Biogas Enhancement Through Anaerobic Co-digestion Of Oil Refinery Wastewater With Three Different Organic Waste Materials

Oil refinery wastewater(ORWW)is one of the considerable parts of industrial wastewaters and aquatic substrates.That contains different petroleum hydrocarbons,water,heavy metals,salts and solid particles.Although different effective methodologies had been investigated to treat ORWW,improving its anaerobic digestion and biodegradability as a reliable and natural treatment had been focused during recent years.Anaerobic co-digestion(AcoD)of different substrates to improve their AD process is well-known technique.This was applied to improve the AD process of ORWW with three various organic substrates including food waste(FW),chicken manure(CM),and sugarcane bagasse(SCB).Besides,their co-fermentations can be improved the AD process of selected organic substrates.To study the AcoD of ORWW with FW,CM,and SCB;six different general experiments had been conducted.And their fermentation process parameters in terms of biogas production(BGP),bio-methane production(BMP),removal soluble chemical oxygen demand(CODs)efficiency and retention time were evaluated.Briefly,these are summarized below:1.Six different ORWW:FW mixing ratios were prepared and fed to the anaerobic batch digesters.To model the cumulative BMP kinetic of AcoD process,three different mathematical models including the modified Gompertz model,Transfer function model and Logistic function model were applied and their reliabilities were evaluated.The results oriented that the 1:1,2:1 and 4:1 mixing ratios could support microbial growth and produce remarkable BGP and BMP content.The highest BGP and BMP values were(179.21±28.45)mL/g.VS and(104.66±20.34)mL/g.VS,respectively,which were produced by the digester with 4:1 AcoD mixture during 25 days,as the shortest retention time.Also,the highest CODs removal efficiency was observed from the same AcoD treatment,which was(75.00±0.66)%.Semi-empirical modeling revealed that the modified Gompertz model had good fitting to the experimental data for all AcoD mixtures.While the Transfer function model indicated a better fit for 4:1 AcoD mixture.2.To improve the AcoD of ORWW with an appropriate nutrients pool for microbial and buffer capacity growth,its AcoD with chicken manure was studied.Their AcoD batch experiments of six ORWW:CM ratio treatments(5:0,4:1,3:2,2:3,1:4 and 0:5)under mesophilic condition,were investigated.Then,the modified Gompertz model was well fitted to the experimental cumulative bio-methane production data.The highest soluble chemical oxygen demand removal rate was obtained for 4:1 ratio treatment.The highest biogas production and bio-methane content were achieved for 1:4 ratio treatment.Moreover,the shortest retention time was obtained for 3:2 ratio treatment.By considering the highest oil refinery wastewater portion in the mixtures and the statistical test(LSD0.05)results for the kinetic parameters,it can be concluded that the 4:1 ratio treatment was the optimum treatment.3.To study the optimum process conditions for pre-treatments and AcoD of ORWW with chicken manure,the L9(34)Taguchi's orthogonal array was applied.The biogas production(BGP),bio-methane content(BMP),and chemical oxygen demand stabilization(CODs)rate were evaluated as the process outputs.The optimum conditions were obtained by using Design Expert software(Version 7.0.0).The results indicated the optimum conditions could be achieved with 44%ORWW,36? temperature,30 min sonication,and 6%TS in the digester.The optimum BGP,BMP,and CODs removal rate using the optimum conditions were 294.76 mL/gVS,151.95 mL/gVS,and 70.22%respectively as concluded by the experimental results.In addition,the artificial neural network(ANN)technique was implemented to develop an ANN model to predict BGP yield and BMP content.Finally,the architecture of 9-19-2 for the ANN model was obtained that include nine input,nineteen hiden and two ouput neurons.4.Six different compositions including four different ORWW:SCB mixing ratios and controls were conducted.The negligible BGP by ORWW mono-digestion revealed that it could not support any microbial activity.And,increasing the SCB ratio in the AcoD compositions led to increased BGP and BMP contents.The experimental results using a 1:4 AcoD composition resulted in superior BGP volume with a higher BMP content and lower SCB usage during a shorter retention time.Moreover,the CODs removal rate and BMP content were improved through applying AcoD compositions.The results computed by applying three mathematical models determined that the modified Gompertz model provided the best fitness.Also,implementing three different artificial neural network algorithms including BPNN,RBFNN,and GRNN to model the BGP data revealed that the BPNN(Back Propagation)algorithm was the most fit for the experimental BGP data,with 0.64 and 0.97 for MSE and R2,respectively.5.To investigate the optimum AcoD process of ORWW with SCB under different thermochemical pretreatment,sixteen significant experiments based on the Taguchi's L16(45)orthogonal array were designed and conducted.The lingo-cellulosic analysis confirmed that the treatment of 10%H2O2 with 60 min autoclaving time performed the highest lignin removal rate.Multi-response optimization results revealed that the optimum conditions including 30%ORWW portion in digester,5 min irradiation time to treat ORWW,5%H2O2 concentration to pretreat ORWW,30 min autoclaving time to treat SCB and 0%H2O2 concentration to treat SCB may perform the optimum AcoD process performance.Confirming experiments to prove the optimum conditions performed 481.53 mL/gVS,280.79 mL/gVS,and 49.65%as the BGP and BMP yields and CODs removal rate.Furthermore,implementing ANN technique introduced a proper model with architecture of 5-5-2 to predict BGP yield and its bio-methane content.That include five input,five hiden and two ouput neurons.6.Based on the results from AcoD batch experiments,the mixture of ORWW-FW had been selected as an appropriate treating mixture.And according to the Taguchi's L9(34)orthogonal array,nine different semi-continuous experiments under different mixing ratios,hydraulic retention time(HRT),and sonication time were conducted.The multi-responses optimization results showed that the 684.66 mL/gVS,352.79 mL/gVS,and 3.38%were observed as the optimum values of BGP yield,BMP yield,and volatile solids(VS)removal rate.The optimum process responses were performed by the process conditions of 80%ORWW content,12 days HRT,and 15 min sonication time to pre-treat ORWW and FW.Moreover,a proper ANN model to predict BGP yield and its bio-methane content was developed.And the architecture of 11-15-2 was designed.That include eleven input,fifteen hiden and two ouput neurons.The values of 5843.90 and 0.97 were obtained as the optimum validation performance and correlation co-efficient(R)of all data sets,respectively.As a summarized conclusion of this study,it may be stated that the ORWW was investigated as a proper co-substrate to do co-fermentation with studied organic substrates.And their positive synergetic effects such as providing nutrient pool to improve microbial activities and diluting inhibitants in the digesters were confirmed.