/g-VS_added); P is the maximum cumulative methane yield (mL-CH₄/g-VS_added); Rm is the maximum methane production rate (mL-CH₄/g-VS/d); λ is the lag phase (d); t is the fermentation time, and e is the Euler constant (2.718282). The economic evaluation of each recovery strategy was calculated from the chemical addition cost, human resources, energy consumption, biogas loss, and biogas production during the recovery period according to the current market price and the average price of industrial water in Thailand. Biogas price is 0.21 USD/m³. The cost of NaOH, Ca(OH)₂, and NaHCO₃ was 0.35, 0.2, and 0.3 USD/kg, respectively. The cost of human resources was 10 USD/people/d. The cost of electricity consumption was 0.13 USD/kWh. Biogas loss is the average daily biogas during a good performance period (3,000 m³/d) minus biogas production during the recovery period.

Recovery of AD process imbalance

The AD process imbalance is caused by a low pH and high VFAs accumulation due to organic overload and low degradation efficiency. Inhibited sludge had low pH (3.9) and a high total VFA (4.8 g/L) with butyric acid (2.5 g/L) and acetic acid (1.8 g/L) as the main VFA (Table 1). It also had a low SS and VSS of 14.5 g/L and 11.0 g/L, respectively, indicating a low number of anaerobic microorganisms in the systems. The specific methanogenic activity (SMA) of the inhibited sludge with glucose, acetic acid, avicel (cellulose), and gelatin was 0.208, 0.344, 0.401, and 0.065 gCH₄-COD/gVSS/d, respectively (Fig. 1). The SMA of acetic acid (0.344 g-CH₄-COD/g-VSS/d) was lower than the SMA of avicel (cellulose), indicating a low number of methanogenic archaea for reducing acetate to CH₄ and CO₂ but a high number of hydrolytic bacteria and acidogenic bacteria for acid production. The pH of inhibited sludge was gradually increased from 5.7 to 7.8. Data files regarding SMA analysis of inhibited microbial sludge, self-recovery sludge, and all recovered strategy were showed in Data S1. The acidic pH directly affected microbial activity, leading to a low methane production rate (40.8 mL-CH₄/d) and extended recovery time (49.4 days). The self-recovery had low hydrolysis constant (k_H) (0.005 d⁻¹) with a long lag phase (21.2 days) and low methane yield (209 mL-CH₄/g-VS_added) were observed in self-recovery (Table 2). The specific methanogenic activity of self-recovery was lower than those of the inhibited sludge for cellulose and glucose as substrate (0.112 and 0.106 gCH₄-COD/gVSS/d, respectively) but higher for acetic acids and gelatin substrate (0.481 and 0.122 gCH₄-COD/gVSS/d, respectively) (Fig. 1). The SMA test using avicel (cellulose), glucose, gelatin, and acetic acid as a substrate is a representative microbial population group of hydrolytic bacteria, acidogenic bacteria proteolytic bacteria, and methanogenic archaea, respectively. Methane production was achieved from self-recovery but the low activity of specific methanogenic activity, indicating low active methane-producing microorganisms in the self-recovery strategy. A more detailed biogas production data of each recovery strategy and self- recovery will be provided in Data S2.

The recovery by dilution strategy with biogas effluent (BE) gave higher methane yield (214–282 mL-CH₄/g-VS_added) and methane production rate (53.4–108.0 mL-CH₄/d) than the recovery by dilution with tap water (TW) (177-190 mL-CH₄/g-VS_added and 40.7–62.5 mL-CH₄/d, respectively) and the self-recovery (Table 2). The lag phase of recovery by dilution with BE (7.1–12.3 days) and dilution with TW (9.9–15.1 days) were significantly shorter than self-recovery (21.2 days). The recovery time of dilution with TW and BE was in the same range (33.2–41.7 days and 32.8–42.0 days, respectively). The dilution with TW could reduce toxicity in the inhibited sludge so that lag phase and recovery time could be decreased, but not enhanced methane-producing microorganisms resulting in low methane yield and methane production rate. The recovery by dilution with BE at a ratio of 8:2 to 5:5 had a 20–32% shorter lag phase and recovery time than the self-recovery. The recovery by dilution with BE could enhance active methane-producing microorganisms in sludge and resulted in a 2.2 fold increase of the methane production rate (91.4–92.8 mL-CH₄/d compared to 40.8 mL-CH₄/d, respectively). The recovery by dilution with BE had specific methanogenic activity higher than self-recovery. The specific methanogenic activity of recovery sludge with glucose, acetic acid, cellulose, and gelatin was 0.234, 0.684, 0.528, and 0.134 gCH₄-COD/gVSS/d, respectively. The methane yield of dilution with BE strategy was 34.9% higher than the self-recovery and showed better growth of methanogens under suitable pH (6.5–7.3). Therefore, the dilution of the inhibited sludge with BE at a ratio of 8:2 was a suitable strategy to accelerate the recovery process of the sludge from the inhibited state.

The recovery by adjusting pH strategy with all concentrations tested of (0.10–0.5% w/v) Ca(OH)₂, (0.10–0.14% w/v) NaOH, (6.00–10.00% w/v) oil palm ash, except (0.85–1.25% w/v NaHCO₃) exhibited higher methane yield and methane production rate than the self-recovery (Table 2). The highest methane yield (383 mL-CH₄/g-VS_added) was achieved from the recovery by adjusting pH with 0.14% w/v NaOH with the methane production rate of 111.7 mL-CH₄/d. On the other hand, the highest methane production rate (226.3 mL-CH₄/d) was achieved from the recovery by adjusting pH with 8.0% w/v oil palm ash with the hydrolysis constant (k_H) of 0.007 d⁻¹ and lag phase of 8.2 days with the recovery time of 32.4 days. In terms of the specific SMA of recovery sludge by three alkaline sources for adjusting pH, Ca(OH)₂ showed the highest SMA on acetic acid and cellulose, as substrates were 0.767 and 0.821 gCH₄-COD/gVSS/d, respectively. In contrast, the highest SMA on glucose and gelatin (0.591 and 0.096 gCH₄-COD/gVSS/d, respectively) were obtained from recovery by adjusting pH with 8.00% w/v oil palm ash. Thus, the pH adjustment strategy using 0.14% w/v NaOH and 0.40% w/v Ca(OH)₂ gave the highest methane yields of 383 and 373 mL-CH₄/g.VS_added, respectively, while using 8.00% w/v oil palm ash and 0.40% w/v Ca(OH)₂ gave the highest methane production rates of 226.3 and 151.4 mL-CH₄/d, respectively. The 0.40% w/v Ca(OH)₂ was demonstrated effective in accelerating recovery of the inhibited sludge by saving 31.6% recovery time (from 49.4 days to 33.8 days) and enhancing the methane yield of 78.5% (from 209 to 373 mL-CH₄/g.VS_added) when compared with self-recovery. Nevertheless, the pH adjustment strategy from the best results of each source of alkaline (0.14% w/v NaOH, 0.40% w/v Ca(OH)₂, 1.25% w/v NaHCO₃, and 8.00% w/v oil palm ash) was selected for economic evaluation compared to the dilution strategy (dilution with BE at 8:2 ratio).

The recovery by bioaugmentation or re-inoculation of active methane-producing sludge at 5–50% into the inhibited sludge was able to accelerate the recovery process in terms of methane yield (212-237 mL-CH₄/g-VS_added), methane production rate (40.4–83.5 mL-CH₄/d), lag phase (15.2–11.0 days) but not reduced the recovery time (43.6–47.2 days) (Table 3). Its k_H value was 0.006–0.008 d⁻¹. The addition of active methane-producing sludge at 15–50% had a methane production rate higher than the self-recovery because it can improve the amount of active biomass. The recovery of this strategy gave a small improvement with a long recovery time of 45.8 days and not difference with self- recovery (49.20 days). In particular, k_H of the addition of active methane-producing sludge strategy was small increased when compared with self-recovery. The amount of active methane-producing sludge at 30%–50% was suitable for recovery of the inhibited sludge with a shorter lag phase and increased tolerance of microorganism to low pH and high VFAs. However, the addition of active methane-producing sludge strategy had lower recovery efficiency than alkali addition due to low buffering capacity.

Economic evaluation

The energy and economic evaluation of each recovery strategy were provided in this study. The price of chemical addition, energy consumption, biogas loss, and human resources was achieved according to the current market price and the average price of industrial water in Thailand. According to Table 4, there were no extra-economic benefits from either of the recovery strategies. The recovery by NaOH addition, dilution with biogas effluent, and oil palm ash addition could reduce the recovery cost of the inhibited AD systems more than other strategies. The net profit of strategies for recovery inhibited sludge is in the following order (USD/m³/d): NaOH addition (−1.76), dilution with biogas effluent (−1.77), oil palm ash addition (−1.79) Ca(OH)₂ addition (−2.49), dilution with tap water (−3.33), and NaHCO₃ addition (−7.2). The recovery by 0.14% w/v NaOH addition corresponding to the addition of NaOH of 1.40 kg/m³-inhibited sludge had the chemical addition cost of 0.49 USD/m³/d. The recovery by dilution with biogas effluent at a ratio of 8:2 corresponding to 0.2 m³/m³-inhibited had no cost for biogas effluent. The recovery by 8.00% w/v of oil palm ash addition corresponding oil palm ash of 80 kg/m³-inhibited sludge had no oil palm ash cost. Oil palm ash is a by-product obtained by burning fibers, shells, and empty fruit bunches as fuel in palm oil mill boilers, while biogas effluent is that from AD digester free of charge. Thus, we suggest that NaOH addition, dilution with biogas effluent, and oil palm ash addition could be economically feasible strategies to recover the inhibited AD system feeding with POME.

The validity of lab-scale results in the full-scale recovery of AD process imbalance

Full-scale recovery was conducted by diluting 50 m³ of inhibited sludge with biogas effluent at a ratio of 8:2, which was added to 6,000 m³ biogas reactors every day. The AD imbalance reactor was maintained by a low feeding rate of 10 m³-POME/d. After dilution, the pH increased from 5.6 to 6.8 in the first week and 7.8 in the second week. After that, the reactor was operated normally with a feeding rate of 200 m³-POME/d. The methane production rates of the first, second, and third weeks of dilution with biogas effluent were 0.8, 2.0, and 2.86 m³-CH₄/m³-reactor /d with the pH values of 6.8, 7.8, and 7.8, respectively (Fig. 2). The SMA values of the first, second, and third weeks of dilution with biogas effluent were also increased to 0.44, 0.70, and 0.71 g-CH₄-COD/g-VSS/d, respectively. Results indicated that the reactor recovered within two weeks after dilution with biogas effluent. Therefore, this recovery time (15 days) gave a better result than lab-scale reactor (36.4 days).

Microbial community responsible for high potential recovery strategies

The microbial community from the four high potential recovery strategies (dilution with BE at a ratio of 8:2, 0.14% w/v NaOH addition, 0.50% w/v Ca(OH)₂ addition, and 8.00% w/v oil palm ash addition) indicated by short time lag phase, short recovery time, high SMA activity, and high methane production was analyzed. The heat map of the bacterial and archaeal communities of these four recovered sludge were higher than those of self-recovery (Fig. 3). The bacterial community of self-recovery strategy was dominated by Desulfotomaculum sp., Bacteroidetes sp., and Lactobacillus sp. (Fig. 3A). The number of Clostridium sp., Anaerostipes sp., Lynsinibacillus sp. increased after 12 days of self-recovery. For the archaea community, self-recovery was dominated by Methanosaeta sp. (Fig. 3B). The numbers of Methanosaeta sp. and Methanosarcina sp. increased after 12 days of self-recovery, while the bacteria in recovery by dilution with BE, at a ratio of 8:2, was dominated by Desulfotomaculum sp., Lactobacillus sp., Bacteroidetes sp., Clostridium sp., Staphylococcus sp., Selenomonas sp., and Lynsinibacillus sp., in which the last four species increased after 5 days of recovery. The archaea community was dominated by Methanosaeta sp. and Methanosarcina sp. in which the latter species decreased after 5 days of recovery. The recovery by 0.14% w/v NaOH addition was dominated by Desulfotomaculum sp., Blautia sp., Lactobacillus sp., Bacteroidetes sp., and Selenomonas sp., with the number of Blautia sp., Clostridium sp., Anaerostipes sp., Lynsinibacillus sp., and Staphylococcus sp. increasing after 5 days of recovery. Furthermore, the archaeal community was dominated by Methanosaeta sp., Methanosarcina sp., and Methanococcoides sp. after 5 days of recovery. The recovery by 0.50% w/v Ca(OH)₂ addition was dominated by Desulfotomaculum sp., Blautia sp., Clostridium sp., Anaerostipes sp., Kurthia sp., Exiguobacterium sp., Staphylococcus sp., Selenomonas sp., Bacillus sp., Lynsinibacillus sp., Lactobacillus sp., and Bacteroidetes sp., in which the last two species did not appear at 5 days of recovery. The archaea community was dominated by Methanosaeta sp., with Methanosarcina sp., and Methanococcoides sp. at 5 days of recovery. The recovery by 8.0% w/v oil palm ash addition was dominated by Desulfotomaculum sp., Blautia sp., Lactobacillus sp., Bacteroidetes sp., Clostridium sp., Anaerostipes sp., Kurthia sp., Exiguobacterium sp., Staphylococcus sp., Selenomonas sp., Bacillus sp., and Lynsinibacillus sp., where the first four species did not appear at 5 days of recovery, while the dominated archaea community is similar with the recovery by 0.50% w/v Ca(OH)₂ .The orders Clostridiales and Bacilli were observed as main bacteria in all recovered sludge. Desulfotomaculum sp. was found in all recovered strategies while Blautia sp., Clostridium sp., and Anaerostipes sp. were remarkably abundant during 12–20 days in pH adjustment with 0.50% w/v of Ca(OH)₂ and 8.0% w/v of oil palm ash addition. During the third week, Bacillus sp., Lactobacillus sp., Staphylococcus sp., Kurthia sp., Lysinibacillus sp., Exiguobacterium sp., and Bacillus sp. were observed in all strategies. The recovery by 0.50% w/v Ca(OH₂) and 8.00% w/v oil palm ash addition was predominant with bacteria belonging to Kurthia sp., Exiguobacterium sp., and Bacillus sp. These bacteria can produce VFA from monomers after hydrolysis, especially Exiguobacterium sp., which was a high number in the third week. The member of Bacteroides sp., and Selenomonas sp. had a low number in self-recovery but abundant in the recovered sludge with 0.14% of NaOH, 0.50% of Ca(OH)₂, and 8.00% of oil palm ash addition. The distribution of the exclusive bacteria in different groups clearly showed that the recovery strategy significantly influences the bacterial community structure and selectively enriches specific acidogenic bacteria during the recovery process. The acetoclastic methanogen (Methanosaeta sp.) was more abundant than hydrogenotrophic methanogens in the recovered sludge. Methanosaeta sp. and Methanosarcina sp. were predominant in the recovered sludge with 0.50%w/v Ca(OH)₂ and 8.00% w/v oil palm ash addition. Molecular identification of bacteria and archaea from recovery strategy by dilution with BE 8:2, addition with 0.14% w/v NaOH, 0.50% w/v Ca(OH)₂, 8.00% w/v oil palm ash and self-recovery by denaturing gradient gel will be offered in Tables S1 and S2, respectively).