An-Najah National University Faculty of Graduate Studies Feasibility of Generating Renewable Energy from Wastewater Treatment Process Using Microbial Fuel Cells: The West Bank as Case Study By Odai Judeh Adel Judeh Supervisor Prof. Dr. Marwan Haddad This Thesis is Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Water and Environmental Engineering, Faculty of Graduate Studies, An-Najah National University, Nablus, Palestine 2017 II Feasibility of Generating Renewable Energy from Wastewater Treatment Process Using Microbial Fuel Cells: The West Bank as Case Study By Odai Judeh Adel Judeh This Thesis was defended successfully on 15/05/2017 and approved by: ureSignat Defense Committee Member  Prof. Dr. Marwan Haddad / Supervisor ………………………  Dr. Nidal Mahmoud / External Examiner ………………………  Dr. Imad Breik / Internal Examiner .…..………….……… III Dedication To my mother, father, brothers and sister IV Acknowledgements First of all, praise to Allah for helping me in making this research possible. I would like to extend thanks to many people who helped me during my research work. Special mention goes to my enthusiastic supervisor Prof. Dr. Marwan Haddad, not only for his tremendous academic support, but also for giving me many wonderful opportunities during my work in Water and Environmental Studies Institute (WESI). Special thanks to Dr. Abdelfattah Hasan for his infinite support during performing my experiments at Environmental Engineering Lab in AN-Najah National University. Many thanks go to my defense committee for their efforts in academic reviewing of my thesis. Special thanks go to Water and Environmental Studies Institute (WESI) staff for helping and encouraging me. Special thanks also are presented to Water and Environmental Studies Institute (WESI) for giving me the chance to work on REWAS project (one of Palestinian-Dutch Academic Cooperation Program on Water Projects). Profound gratitude goes to people who helped me in this thesis: Prof. Dr. Hamdallah Be’erat, Dr. Abdelrahim Abusafa, Dr. Subhi Samhan, Eng. Omar Tuffaha, Eng. Ahmad Khaldi, Mr. Zahran Ashqar, and Eng. Hamees Tubeileh. Special mention goes to my parents, brothers, sister, friends and colleagues. V اإلقرار أنا الموقع أدناه مقدم الرسالة التي تحمل عنوان: Feasibility of Generating Renewable Energy from Wastewater Treatment Process Using Microbial Fuel Cells: The West Bank as Case Study إلشارة اليه حيثما أقر بأن ما اشتملت عليه هذه الرسالة إنما هي نتاج جهدي الخاص، باستثناء ما تم ا و بحثي لدى ورد ، وأن هذه الرسالة ككل، أو أي جزء منها لم يقدم لنيل أي درجة أو لقب علمي أ أي مؤسسة تعليمية أو بحثية أخرى. Declaration The work provided in this thesis, unless otherwise referenced, is the researcher's own work, and has not been submitted elsewhere for any other degree or qualification Student's Name: سم الطالب:ا Signature: التوقيع : Date: : التاريخ VI Table of Contents Dedication ................................................................................................... III Acknowledgements ..................................................................................... IV Declaration ................................................................................................... V Table of Contents ........................................................................................ VI List of Abbreviations and Nomenclature ................................................. VIII List of Tables ............................................................................................... IX List of Figures ............................................................................................. XI Abstract .................................................................................................... XIII 1. INTRODUCTION .................................................................................. 1 1.1. Wastewater Treatment and Energy Issues ........................................ 1 1.2. Importance of the Research in Palestine .......................................... 4 1.3. Objectives and Research Questions ................................................. 4 2. LITERARTURE REVIEW ..................................................................... 6 2.1. Background ....................................................................................... 6 2.2. History and Development of MFCs ................................................. 9 2.3. Review of the Latest Research on MFCs ....................................... 11 2.4. Summary ......................................................................................... 22 3. STUDY AREA AND METHODOLOGY ........................................... 23 3.1. Study Area ...................................................................................... 23 3.2. Methodology ................................................................................... 26 3.2.1. Experimental Setup and Design of Experiments ..................... 26 3.2.1.1. Preparation of Chambers ....................................................... 27 3.2.1.2. Electrodes Preparation .......................................................... 28 3.2.1.3. Preparation of Salt Bridges ................................................... 29 3.2.1.4. Temperature Control System ................................................ 31 3.2.1.5. Mixing System ...................................................................... 31 3.2.1.6. Wastewater and Sludge Sampling ........................................ 32 3.2.1.7. Cathodic Chamber Preparation ............................................. 33 3.2.1.8. Electrical Panel Preparation .................................................. 34 3.2.2. Experimental Program ............................................................. 34 3.2.2.1. Investigation of Effect of Electrode Material ....................... 34 VII 3.2.2.2. Investigation of Effect of Salt bridges Characteristics ......... 34 3.2.2.3. Operation of the Second Stage Experiment .......................... 35 3.2.3. Analytical Procedures and Measurements: .............................. 36 3.2.3.1. Environmental Measurements .............................................. 36 3.2.3.2. Voltage and Power Measurement ......................................... 37 3.2.4. Kinetic Models for COD Decay in the MFCs ...................... 37 3.2.5. Data Management and Analysis: ............................................. 39 4. RESULTS AND DISCUSSION ........................................................... 40 4.1. Effect of Different Conditions on MFC Performance .................... 40 4.1.1. Effect of Electrode Material on Output Voltage ...................... 40 4.1.2. Effect of Salt Bridge Diameter on Output Voltage .................. 43 4.1.3. Effect of Salt Bridge Solution on Output Voltage ................... 46 4.1.4. Effect of Salt Concentration on Output Voltage ...................... 48 4.2. COD Determination and Determining Kinetic Models for COD Removal ................................................................................................... 62 4.3. Relationship Between Output Voltage and Substrate COD ........... 62 4.4. Achieved Output Power and Feasibility Estimation ........................ 64 5. CONCLUSIONS AND RECOMMENDATIONS ............................... 68 References ................................................................................................... 69 APPENDICES ............................................................................................. 84 VIII List of Abbreviations and Nomenclature Abbreviation Meaning BOD Bio-Chemical Oxygen Demand COD Chemical Oxygen Demand MAC-MFC Multiple Anode Chamber MFC MDCs Microbial Desalination Cells MFCs Microbial Fuel Cells MPL Micro Porous Layer Mt Million tonnes Mtoe million tonnes of oil equivalent OsMFCs Osmotic Microbial Fuel Cells PCR power to cost ratio PEM Proton Exchange Membrane pH Potential Hydrogen (Acidity or Alkalinity Scale) POME palm oil mill effluent PVDF- AC Activate Carbon with Polyvinylidene Fluoride as diffusion layer TSS Total Suspended Solids WW Wastewater WW-MFC MFC operated by wastewater as substrate WWTP Wastewater Treatment Plant IX List of Tables Table 2.1 Achieved power output in various substrates used in MFCs ..... 14 Table 2.2Recorded power densities for various electrode materials.......... 18 Table 3.1 Details of substrate used in the MFCs ........................................ 27 Table 4.1 Summary of sludge COD measurement ..................................... 53 Table 4.2 Summary of wastewater COD measurement ............................. 53 Table A1 Achieved output voltage for different electrodes used in MFCs 85 Table A2 Achieved output voltage for different salt bridge diameters used in MFCs ..................................................................................... 85 Table A3 Achieved output voltage for different salts used in salt bridges for MFCs ......................................................................................... 86 Table A4 Achieved output voltage for different salt concentrations used in salt bridges for MFCs ................................................................ 86 Table B1 Shapiro-Wilk test results on SPSS to check normality of the data obtained from electrode materials experiment .......................... 97 Table B2 Levene's test results from SPSS to check homogeneity of variances of electrode material experiment............................................... 98 Table B3 Shapiro-Wilk test results on SPSS to check normality of the data obtained from salt bridge diameter experiment ........................ 98 Table B4 Levene's test results from SPSS to check homogeneity of variances of salt bridge diameter experimentError! Bookmark not defined. Table B5 Shapiro-Wilk test results on SPSS to check normality of the data obtained from salt solution type used in salt bridges experiment ................................................................................. 99 Table B6 Levene's test results from SPSS to check homogeneity of variances of salt solution type used in salt bridges experiment .............. 100 Table B7 Shapiro-Wilk test results on SPSS to check normality of the data obtained from salt solution concentration used in salt bridges experiment ............................................................................... 100 Table B8 Levene's test results from SPSS to check homogeneity of variances of electrode material experiment............................................. 101 Table B9 One way ANOVA test results for data obtained from electrode materials experiment using SPSS ........................................... 101 Table B10 Tukey HSD test for the data obtained from electrode materials experiment using SPSS ........................................................... 102 X Table B11 One way ANOVA test results for data obtained from salt bridge diameter experiment using SPSS ............................................ 105 Table B12 Tukey HSD test for the data obtained from salt bridge diameter experiment using SPSS ........................................................... 106 Table B13 One way ANOVA test results for data obtained from salt solution type used in salt bridge experiment using SPSS ..................... 107 Table B14 One way ANOVA test results for data obtained from salt solution concentration used in salt bridges experiment using SPSS .... 108 Table B15 Tukey HSD test for the data obtained from salt solution concentration used in salt bridges experiment using SPSS .... 109 Table C1 Details of COD calculations for all measurements .................. 111 Table D1 COD and output voltage measurements for MFC 1, 2 and 3 ... 117 Table D2 COD and output voltage measurements for MFC 4, 5 and 6 ... 118 Table D3 COD and output voltage measurements for MFC 7, 8 and 9 ... 119 Table D4 COD and output voltage measurements for MFC 10, 11 and 12 ............................................................................................. 120 Table E1 Kt values for COD reduction in MFC1 to determine the kinetic model ....................................................................................... 121 Table E2 Kt values for COD reduction in MFC2 to determine the kinetic model ....................................................................................... 122 Table E3 Kt values for COD reduction in MFC3 to determine the kinetic model ....................................................................................... 123 Table E4 Kt values for COD reduction in MFC4 to determine the kinetic model ....................................................................................... 124 XI List of Figures Figure 1.1 World total final energy consumption (Mtoe) through 1971 to 2013 ............................................................................................. 2 Figure 1.2 CO2 emissions from energy consumption (Mt of CO2) through 1971 to 2013 ................................................................................ 3 Figure 2.1 Typical microbial fuell cell ......................................................... 8 Figure 2.2 Schematic design of single chambered MFC ............................. 8 Figure 2.3 Schematic design of multiple anode chamber MFC developed by Mathuriya ............................................................................. 15 Figure 3.1 Maps and main districts of Gaza Strip and West Bank ............ 23 Figure 3.2 Wastewater sampling from primary sedimentation tank effluent ...................................................................................... 25 Figure 3.3 Schematic of DS-MFC used ..................................................... 28 Figure 3.4 Prepared glass U-shaped tubes for use in salt bridges .............. 30 Figure 3.5 Prepared salt bridges soaked in KCl solution ........................... 31 Figure 3.6 Used mixing system .................................................................. 32 Figure 3.7 Output voltage measurement .................................................... 38 Figure 4.1 Output voltage in MFCs of different electrode materials ......... 41 Figure 4.2 Output voltage of MFCs with different salt bridge diameters .. 44 Figure 4.3 Output voltage for KCl and NaCl Salt bridges ......................... 47 Figure 4.4 Output voltage for KCl and NaCl Salt bridges ......................... 49 Figure 4.5 Comparison between Kt values of third order reaction and linearized Kt values for COD behaviour in MFCs 1,2 and 3 ... 55 Figure 4.6 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 .... 55 Figure 4.7 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 .... 56 Figure 4.8 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 .... 56 Figure 4.9 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 .... 57 Figure 4.10 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 .... 57 Figure 4.11 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 .... 58 XII Figure 4.12 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 .... 58 Figure 4.13 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 .... 59 Figure 4.14 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 .... 59 Figure 4.15 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 .... 60 Figure 4.16 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 .... 60 Figure 4.17 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 ............................................................................................... 61 Figure 4.18 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 ............................................................................................... 61 Figure 4.19 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 ............................................................................................... 62 Figure 4.20 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 ............................................................................................... 62 Figure 4.21 Relationship between COD and output voltage ..................... 63 Figure B1 Steam and leaf plots for electrode materials experiment data to check for outliers using SPSS during the seven days of operation .................................................................................... 89 Figure B2 Steam and leaf plots for salt bridge diameter experiment data to check for outliers using SPSS during the seven days of operation .................................................................................... 92 Figure B3 Steam and leaf plots for salt bridge type experiment data to check for outliers using SPSS during the seven days of operation .................................................................................... 94 Figure B4 Steam and leaf plots for salt concentration used in salt bridges experiment data to check for outliers using SPSS during the seven days of operation ............................................................. 96 XIII Feasibility of Generating Renewable Energy from Wastewater Treatment Process Using Microbial Fuel Cells: The West Bank as Case Study By Odai Judeh Adel Judeh Supervisor Prof. Dr. Marwan Haddad Abstract Microbial fuel cells (MFCs) technology, is an innovative and relatively new technology by which organic matter can be simultaneously biodegraded anaerobically and generate electrical energy directly. MFCs can be used in various applications such as: water desalination, biosensors as well as water and wastewater treatment. Through this study, MFCs as a wastewater treatment system was investigated for the first time in Palestine. Palestine is a developing country that suffers from improper wastewater collection and treatment systems in addition to water supply shortage that can cause increasing organics concentrations in the discharged wastewater. MFC model used in this research was double chambered-MFC (DS-MFC), it was operated by primary effluent wastewater as substrate, salt bridge was used as proton exchange media and water saturated of dissolved oxygen was used as cathodic solution. This research consists of two main parts: First, investigation of many parameters that may affect MFCs efficiency, such as: electrode material type, electrode size, salt bridge diameter, type of salt solution that used in salt bridge and concentration of salt solution used in the salt bridge. This part was conducted by constructing and operating different MFCs to investigate XIV each parameter (variable) individually. All conditions, except the concerned parameter, were fixed for each parameter experiment. Three duplicates- MFCs for each variable value were used to obtain reliable results. Output open circuit voltage (OCV) was measured one time per day and for one week for each variable and then the obtained output voltage data were analyzed as a trial to find the most suitable conditions. The second part is aimed to understand and model the relationship between COD of substrate in MFC at any time and output voltage from the MFC at the same time in addition to define the kinetic reaction order of COD removal process. Four different COD-MFCs were constructed, three duplicates for each. Initial COD value was approximately fixed for each 3 duplicates, and other parameters were set and fixed as found from the first part of this research. After 15 days-startup period the MFCs were operated for 30 days. COD was measured for the twelve MFCs each two days and output voltage was measured each 24 hours. Analysis of the obtained data from performed experiments, showed that MFCs with copper electrode produce output voltage significantly higher than MFCs with carbon brushes electrodes which, in turn, achieved output voltage significantly higher than both that achieved by MFCs with zinc electrodes and MFCs with manufactured carbon electrodes. It was found that diameter of salt bridge affects the output voltage of MFCs; MFCs with 10 mm salt bridge shown significantly higher output voltage than MFCs with both 16 and 24 mm salt bridges. This behavior could be interpreted by increasing the electrical resistance when the diameter is increased. XV It was found that KCl salt bridge is significantly more efficient than NaCl salt bridges when they were used in MFCs for wastewater treatment. It was found also that MFCs with 1M KCl salt bridges can produce output voltage significantly higher than that produced by MFCs with 3M KCl salt bridges. Results revealed that the COD of the substrate used in MFC at any time is related proportionally to output voltage from that MFC at the same time, this relationship can be fitted as natural logarithmic model as following: COD (mg/L) = 229.85 Ln (V)-1039.6; where V is output voltage (mV), and this model can be used with large number of limitations to indicate COD for a wastewater sample by measuring output voltage of MFC operated by that sample. Maximum COD removal percentage achieved in this study was 87.1 % which is comparable to published achievements. It was found that COD removal behavior in this study was ranged between 1st and 2nd order kinetic reactions. A maximum output power achieved was 0.585 W/m3 with an average output power of 0.251 W W/m3. 1 Introduction 1.1. Wastewater Treatment and Energy Issues Wastewater treatment has its positive impacts on sustainable water management and environment protection; however, it is considered an energy-intensive process (Foley, 2010; Gikas and Tsoutsos, 2015). Conventional wastewater treatment requires circa 1.1-2.4 MJ/m3 of electrical energy (Tchobanoglous et al., 2003). Energy needs could be higher significantly for small wastewater treatment plants i.e. for wastewater treatment plant with flow less than 5000 m3/day (Gikas and Tsoutsos, 2015). As more investments are needed to increase the wastewater treatment capacity and efficiency, it becomes a growing challenge to meet the associated energy demands. In future, energy demand for wastewater treatment is expected to grow due to increasing restricts on discharge requirements and resulted in increasing required treatments and energy needs. Inevitable population growth and consequently increasing of wastewater generation could increase energy needs for wastewater treatment in future. Studies reveals that new pollutants such as endocrine disrupting materials, pharmaceuticals and personal care products are contributing to increase energy need for wastewater treatment; because removal of such pollutants can’t be performed efficiently using conventional wastewater treatment processes; so other energy-intensive and special treatment processes are 2 required (Adam et al., 2002; Petrovicè et al., 2001; Ternes et al., 2002; Westerhoff et al., 2005; Zwiener and Frimmel, 2000). Current global energy consumption is about 9301 million tonnes of oil equivalent (1 ton=11,630 kWh) “Mtoe”, 84.5 % of the global consumption is of non-renewable resources (International Energy Agency, 2015). Figure 1.1 shows world total energy consumption from 1971 to 2013. Figure 1.1 World total final energy consumption (Mtoe) through 1971 to 2013 Source: (International Energy Agency, 2015) Non-renewable sources of energy are dramatically decreasing worldwide; due to the heavily increasing consumption of energy worldwide. Moreover, increasing carbon emissions from fossils fuels resources (Figure 1.2), followed by change in carbon percentage in the atmosphere have a great input into global climate change; so extensive research was started globally to find renewable, cheap and sustainable sources of energy. 3 Figure 1.2 CO2 emissions from energy consumption (Mt of CO2) through 1971 to 2013 Source: (International Energy Agency, 2015) Renewable energy is the energy that created by naturally replenished sources such as: winds, sunlight, tides, etc. Other energy sources such as: biomass, anaerobic digestion and biofuel are widely considered renewable energy (You, 2016). Wastewater went to be considered as a resource for water, energy and plant nutrients more than as waste. Actually, wastewater contains energy folds of that is required for its treatment (Gude, 2015; McCarty et al., 2011; Shoener et al., 2014). Net-zero energy wastewater treatment or net-positive energy wastewater treatment is an evolving trend that could be achieved through: energy conservation in water and wastewater treatment, hydraulic energy recovery, heat recovery, anaerobic digestion, algae growth for biofuels, microbial fuel cells (MFCs), and microbial desalination cells (MDCs), etc. (Gude, 2015) Microbial Fuel Cells (MFCs) technology provide an innovative option in generating sustainable energy from waste water simultaneously with its 4 treatment; so wastewater treatment energy needs might be partially self- covered. Microbial Fuel Cells (MFCs) are devices used to generate electrical energy in a biochemical process; in which living microorganisms digest and oxidize organic matters and producing electrons and protons which could be exploited (Borah et al., 2013; Du et al., 2007; Fenget al., 2009; Hisham, et al., 2013; Huggins et al., 2013; Karmakar et al., 2010; Kim et al., 2007; Lee et al., 2012; Lovley, 2006; Logan et al., 2006). 1.2. Importance of the Research in Palestine To our knowledge, this is the first research on MFCs conducted in Palestine. Palestine is a developing country that suffers from Israeli occupation. Energy situation in Palestine is very critical; due to absence of conventional fossil fuels resources and the lack to import 100 % petroleum needs and 92% of electrical needs from Israel. Wastewater treatment sector in Palestine is a budding sector and needs enormous efforts and budgets to be developed to satisfy environmental standards and policies. In the developed countries energy needs for wastewater treatment is significant portion of the total energy needs. MFCs could be an innovative solution to treat wastewater and produce energy that could cover large portion of wastewater treatment energy needs, and so could be of high economic and technical value for Palestine. 1.3. Objectives and Research Questions The main objectives for this research are: 5 1) Design, construct and investigation the behavior of microbial fuel cells for primary effluent wastewater treatment. 2) Data acquisition from pilot experiments: COD removal and output voltage data will be collected. 3) Manage data obtained from pilot experiment: including statistical analysis and modeling to characterize MFC behavior. This research will try to answer following key questions: 1) Which are the most suitable operating conditions and settings for MFC (salt bridge characteristics and electrodes material) that would reflect optimal energy generation and treatment efficiency in wastewater treatment? 2) What is the relationship between organic content of wastewater and output energy from MFC? 6 2. Literarture Review 2.1. Background MFCs are devices that directly convert the chemical energy stored in the organic matters into electrical energy (Mahendra and Mahavarkar, 2013; Min et al., 2005). Typical MFC consists of two compartments: Anodic chamber and cathodic chamber, separated by proton exchange membrane (PEM) (Karmakar et al., 2010; Mahendra and Mahavarkar, 2013; Du et al., 2007). Substrate (organic-rich matter) is added to the anodic chamber, where oxidation of organic matter occurs by anaerobic microorganisms (Delaney et al., 1984; Mathuriya and Sharma, 2009). Electrons and protons are produced from oxidation half-reaction of the substrate, electrons transfer from the anode to the cathode through external electrical load (resistance); where protons transfer to the cathode through PEM or salt bridge (Lithgowet al., 1986; Mohan et al., 2007). Equation 1 shows the typical anode reaction using acetate as example reaction. CH3COO- + 2H2O (anaerobic microorganisms) → 2CO2 + 7H+ + 8e- (1) Through digestion of organic matters, electrons are transferred to the anode by microorganisms in three mechanisms: First, by exogenous mediators, which are added externally, such as Thionine. Second, Mediators naturally produced by microorganisms. Third, direct transfer from the cytochromes to 7 the anode (Min and Logan, 2004). One of the disadvantages of synthetic mediators is that they are unstable and toxic to most microorganisms. In the cathodic chamber; where reduction half-reaction occurs; catholite solution is added and it is working as electron acceptor. The most common electron acceptor (catholites) are ferricyanide (K3[Fe(CN)6]) or oxygen, which is the most suitable electron acceptor due to its high oxidation potential. At the cathodic chamber electrons and protons are combined with oxygen to form water as illustrated in Figure 2.1 (Lithgow et al., 1986; Rabaey et al., 2004). Equation 2 shows the typical cathodic reaction using acetate as example reaction. O2 + 4e- + 4H+ → 2H2O (2) MFCs that contains two chambers, are called double-chambered MFCs (DS- MFC), see Figure 2.2. Cathodic chamber can be removed, and the cathode is placed at the side of the anodic chamber and exposed to air at the other side, then it will be called single-chambered MFC (SC-MFC) or air-cathode MFC (Karmakar et al., 2010). Figure 2.1 shows the main parts and processes in a typical MFC, where Figure 2.2 show schematic drawing for single chambered microbial fuel cell. 8 Figure 2.1 Typical microbial fuell cell Source: (Oji et al., 2012) Figure 2.2 Schematic design of single chambered MFC MFCs have many advantages comparing to conventional anaerobic digestion processes: MFCs are applicable for low concentration substrates and they can be operated at ambient temperatures 20oC and below (Pham et al., 2006). 9 MFCs have many potential applications including: wastewater treatment, biofuel production, water desalination, remote power sources and biosensors. In wastewater treatment MFCs can remove organics and inorganics such as nitrate and sulfide. Non-conventional applications and functions of MFCs were navigated; Zhen He studied the merger of MFC and reverse osmotic pressure to create osmotic microbial fuel cells (OsMFCs) to produce high quality water simultaneously with generating electrical energy (He, 2012). 2.2. History and Development of MFCs The linkage between chemistry, biology and electricity isn’t an emerging discovery; scientists have long known about this relationship (Potter, 1911). In the late 1700s, Luigi Galvani, noted that a detached frog’s leg twitched due to the electrical charges in the atmosphere; this discovery leaded to find electro-chemistry Science (Rogers, 2010). In the beginnings of the 20th century, researches published on electricity production using microorganisms. The first attempt proved that biological processes could produce electrical energy was in 1911 by Potter; he published the earliest report on the ability of producing electrical energy from oxidizing organic matters by microorganisms and he had demonstrated it (Bullen et al., 2006; Shuklaet al., 2004). However, this attempt was neglected for about two decades until Cohen approved that a batch biological fuel cell could produce more than 35 volts (You, 2016; Cohen, 1931). 10 The USA NASA space program in the sixth decade of the previous century, was interested in biological fuel cell technology to produce energy in space ships from organic wastes (Putnam, 1971). A clear principle for biological fuel cells was identified in 1976 by Rao (Rao et al., 1976; You, 2016). In the 1980s, Bennetto succeeded to extract electrical current from MFCs that were operated using wastewater, pure cultures and artificial electron mediators; to facilitate electrons transfer to the anode (Bennetto et al., 1985; Bennetto et al., 1983). In the 1980’s, it was discovered that the generated electricity could be significantly increased if electron mediators were added (Du et al., 2007). The outer layers of the majority of microorganisms consists of non- conductive lipid membrane, peptididoglycans and lipopolysaccharides which hinder the direct electron transfer to the anode. Electron mediators such as: thionine, methyl blue, methyl viologen, humic acid, neutral red and others can accelerate the transfer (Davis and Higson, 2007). Only anodophiles (exoelectrogens) such as Shewanella putrefaciens, Geobacteraceae sulfurreducens, Geobacter metallireducens and Rhodoferax ferrireducens are bioelectrochemically active organisms and they can form a biofilm on the anode surface and also transfer electrons to the anode directly by conductance through their cell membrane (Du et al., 2007). Habermann and Pommer demonstrated that some microorganisms can produce electron shuttles naturally, so no exogenous mediators are necessary (Habermann and Pommer, 1991). 11 However, in the last two decades, too many scientists and researchers have worked on development of MFCs by attempting to increase efficiency of treatment and output power. 2.3. Review of the Latest Research on MFCs MFCs technology is an innovative solution for bioelectricity production from wastewater treatment simultaneously with its treatment; so, energy needs of wastewater treatment could be partially or totally recovered from the wastewater itself. Today many researchers are working on developing MFC’s and other electrochemical technologies. As a comparison between MFCs and conventional aeration treatment of wastewater, it was found that both of them able to remove more than 90% of COD; but with shorter time in conventional aeration (8 days) than MFCs (10 days) (Huggins et al., 2013). In the case of high COD concentration, MFCs showed lower removal efficiency compared to aeration, but much higher efficiency for low COD concentration. Suspended solids measurements showed that MFCs reduced sludge production by 52-82 % compared to aeration and MFCs also can save energy consumption in aeration since that MFC process is anaerobic (Huggins et al., 2013). MFCs can be operated by a wide range of substrates as long as they contain organic matters (You, 2016). Lee with his partners used sulfide as substrate, which is considered as one of the common inorganic pollutants in the wastewater from livestock farming. In Lee’s study, sulfide was added to a synthetic wastewater with a 12 concentration of sulfide ranged from 50 to150 ppm and autotrophic denitrifier “Pseudomonas” was used as pure culture. They used two types of electrodes; carbon cloth electrode and carbon felt electrode and found that carbon cloth electrode can generate more voltage than the carbon felt (Lee et al., 2012). The achieved voltage density, power density, current density, percentage of sulfide converted to thiosulfate and percentage of sulfide converted to elementary sulfur in Lee’s study were 65 mV/m2 of anode area, 29.3 mW/m2 of anode area, 108 mA/m2 of anode area, 40.7% and 57.6 % respectively (Lee et al., 2012). Hisham with his co-authors used three different substrates to operate MFCs using carbon paper as electrodes for 96 hours; they were activated sludge, palm oil mill effluent (POME) and leachate from food waste. The highest voltage among three substrates was obtained by leachate-MFC (0.455 V), followed by POME-MFC (0.444 V) and then activated sludge-MFC (0.396 V). Where the highest treatment efficiency in terms of COD was achieved by activated sludge-MFC (37.5%), then leachate-MFC (6.11%) and POME- MFC achieved zero-COD removal. In terms of nitrogen removal, Sludge- MFC achieved the highest efficiency (65.28%) followed by POME-MFC (48.12%) and leachate-MFC (25.15%) (Hisham et al., 2013). Guo with his partners, used anaerobic sludge as substrate with adding glucose as extra carbon source in the startup period, nitroaromatic antibiotic chloramphenicol solutions with different concentrations were added to the sludge in order to study its removal efficiency (Guo et al., 2016). 13 Four types of MFCs were used in Guo’s experiments: normal MFC, open- circuit MFC, no extra carbon source MFC and abiotic (anaerobic sludge was autoclaved before inoculation) MFC. Results showed that the maximum removal efficiency of nitroaromatic antibiotic chloramphenicol was achieved in normal MFC, followed by open circuit MFC, then abiotic MFC and the minimum removal efficiency was achieved in MFC with no extra carbon source; given that equal initial concentration was applied. Results also showed an inverse relationship between removal efficiency and initial concentration of nitroaromatic antibiotic chloramphenicol (Guo et al., 2016). Table 2.1 shows various types of substrates used in previous MFC studies and power generation for each type (Maria and Dharmendra, 2016). 14 Table 2.1 Achieved power output in various substrates used in MFCs Substrate MFC Type Power generation Reference Swine Single chamber 261 mW/m2 (Köroğlu et al., 2014; Min et al., 2005) Swine Double chamber 45 mW/m2 (Köroğlu et al., 2014; Min et al., 2005) Saline domestic sewage sludge Double chamber 41 W/m3 (Karthikeyan et al., 2016) Nitrogen containing organic compounds (pyridine and methyl orange) Single chamber 502.5 mW/m2 (Wang et al., 2015) Nitrogen containing organic compounds (pyridine and methyl orange) Two single chambers connected in series 401.6 mW/m2 (Wang et al., 2015) Sulfide containing synthetic WW Double chamber 29.3 mW/m2 (Lee et al., 2012) Paper recycling WW+ 100 mM PBS Single chamber 672 mW/m2 (Huang and Logan, 2008) Urine Double chamber 8 mA/m2 (Ieropoulos et al., 2012) Lemon peel waste Double chamber 371 mW/m2 (Miran et al., 2016) Domestic WW Air-cathode 422 mW/m2 (Köroğlu et al., 2014; Ahn and Logan, 2010) Leachate Air-cathode 344 mW/m2 (Köroğlu et al., 2014; Puig et al., 2011) Starch Air-cathode 239.4 mW/m2 (Köroğlu et al., 2014; Lu et al., 2009) Beer brewery Air-cathode 205 mW/m2 (Köroğlu et al., 2014; Feng et al., 2008) Beer brewery Single chamber 170 mW/m2 (Köroğlu et al., 2014; Feng et al., 2008) Textile WW Single chamber 812 mW/m2 (Mise and Saware, 2016) A Multiple anode chamber MFC (MAC-MFC) design (see Figure 2.3) was developed by Mathuriya and its performance was compared to a single cathode chamber MFC (SC-MFC). During 60 days of operation with different wastewaters, it was found that the MAC-MFC generated stable and higher power outputs in comparison to SC-MFC (Mathuriya, 2016). 15 Figure 2.3 Schematic design of multiple anode chamber MFC developed by Mathuriya MFCs can be operated by pure culture, but it is more suitable to be operated by mixed cultures especially when the used substrate is complex such as wastewater (Kim et al., 2007; Prasad et al., 2006). Researchers have done intensive research on the cultures that could operate MFCs. Pure cultures such as Clostridium acetobutylicum, Clostridium thermohydrosulfuricum, Shewanella putrefaciens, Geobacter sulfurreducens, Desulfobulbus propionicus, Rhodoferax ferrireducens and Saccharomyces cerevisae (yeast) are used in bioelectricity production (Mathuriya and Sharma, 2009; Mokhtarian et al., 2012; Mathuriya and Sharma, 2010). Borah with his co-workers, isolated Bacillus megaterium and other 24 types of microorganisms from soil, and they used each type to operate individual MFC. They found that Bacillus megaterium is the most efficient type among the 25 types (Borah et al., 2013). 16 Lee used autotrophic denitrifier, isolated from an expanded granular sludge, sp. C27, to operate an MFC. Pseudomonas sp. C27 can convert sulfide to elementary sulfur, the generated power density was 29.3 mW/m2 (Lee et al., 2012). Lee’s findings are considered as an introduction for treatment of sulfides containing wastewater simultaneously with harvesting electricity. Sheikh et al. used isolated facultative anaerobic bacteria as biocatalysts in MFCs. These MFCs were operated by various substrates. They achieved a maximum voltage of 0.30 V using urea as substrate (Sheikh et al., 2015). Enterobacter sp. ALL-3 culture was used by Tkach and his coworkers to operate SC-MFC at different temperatures (ranges from 5 oC to 25 oC), using acetate as substrate. The maximum achieved power densities at 5 oC, 10 oC and 25 oC were 293, 213 and 84 mw/m2 respectively (Tkach et al., 2016). Startup temperature in Tkach’s experiment was 5 oC for 30 days, a maximum voltage of 500 mV was achieved. After 8-10 days, the voltage was reduced to less than 50 mV and the MFC was resembled again. Then Temperature was increased to 10 oC and a maximum voltage achieved was 530 mV. Finally, temperature was increased to 25 oC and the maximum voltage achieved was 480 mV. It can be noticed that the time in which acetate was consumed at 10 oC range was less than that for 25 oC; this is because digestion of acetate was quicker. But still these results can be accredited for Enterobacter sp. ALL-3 culture and can’t be generalized for all cultures (Tkach et al., 2016). Genetically engineered bacterial strain was used to operate an MFC with urine substrate and was observed by Shreeram and his partners. An increase 17 of 2.7-fold in peak power density was achieved using genetically engineered strain comparing to the wild-type strain (Shreeram et al., 2016). Type of microorganisms doesn’t alone affect MFC efficiency; cells count and biofilm density also have effects on output power generated by MFC. Power density depends directly on biofilm growth; so, it is increased significantly during initial growth period of biofilm (Ramasamy et al., 2008). Previous research found that using ferricyanide solution as cathodic solution (electron acceptor) is more efficient, in terms of both treatment and power generation, than using oxygen. Mohan, with his partners, tested both cathodic solutions under the same conditions, they found that generated power is slightly lower in case of using dissolved oxygen. Where COD removal percentage is significantly higher in the case of using ferricyanide solution (Mohan et al., 2007). Studies demonstrated that nitrite can be used as cathodic solution (electron acceptor). Ammonium-containing effluent from the carbon-utilizing anode was fed to external biofilm-based aerobic reactor for nitrification, the nitrified liquor was then fed to the cathodic chamber in an MFC in order to reduce nitrate (Virdis et al., 2008). A maximum power of 34.6 ±1.1 W/m3, maximum current of 133.3 ±1.0 A/m3 and nitrite removal rate of 0.41 kg NO3—N/m3.day of net volume of the cathodic chamber were achieved (Virdis et al., 2008). COD removal rate was 2 kg COD/m3.day of the net volume of the cathodic chamber. Achieved 18 COD/N ratio in MFCs was 4.5 compared to COD/N ratio in conventional aeration processes of more than seven. It is important to mention that the electrodes, especially anode, are the key components to set the efficiency of MFC’s (Logan et al., 2006; Offei et al., 2016). Many types of electrodes are available and can be used; e.g. graphite, zinc, copper, aluminum, carbon, stainless steel, mild steel, etc. The electricity generation from MFCs is directly proportional to the surface area and type of the anode (Ashoka et al., 2012). Achieved power densities of previous studies on MFCs, using the most common electrodes, are summarized in Table 2.2. Table 1.2 Recorded power densities for various electrode materials Electrode Material Power Density Reference Carbon paper 148 mW/m2 (Thygesen et al., 2011) Graphite fiber brush 422 mW/m2 (Ahn and Logan, 2010) Granular activated carbon 2981 mW/m2 (Nam et al., 2010) MPL-Carbon viel 60.7 mW/m2 (You et al., 2014) MPL-Carbon cloth 50.6 mW/m2 (You et al., 2014) PVDF-AC 1400 mW/m2 (Yang et al., 2015) Carbon felt 356 mW/m3 (Aelterman et al., 2008) Graphite felt 386 mW/m3 (Aelterman et al., 2008) Graphite wool 321 mW/m3 (Aelterman et al., 2008) Graphite granules 257 mW/m3 (Aelterman et al., 2008) Activated carbon 1740 mW/m3 (Offei et al., 2016) Mediator-less DS-MFC, using carbon nanotube CNT-doped PEM, carbon zinc electrodes and E. coli culture was constructed by Vijay and others. Maximum output voltage and maximum current were 1.23 volts and 8.08 mA respectively (Vijay et al., 2014). 19 Electrodes used in MFCS should have satisfy various conditions to be efficient, such as: reasonable cost, good electrical conductivity, high surface area, low electrical resistance, non-corrosive, biocompatible, distance between electrodes should be as close as possible and they should be chemically and mechanically stable (Jang et al., 2004; Ashoka et al., 2012; Singh et al., 2010). The maximum current was achieved in the literature when the platinum-coated graphite was used as electrode (Ashoka et al., 2012). Another study focused on electrodes used in MFCs, MFCs were constructed and monitored using cowdung as substrate, nylon as PEM and methylene blue mediator. Five different metals were used as electrodes: copper (Cu), aluminum (Al), stainless steel (SS), zinc (Zn), carbon (C) and mild steel (MS). Twenty-one electrodes combinations were applied, with the same surface area of each electrode. The highest four combinations were Cu/Zn, SS/SS, C/C and Al/SS respectively (Ashoka et al., 2012). However, other questions should be raised; does copper stay stable without reactions with substrate? If there were any reactions, how do these reactions affect the microorganisms? Activated carbon (AC) are now used frequently as electrodes for MFCs, due to its low price and good catalytic activity (Wang et al., 2013; Offei et al., 2016). Output power from MFC follows saturation kinetics as a function of provided substrate (anodic solution). Generated voltage from MFC is decreased linearly with time (Barua and Deka, 2010). 20 MFC is anaerobic process that has many advantages such as: it produces less carbon emissions, requires less energy, produces less sludge through treatment, requires less nutrients and higher potential energy recovery because most the organics transferred to energy (Hwang et al., 2004; Du et al., 2007). Conditions that affect MFC efficiency are: pH, temperature, proton exchange material, electrode material, ratio of membrane surface to anodic chamber volume, ratio of anode surface area to the volume of anodic chamber and dissolved oxygen concentration in the cathodic chamber (Karmakar et al., 2010; Oji et al., 2012). Feng et al. tried to operate MFC under different temperatures, they used three temperatures: 30, 20 and 15 oC. They found that generated energy is significantly reduced when temperature reduced by 5 oC (from 20 to 15 oC), and is reduced less significantly when temperature reduced from 30 to 20 oC. For COD removal efficiency, they found that slight differences between the 3 temperatures (Feng et al., 2009). Protons exchange from the anode to cathode can be achieved through PEM mainly, the main disadvantage of PEM is the high cost (Du et al., 2007). To overcome cost problem of PEM, salt bridges can be used for protons exchange but with less efficiency (Sevda and Sreekrishnan, 2012). Nafion membrane is the most popular PEM due to its highly selective permeability of protons (Du et al., 2007). Park and Zeikus used porcelain septum made from kaolin as proton exchange media, a maximum power and 21 voltage achieved, using sewage sludge as substrate, were 788 mW/m2, 045 V (Park and Zeikus, 2003). Grzebyk and Poźniak have prepared interpolymer cation exchange membranes with polyethylene/poly (styrene-co-divinylbenzene) by their sulfonation with a solution of chlorosulfonic acid in 1,2-dichloroethane, and they used them in MFC as protons exchange media; but they achieved low power output (Grzebyk and Poźniak, 2005). Min et al. compared between salt bridges and PEM as proton exchange systems in MFCs, using both pure and mixed cultures, and they found that generated power from PEM system, 38 mW/m2, was order of magnitude higher than power generated from salt bridges, 2.2 mW/m2. According to their conclusion, salt bridge has internal electrical resistance that is extremely higher than that for membrane (Min et al., 2005). However, other researchers have worked on development of salt bridges and improvement of its efficiency. Sevda and sreekri have investigated the effect of salt concentration of salt bridges on electricity generation from synthetic WW-MFC. Maximum output power, was achieved at 5% concentration of NaCl, was 84.99 mW/m2 with 88.41 % COD removal (Sevda and Sreekrishnan, 2012). To our knowledge, there is no previous research on the effect of salt concentration in salt bridges on output power when natural wastewater is used as substrate. Also, no previous studies performed on the effect of salt concentration used in salt bridges for other salts such as potassium chloride. 22 COD removal characteristics in SC-MFC were studied by Zhang with his partners (Zhang et al., 2015). However, to our knowledge, COD removal characteristics for DS-MFC were not studied before and this gap will be covered by this study. However, to our knowledge, there are no previous research on other conditions that affect MFC efficiency. 2.4. Summary It can be notice from the performed literature review that much subjects regarding MFCs and its application in wastewater treatment were covered in previous research. Latest research is mainly covering: substrates used to operate MFCs, microorganisms’ cultures, electrode material, system temperature, cathodic solutions, proton exchange mechanism and COD removal characteristics for single chambered MFCs. In this research, many new topics is to be covered; i.e. wastewater mixed with anaerobic sludge will be used as substrate, salt bridge characteristics effects on MFC behavior and COD removal characteristics for double chambered MFCs. This research is good to be conducted in Palestine for the first time, and could be a starting point for further research for Palestine in future. MFCs could be an innovative solution for wastewater treatment and energy problems occur in Palestine. 23 3. Study Area and Methodology 3.1. Study Area Occupied Palestinian Territories consists of two disconnected parts; Gaza Strip and the West Bank, including East Jerusalem, with a total land area of 6,020 Km2 (See Figure 3.1). It has a total population of 4.55 million: 2.79 million in the West Bank 2.79 and 1.76 million in the Gaza Strip ( Palestinian Central Bureau of Statistics, 2015). Figure 3.1 Maps and main districts of Gaza Strip and West Bank (Source: www.globalsecurity.org/military/world/palestine/images/palestinemap.gif) http://www.globalsecurity.org/military/world/palestine/images/palestinemap.gif 24 The wastewater samples were collected from Nablus West wastewater treatment plant. This WWTP is located to the west of Nablus as shown in Figure 3.1. In the West Bank, the annual discharge of wastewater was estimated at about 62 million m3, about 31% of the communities are connected to the sewer system (Palestinian Water Authority , 2011). About 15% of this wastewater is treated in six major wastewater treatment plants (capacities range from 1,500 to 15,000 m3/d) and 11 small wastewater treatment plants (10-120 m3/d capacity) (Judeh, 2015). One of the major WWTP use solar energy as energy source which is Jericho WWTP, which has 440 PV solar panels generating approximately 100 kWh per day in summer days (Khalaf, 2015). Nablus West wastewater treatment plant serves the western parts of Nablus city and the nearby villages. It was constructed by the German Government fund through the German Development Bank (KfW) (Palestinian Water Authority , 2011). West Nablus WWTP serves a total population of 110,000 capita. Treatment system is conventional Activated Sludge System, with an actual average flow of about 11,000 m3/day where the design flow is 15,000 m3/day (Homaeidan, 2014). Nablus West plant contains two main lines; first one is wastewater treatment line including: grit chamber, primary sedimentation tank, aeration tanks, final sedimentation tanks, filtration and disinfection. Where the second one is sludge treatment line including thickener, anaerobic digester, sludge 25 drying basin, sludge storing, liquor storage tank, gas holder and gas flare (Abu-Ghosh et al., 2014). Influent wastewater to West Nablus WWTP is mostly domestic wastewater with few slaughter houses wastewater and three halva factories. Average influent characteristics are: COD=990 mg/L, BOD=400 mg/L, TSS=410 mg/L, pH=7.8 and conductivity=1500 𝜇s/cm (Nablus Municapility, 2016). Samples were collected from the effluent of primary sedimentation tank; in order to get rid of unnecessary solids (See Figure 3.2). Figure 3.2 Wastewater sampling from primary effluent from Nablus West WWTP 26 3.2. Methodology This study was performed mainly by laboratory work, all laboratory work was done in Environmental Engineering Laboratory of the Civil Engineering Department at An-Najah National University, Nablus. Laboratory work was performed in two stages: the first stage consists of all experiments that have be done to investigate parameters that affect MFC performance, i.e. salt bridge diameter and electrode materials. While in the second stage, four different COD primaries effluent wastewater-MFCs, for each concentration three MFCs duplicates were constructed, all MFCs were operated and monitored for 30 days in addition to 15 days-startup period. The research methodology is presented in the following sections: 3.2.1. Experimental Setup and Design of Experiments The main goal of this study is to investigate MFCs behavior under different parameters and trying to characterize the relation of the output power of MFCs with substrate COD. Double chambered-MFCs were used, with carbon brushes electrodes (purchased from pumps spare parts market), salt bridges for protons exchange, primary effluent wastewater as substrate and anaerobic sludge as source of anaerobic microorganisms; it worth to mention that sludge will cause substantial increase of the substrate COD. To obtain various range of data, four MFCs were fed with different COD- substrate’s. To obtain more precise results, three duplicates of each concentration were used; so, twelve MFCs were constructed and operated at 27 total. Table 3.1 shows the details of the used substrates for each concentration. COD for the anaerobic sludge and the used wastewater were measured using titrimetric method according to “Standard Methods for the Examination of Water and Wastewater” (American Public Health Association, American Water Works Association, & Water Environment Federation, 2012); they were found to be 48,000 mg/L and 547 mg/L respectively as presented in Appendix A. Volume of substrate used for each MFC is 800 mL, COD for the mixed sludge-wastewater was selected to be within 300-1700 mg/L; volume of the sludge was calculated using equation 3. COD total = (WW VolumeX WW COD) + (Sludge Volume X Sludge COD) 𝑇𝑜𝑡𝑎𝑙 𝑣𝑜𝑙𝑢𝑚𝑒 (3) Table 3.1 Details of substrates used in the MFCs Number WW Volume (mL) Distilled Water Volume (mL) Sludge Volume (mL) Expected Mixture COD (mg/L) 1 500 300 0 342 2 795 0 5 844 3 785 0 15 1437 4 780 0 20 1733 3.2.1.1. Preparation of Chambers MFCs were constructed using glass jars for both anode and cathode chambers as shown in Figure 3.3. Volume of the used jars was 1 liter, 28 whereas the used volume was 800 mL; this is to prevent substrate and/ aerated water from dropping outside the jars during mixing/shaking. Figure 3.3 Schematic of DS-MFC used 3.2.1.2. Electrodes Preparation The first experiment of the part was aimed to investigate different electrode materials effect on MFCs performance. Four types of electrodes (Zinc plate, Copper plate, Carbon brushes and pre-treated Carbon electrodes) were prepared and inserted into twelve different WW-MFCs, three duplicates for each one of the four electrodes, dimensions of all electrodes were 7 cm X 24 cm. DS-MFC were used in this step, salt bridges of 16 mm diameter and filled with KCl salt solution with 1 M concentration used and dissolved oxygen in distilled water by fish aerators was used as cathodic solution. Area of anodes where equal in all MFCs and equal to 28 cm2. Carbon brushes were selected to be used in MFCs in the second stage of this study. Carbon brushes electrodes were treated by soaking them in 1 M-HCl 29 solution for 24 hours in order to increase their electro-efficiency. Electrodes were then soaked into mixed solution of anaerobic sludge and primary effluent wastewater (20% sludge in terms of volume) at 35 oC for 3 days; in order to allow culture to form on the anodes surface. 3.2.1.3. Preparation of Salt Bridges Salt bridges were used as proton exchange media due to its low cost and availability comparing to PEMs. Pyrex Glass tubing were used as structure of the bridges; because it is an inert material chemically and electrically. U-shaped salt bridges were constructed from straight glass tubes and using Bunsen flame for bending to form U-shaped tubes (See Figure 3.4). Procedure followed in preparing of 1 M KCl-salt bridges can be summarized as:  Preparation of glass tubes (30 cm length for each).  Use Bunsen flame to bend glass tubes at two points (first and second thirds along tubes), bending angle was 90o to form U-shaped tubes.  Allow bent tubes to cool in room temperature for excessive time of 24 hours (See Figure 3.4).  Preparation of salt solution by weighing 74.5513 gm of 99% pure KCl using accurate balance (ABT 100-5M) and then adding salt to 1 L distilled water in a 2 Liter beaker.  Salt solution was heated and mixed using hot plate mixer, then agar is added gradually to the solution (3 % weight to volume ratio). 30  The beaker was kept covered during heating and mixing.  Keep the solution heated and stirred until all the agar become dissolved, the solution become clear and bubbles are just beginning to form.  Once the solution prepared, filling U-shaped tubes with solution was performed immediately using funnel and 200 mL-beaker.  Mineral wool was used to close the ends of the glass tubes, and then salt bridges were immediately place in 1M-KCl solution to prevent agar in the tubes from shrinking during cooling (See Figure 3.5).  After cooling (approximately 2 hours), salt bridges were stored in the refrigerator for few days before using them. Figure 3.4 Prepared glass U-shaped tubes for use in salt bridges 31 Figure 3.5 Prepared salt bridges soaked in KCl solution 3.2.1.4. Temperature Control System Temperature control system was constructed and applied to the hood were all MFCs were put; in order to keep temperature within 34-36 oC during operation of the MFCs. Temperature control system consists of: Arduino, two water proof temperature sensors, two air sensors, heater, Bluetooth device for monitoring and microcontroller. 3.2.1.5. Mixing System One of the obstacles that was faced in this study is mixing of WW and sludge. Hot plate stirrer was non-practical solution to solve this dilemma due to two reasons: required number of stirrers was 12 which is impossible to get, the second reason that the mechanism by which hot plate stirrer works is 32 questionable for our application; i.e. hot plate stirrer works by applying magnetic field which can affect electrical behavior of the MFCs. Mechanical system was the main suggestion to solve mixing problem. The first idea was to construct vertical or horizontal mixers that connected to the 12 MFCs with one strong motor; but this idea was rejected because of the existence of salt bridges and electrodes in the anodic chambers. Finally, shaking plate was proposed and shown to be the best solution. A mechanical shaker was designed and constructed to perform mixing task for the samples, it consists of: geared motor, transition mechanism, bearing, caster wheel and movable box. See Figure 3.6. Figure 3.6 Used mixing system 3.2.1.6. Wastewater and sludge Sampling Primary effluent wastewater was used as substrate for all the performed experiments in this thesis, the reason for that was to avoid solids in the primary sludge despite that some organics are lost with sludge; i.e. COD is expected to be decreased by about 20-30 % through primary sedimentation. 33 Since that many areas in Nablus have combined sewer system for WW and storm water, and storm water can enter the WWTP and then cause dilution for the WW, sampling of wastewater from Nablus West WWTP was done in non-rainy days, and after at least 72 hours of any raining fall, to assure that no storm water is mixed with the collected wastewater. Sampling of WW was done from the weirs of the primary sedimentation tank in the WWTP and from various locations along the weir. WW is collected in a cleaned plastic container with a volume of 10 liters, and then stored at 35 oC till use after one-two days, in order to assure keeping microorganisms activity. Sludge was collected in glass container from the anaerobic digester in the WWTP and stored at 35 oC until use after one-two days. Sampling from MFCs was performed using pipette; in all sampling times, approximately half of the sample is collected from the first top third of the anodic chamber and the rest from the second top third of the chamber. 3.2.1.7. Cathodic chamber preparation Oxygen is the best efficient electron acceptor as mentioned before. In this experiment, the cathodic solution used was aerated water. Distilled water was used, 800 mL distilled water was used in each cathodic chamber. Evaporation of distilled water was noticed due to heater and aeration effect; so cathodic chambers were refilled on daily basis. Fishing aerators were applied to aerate the cathodic chambers; operating was done for 15 minutes intervals; each operating interval was followed by 34 15 minutes break in order to prevent exhausting of the fishing aerators. It was assured that the amount of aeration is approximately equal in all cathodes. 3.2.1.8. Electrical Panel Preparation To ease the voltage and power measurements, an electrical panel was prepared. For each MFC, a 1000-ohm resistance was fixed at the electrical panel and connected with the electrodes with copper wires, for all MFCs the length of the copper wires was equal (1 meter). 3.2.2. Experimental Program The experimental program is discussed in the following sections. 3.2.2.1. Investigation of Effect of Electrode Material MFCs with different electrodes (Zinc, copper, manufactured carbon and carbon brushes) were operated for 7 days, and behavior of each MFC was inspected in terms of output voltage. Voltcraft M-3860M multimeter (manufacturer: METEX) was used to measure output open circuit voltage on daily basis. 3.2.2.2. Investigation of Effect of Salt bridges characteristics 35 Three different diameters salt bridges (10, 16 and 24 mm) were studied in DS-MFCs using carbon brushes electrodes and 1M-KCl salt. They were compared to each other in terms of output power. Two types of salt solutions, 1 M KCl and 1 M NaCl, were used in WW-MFC and compared each to other in terms of output voltage. Then, three different concentrations of KCl solutions (1M, 2M, and 3M) were used in WW-MFCs, three duplicates for each concentration, and compared each to other in terms of output voltage. Efficiency of salt bridges depends on its diameter, length, and type of salt used and concentration of salt solution. Salt bridges length was specified practically, in a way that the ends of the bridges were approximately reach the middle of the chambers. For the selection of the other characteristics of the salt bridges, three experiments were performed. All output voltage measurements were performed on daily basis for all salt bridge experiments. 3.2.2.3. Operation of the Second Stage Experiment Twelve MFCs were constructed and fed mainly with wastewater and small quantity of anaerobic sludge. Metabolic behavior is highly affected by the surrounding conditions such as temperature and pH. Anaerobic sludge was used in this study as source of anaerobic microorganisms, it was collected from anaerobic digester in which the temperature is more than 40 oC and it was there for 12–60 days depending on the temperature. Since that temperature in the anaerobic digester is near thermophilic conditions (40 oC – 60 oC), then it is expected that mesophilic 36 microorganisms (20 oC – 40 oC) are very weak in the anaerobic sludge, therefore a startup period is required in order to assure that mesophilic microorganisms were become strong enough (Bitton, 2005). After mixing WW with anaerobic sludge, 15 days startup period was performed at 35 oC temperature and without connecting salt bridges and electrical circuit in order to allow microorganisms to adapt with the experiment conditions. After startup of the experiment, COD for each MFC was measured each two- day using titrimetric method. Output voltage was measured on daily basis using Voltcraft M-3860M multimeter. 3.2.3. Analytical Procedures and Measurements: Since that the main idea of this research is to find the relationship between organic contents in the different COD-WW samples and the output power, then two types of measurements were concerned; quality (environmental) measurements and energy measurements. It worths to mention that environmental measurements were performed according to standard methods for examination of water and wastewater (American Public Health Association, American Water Works Association, & Water Environment Federation, 2012). Details of all measurements performed are summarized in the following sections. 3.2.3.1. Environmental Measurements 37 The main environmental/quality measurement in this study is organic matter contents or COD. COD measurements were taken one time each 48 hour as following:  COD measurement: COD was measured on 48 hours basis during the operating of the MFCs started at the end of startup period in the second stage. COD measurements were performed according to standard methods for examination of water and wastewater (see Appendix A) (American Public Health Association, American Water Works Association, & Water Environment Federation, 2012). COD was calculated according to the following equation: COD (mg O2/l) = (𝐴−𝐵) 𝑋 𝑀 𝑋 8000 𝑚𝐿 𝑠𝑎𝑚𝑝𝑙𝑒 Where: A: mL of F.A.S used for blank, B: mL of F.A.S used for the sample, M: molarity of F.A.S (0.05 M) and 8000: milliequivalent weight of oxygen. 3.2.3.2. Voltage and Power Measurement Voltcraft M-3860M Multimeter was used to measure output voltage (See Figure 3.7). 38 Figure 3.7 Output voltage measurement Writing down the measurements was performed after the reading fixed on the multimeter screen. 3.2.4. Kinetic models for COD decay in the MFCs The commonly used kinetic models for environmental applications are 0, 1st, 2nd and 3rd order kinetic equations: Zero order kinetic reaction: CA= Co-Kot First order kinetic reaction: CA= Coe -K1t Second order kinetic reaction: (1/CA) = (1/ Co) + K2t Third order kinetic reaction: (1/ CA)2 = (1/ Co) 2 + 2K3t Since we have COD vs time data, we can use the data to obtain the most suitable kinetic model for each MFC. The four kinetic equations can be linearized by finding Kt value for each point (COD, time) and plotting t vs 39 Kt for the four kinetic models and determine which the most representative model is. Linearize kinetic models, obtained: - 0 order: (Co- CA) = Kot, - 1st order: Ln(Co / CA)= K1t - 2nd order: ((Co/CA)-1)/ Co=K2t, - 3rd order: 0.50(CA -2 - Co -2) =K3t. 3.2.5. Data management and analysis: Obtained measurements data were analyzed statistically using testing hypothesis procedures in order to obtain reliable results. One-way ANOVA was used as a tool for statistical analysis to find the significant differences between different conditions investigated. Linearization using EXCEL was utilized to characterize COD kinetic model. Excel was used also to find the relationship between COD of substrate in MFC and output voltage of the same MFC. 40 4. Results and Discussion This study consists of two stages; first one is to investigate the effect of different parameters/conditions on the performance of MFC where the second one is to characterize the relationship between organic content in the substrate used in MFC with output power generated by MFC. Results of both stages are summarized in the following sections. 4.1. Effect of different conditions on MFC performance In this stage four individual parameter were investigated, one experiment is performed to investigate each parameter. 4.1.1. Effect of electrode material on output voltage In this experiment four different electrodes, were used in 4 different MFCs, three MFC-duplicates for each electrode. Used electrode materials were: Zinc sheets, Copper sheets, manufactured pretreated carbon electrodes and Carbon brushes. All used electrodes have the same dimensions (7 cm X 4 cm). Figure 4.1 represents the measurements of the output voltage for each electrode material during 7 days of operation. Output voltage measurements are provided in Table A1 provided in Appendix A. 41 Figure 4.1 Output voltage in MFCs of different electrode materials As shown in Figure 4.1, Copper sheets electrode-MFC looks to be the most efficient one, then Carbon brushes-MFC, then Zinc sheets-MFC and finally manufactured Carbon electrodes-MFC. Statistical analysis is used to approve or to disclaim the initial conclusion from the previous figure. One-way ANOVA and post hoc tests are used here applying IBM SPSS statistics 2.0 software. The purpose here is to check if there is a significant difference between the different electrodes materials used and to allocate where the difference is. In these experiments, all conditions were fixed except the electrode material; output voltage vs time was plotted for each electrode type. Since we have one independent variable here, i.e. electrode material, we can use One Way ANOVA test. Checking the ANOVA assumptions: - Continuous dependent variable: Output voltage data can be considered as continuous data-interval continuous variable. 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 200.00 0 1 2 3 4 5 6 7 8 V O LT A G E (M V ) TIME (DAYS) Copper Electrodes Zinc Electrodes Manufactured Carbon Electrodes Carbon Brushes Electrodes 42 - Independent variable should consist of two or more groups: here we have 4 groups (4 electrode materials). - Independence of observations, which means that there is no relationship between the observations in each group or between the groups themselves: this condition is satisfied here. - There should be no significant outliers: SPSS was used to check outliers and found that there are no outliers in the data as shown in steam and leaf plots for the data provided in Appendix B, Figure B1. - Dependent variable should be approximately normally distributed: Shapiro-Wilk test was used to check normality and it was found that all the distributions are approximately normal distributed (All sig. values were more than 0.05 as shown in Appendix B, Table B1). - Homogeneity of variances: Levene’s test for homogeneity of variances was used in SPSS to check for homogeneity and it was found that this condition was satisfied as shown in Appendix B, Table B2. As a result, One Way ANOVA can be used here to compare between the 4 groups we have; Copper, Zinc, Manufactured Carbon and Carbon Brushes. Using ANOVA in SPSS it was found that there is a significance difference between electrode materials used in terms of output voltage (significance less than 0.05) as shown in Appendix B, Table B9. To allocate which electrode materials have significance difference between them, Post Hoc tests was performed on SPSS. Results are summarized in Appendix B, Table B10. It can be noticed that Copper electrodes-MFCs are significantly more efficient than Carbon brushes-MFCs and Zinc electrodes- 43 MFCs. Carbon brushes-MFCs are significantly more efficient than zinc electrodes-MFCs; whereas Zinc-MFCs are significantly more efficient than manufactured carbon-MFCs in the first two days only. It can be noticed that, in general, output voltage is decreasing with time; this can be justified by the decreasing of nutrients with time in the substrate. Variations in output voltage between duplicates for the same electrode material can be attributed to many causes: variation in substrate constituents, variation in microorganism’s cultures, and variation in salt bridges efficiency and variation in electrode position in the anode. 4.1.2. Effect of salt bridge diameter on output voltage In this experiment three salt bridges diameters, 10 mm, 16 mm and 24 mm, were used in 3 different MFCs, three MFC-duplicates for each diameter. Nine DS-MFCs were constructed, 1M-KCl salt was used, Carbon brushes electrodes were used, area of each electrode was 28 cm2 and temperature maintained to 35 oC. Primary effluent wastewater was used as substrate (790 mL) in addition to anaerobic sludge (10 mL). Dissolved oxygen in water was used as cathodic chamber and this experiment was running for one week. Figure 4.2 represents the results of this experiment. Detailed results of this experiment are shown in Appendix A, Table A2. 44 Figure 4.2 Output voltage of MFCs with different salt bridge diameters As shown in Figure 4.2, and based on statistical analysis performed, the best output voltage is achieved by the 10-mm salt bridge-MFC, then by 16 mm and the 24-mm salt bridge-MFC. The purpose here is to check if there is a significant difference between the different diameters of salt bridges used and to allocate where the difference is. In these experiments, all conditions were fixed except the salt bridge diameter; output voltage vs time was plotted for each electrode type. Since we have one independent variable here, i.e. salt bridge diameter, we can use One Way ANOVA test. Checking the ANOVA assumptions: - Continuous dependent variable: Output voltage data can be considered as continuous data-interval continuous variable. - Independent variable should consist of two or more groups: here we have 3 groups (3 salt bridge diameters). 0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00 450.00 0 1 2 3 4 5 6 7 8 V O LT A G E (M V ) TIME (DAYS) 10 mm-Daimeter 16 mm-Diameter 24 mm-Diameter 45 - Independence of observations, which means that there is no relationship between the observations in each group or between the groups themselves: this condition is satisfied here. - There should be no significant outliers: SPSS was used to check outliers and found that there are no outliers in the data as shown in steam and leaf plots for the data captioned from SPSS output provided in Appendix B, Figure B2. - Dependent variable should be approximately normally distributed: Shapiro-Wilk test was used to check normality and it was found that all the distributions are approximately normal distributed (All sig. values were more than 0.05 as shown in Appendix B, Table B3 obtained from SPSS output). - Homogeneity of variances: Levene’s test for homogeneity of variances was used in SPSS to check for homogeneity and it was found that this condition was satisfied as shown in Table B4 in Appendix B. As a result, One Way ANOVA can be used here to compare between the three groups we have; 10 mm, 16 mm and 24 mm salt bridges. Using ANOVA in SPSS it was found that there is a significance difference between salt bridges diameters used in terms of output voltage (significance less than 0.05) as shown in Table B11, Appendix B. To allocate which diameters have significant difference between them, Post Hoc tests was performed on SPSS. Results are summarized in Table B12 in Appendix B. It can be noticed that 10 mm salt bridges-MFCs are significantly more efficient than 16 mm salt bridges-MFCs; whereas 16 mm 46 salt bridges-MFCs are significantly more efficient than 24 mm salt bridges- MFCs in the first one days only. It can be noticed that, in general, output voltage is decreasing with time; this can be justified by the decreasing of nutrients with time in the substrate. Variations in output voltage between duplicates for the same electrode material can be attributed to many causes: variation in substrate constituents, variation in microorganism’s cultures, and variation in salt bridges efficiency and variation in electrode position in the anode. 4.1.3. Effect of salt bridge solution on output voltage In this experiment two salt solutions were used in salt bridge preparation, KCl and NaCl. Six DS-MFCs were constructed, 1M-KCl and 1M-NaCl salts was used in 10 mm-salt bridges, Carbon brushes electrodes were used, area of each electrode was 28 cm2 and temperature maintained to 35 oC. Primary effluent wastewater was used as substrate (790 mL) in addition to anaerobic sludge (10 mL). Dissolved oxygen in water was used as cathodic chamber and this experiment was running for one week. Figure 4.3 represents the results of this experiment. Detailed results of this experiment are shown in Appendix A, Table A3. 47 Figure 4.3 Output voltage for KCl and NaCl Salt bridges As shown in Figure 4.3 the best output voltage looks to be achieved by the KCl salt bridge-MFC. The purpose here is to check if there is a significant difference between the two different salts used. In these experiments, all conditions were fixed except the salt type; output voltage vs time was plotted for each salt type. Since we have one independent variable here, i.e. salt type, we can use One Way ANOVA test. Checking the ANOVA assumptions: - Continuous dependent variable: Output voltage data can be considered as continuous data-interval continuous variable. - Independent variable should consist of two or more groups: here we have 2 groups (2 salt types). - Independence of observations, which means that there is no relationship between the observations in each group or between the groups themselves: this condition is satisfied here. 0 50 100 150 200 250 300 350 400 0 1 2 3 4 5 6 7 8 V O LT A G E (M V ) TIME (DAYS) NaCl-Salt Bridge KCl-Salt Bridge 48 - There should be no significant outliers: SPSS was used to check outliers and found that there are no outliers in the data as shown in steam and leaf plots for the data captioned from SPSS output provided in Appendix B, Figure B3. - Dependent variable should be approximately normally distributed: Shapiro-Wilk test was used to check normality and it was found that all the distributions are approximately normal distributed (All sig. values were more than 0.05 as shown in Appendix B, B6 obtained from SPSS output). - Homogeneity of variances: Levene’s test for homogeneity of variances was used in SPSS to check for homogeneity and it was found that this condition was satisfied as shown in Table 2.15, Appendix B. As a result, One Way ANOVA can be used here to compare between the 2 groups we have; NaCl and KCl salt bridges-MFCs. Using ANOVA in SPS it was found that there is a significance difference between both salts as shown in Table B13, Appendix B. 4.1.4. Effect of salt concentration on output voltage In this experiment three KCl salt concentrations, 1M, 2M and 3M, were used in salt bridge preparation. Nine DS-MFCs were constructed, 10 mm-salt bridges were used, Carbon brushes electrodes were used, area of each electrode was 28 cm2 and temperature maintained to 35 oC. Primary effluent wastewater was used as substrate (790 mL) in addition to anaerobic sludge (10 mL). Dissolved oxygen in water was used as cathodic chamber and this 49 experiment was running for one week. Figure 4.4 represents the results of this experiment. Detailed results of this experiment are shown in Appendix A, Table A4. Figure 4.4 Output voltage for KCl and NaCl Salt bridges As shown in Figure 4.4 and based on statistical analysis the best output voltage is achieved by the KCl salt bridge-MFC. The purpose here is to check if there is a significant difference between the different salt concentrations used and to allocate where the difference is. In these experiments, all conditions were fixed except the salt concentration in the salt bridges; output voltage vs time was plotted for each concentration. Since we have one independent variable here, i.e. salt concentration, we can use One Way ANOVA test. Checking the ANOVA assumptions: - Continuous dependent variable: Output voltage data can be considered as continuous data-interval continuous variable. 0 100 200 300 400 500 600 0 1 2 3 4 5 6 7 8 V O LT A G E (M V ) TIME (DAYS) 1M KCl 2M KCl 3M KCl 50 - Independent variable should consist of two or more groups: here we have 3 groups (3 concentrations). - Independence of observations, which means that there is no relationship between the observations in each group or between the groups themselves: this condition is satisfied here. - There should be no significant outliers: SPSS was used to check outliers and found that there are no outliers in the data as shown in steam and leaf plots for the data captioned from SPSS output provided in Appendix B, Figure B4. - Dependent variable should be approximately normally distributed: Shapiro-Wilk test was used to check normality and it was found that all the distributions are approximately normal distributed (All sig. values were more than 0.05 as shown in Table B7, Appendix B obtained from SPSS output). - Homogeneity of variances: Levene’s test for homogeneity of variances was used in SPSS to check for homogeneity and it was found that this condition was satisfied as shown in Table B8, Appendix B. - As a result, One Way ANOVA can be used here to compare between the 3 groups we have; 1M, 2M and 3M KCl salt bridges-MFCs. Using ANOVA in SPSS it was found that there is a significance difference between salt concentrations used for the salt bridges in terms of output voltage (significance less than 0.05) as shown in Table B14, Appendix B. 51 - To allocate which groups have significance difference between them, Post Hoc tests was performed on SPSS. Results are summarized in Table B15, Appendix B. It can be noticed that 1M KCl salt bridges- MFCs are significantly more efficient than 3M KCl salt bridges- MFCS. 1M KCl salt bridges-MFCs are significantly more efficient than 2M KCl salt bridges-MFCs in the first three days only; the same thing as in a comparison between 2M KCl salt bridges-MFCs and 3M KCl salt bridges-MFCs. 4.2. COD Determination and Determining Kinetic Models for COD Removal COD was measured using titrimetric method according to “Standard Methods for the Examination of Water and Wastewater” (American Public Health Association, American Water Works Association, & Water Environment Federation, 2012). Dilution of samples using distilled water was applied when the expected COD higher than the allowed range (40-400 mg/L). Samples were mixed with: standard potassium dichromate digestion solution, sulfuric acid reagent and sulfamic acid in quantities as given in the standard methods. After heating of mixture for 2 hours at 150 oC and cooling, the mixture was titrated against standard ferrous ammonium sulfate (F.A.S) and using ferroin as indicator. COD was measured for all MFCs substrates day after day, at each time a blank sample was prepared from distilled water and the same reagents used for the samples. 52  COD measurement -COD of anaerobic sludge used Since that the expected COD of the sludge is too large; it is required to dilute sludge samples using distilled water. 10 mL of anaerobic sludge were taken from the well mixed sample using 10 mL-graduated cylinders, and diluted with 990 mL distilled water. The resulted dilution factor is: P1= initial volume of sample/ final volume = 10/ (990+10) = 0.01 The resulted 1 L sample was well mixed using magnetic stirrer and three COD samples, each 1 mL, were taken. The samples were taken from top, middle and bottom of the beaker containing sample. According to Table 4.2220: I in “Standard Methods for the Examination of Water and Wastewater”, when ampules are used the required volumes of reagent are: 2.5 mL sample, 1.5 mL digestion solution and 3.5 mL sulfuric acid reagent. Lower sample volume can be taken and diluted to 2.5 mL to detect higher COD values. Sample volume taken from diluted sludge were 1 mL for each, so the total dilution factor is: P=P1 X P2= .01 X (1/2.5) = 0.004 Results are represented in the following table: 53 Table 2.1 Summary of sludge COD measurement Sample A (mL) B (mL) A-B (mL) Dilution factor COD (mg/L) 1 2.9 1.6 1.3 0.004 52,000 2 2.9 1.8 1.1 0.004 44,000 3 2.9 1.7 1.2 0.004 48,000 Blank 2.9 Average COD for the three samples (mg/L) 48,000  COD measurement -COD of used wastewater COD of the primary effluent was expected to be around 500-700 mg/L. Dilution was performed to assure that the measured COD is within allowed range (40-400 mg/L). Volume of sample, taken from the well mixed 10 L origin sample, was 1 mL and diluted into 2.5 mL using distilled water; so, dilution factor is 0.40. Results are presented in the following table. Table 4.2 Summary of wastewater COD measurement Sample A (mL) B (mL) A-B (mL) Dilution factor COD (mg/L) 1 2.7 1.1 1.6 0.4 640 2 2.7 1.4 1.3 0.4 520 3 2.7 1.5 1.2 0.4 480 Blank 2.7 Average COD for the three samples (mg/L) 547  Sampling calculations Since that titrimetric method for COD determination is applicable within the range (40 mg/L < COD < 400 mg/L) and dilution factor was taken 0.40 (1 mL sample diluted into 2.5 mL) then the expected COD of samples should be lie in this range and considering used dilution factor. 54 At determined COD= 40 mg/L and dilution factor= 0.40, the origin COD is 40/0.4 = 100 mg/ L. The same as at COD=400 mg/L and dilution factor= 0.2, the origin COD is 2000 mg/L. Resulting that COD of samples should be within the range 200 to 2000 mg/L. 300, 900, 1400 and 1700 mg/L were selected considering the previous points. Samples contain WW and sludge, considering that sludge will increase total COD of the samples and using the following equation: COD = (WW VolumeX WW COD)+(Sludge Volume X Sludge COD)+(D.W. 𝑉𝑜𝑙𝑢𝑚𝑒𝑋𝐷.𝑊. 𝐶𝑂𝐷) 𝑇𝑜𝑡𝑎𝑙 𝑣𝑜𝑙𝑢𝑚𝑒 Where: WW volume: volume of WW used in the mixture (mL), WW COD: COD of the influent WW, measured to be 547 mg/L, Sludge volume: volume of sludge used in the mixture (mL), Sludge COD: COD of the used sludge and equals 48,000 mg/L, D.W. Volume: volume of distilled water used to dilute samples in low COD samples (mL), D.W. COD: COD of distilled water and equals to 0 mg/L, Total volume: volume of the total mixture and equals to 800 mL. To determine the kinetic models for COD decay in each MFC, using the approach discussed in section 3.2.4. After linearization of each model for each MFC, find R2 value for each model using Excel as shown in Tables D1-D4 (Appendix D) and Figures 4.5-4.20: 55 MFCs 1, 2 and 3 Figure 4.5 Comparison between Kt values of third order reaction and linearized Kt values for COD behaviour in MFCs 1,2 and 3 Figure 4.6 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 R² = 0.6091 -0.000050 0.000000 0.000050 0.000100 0.000150 0.000200 0 5 10 15 20 25 30 35 Kt Time (days) Kt vs time-3rd order Kt vs time-3rd order Linear (Kt vs time-3rd order) R² = 0.8418 0.000000 50.000000 100.000000 150.000000 200.000000 250.000000 0 2 4 6 8 10 12 14 16 18 K t Time (days) Kt vs time-0 order 56 Figure 4.7 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 Figure 4.8 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 1, 2 and 3 2nd order is the best suitable model for MFC1. MFCs 4, 5 and 6: R² = 0.8388 0.000000 0.200000 0.400000 0.600000 0.800000 1.000000 1.200000 1.400000 1.600000 1.800000 0 5 10 15 20 25 30 35 K t Time (days) Kt vs time-1st order R² = 0.9191 0.000000 0.001000 0.002000 0.003000 0.004000 0.005000 0.006000 0 2 4 6 8 10 12 14 16 18 K t Time (days) Kt vs time-2nd order 57 Figure 4.9 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 Figure 4.10 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 R² = 0.6424 -0.000020 0.000000 0.000020 0.000040 0.000060 0.000080 0 5 10 15 20 25 30 35 Kt Time (days) Kt vs time-3rd order Kt vs time-3rd order Linear (Kt vs time-3rd order) R² = 0.8375 0.000000 50.000000 100.000000 150.000000 200.000000 250.000000 300.000000 350.000000 400.000000 450.000000 0 5 10 15 20 K t Time (days) Kt vs time-0 order 58 Figure 4.11 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 Figure 4.12 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 4, 5 and 6 1st order kinetic model is the best suitable model for MFC2. MFCs 7, 8 and 9 R² = 0.9527 -0.500000 0.000000 0.500000 1.000000 1.500000 2.000000 2.500000 0 5 10 15 20 25 30 35 K t Time (days) Kt vs time-1st order R² = 0.8814 0.000000 0.000500 0.001000 0.001500 0.002000 0.002500 0 2 4 6 8 10 12 14 16 K t Time (days) Kt vs time-2nd order 59 Figure 4.13 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 Figure 4.14 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 R² = 0.6536 -0.000005 0.000000 0.000005 0.000010 0.000015 0.000020 0 5 10 15 20 25 30 35 Kt Time (days) Kt vs time-3rd order Kt vs time-3rd order Linear (Kt vs time-3rd order) R² = 0.9123 0.000000 200.000000 400.000000 600.000000 800.000000 0 5 10 15 20 K t Time (days) Kt vs time-0 order 60 Figure 15 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 Figure 16 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 7, 8 and 9 1st order is the most suitable model for MFC3. MFCs 10, 11 and 12: R² = 0.9599 -1.000000 0.000000 1.000000 2.000000 0 5 10 15 20 25 30 35 K t Time (days) Kt vs time-1st order R² = 0.9144 0.000000 0.000200 0.000400 0.000600 0.000800 0.001000 0.001200 0 2 4 6 8 10 12 14 16 K t Time (days) Kt vs time-2nd order 61 Figure 4.17 Comparison between Kt values of third order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 Figure 4.18 Comparison between Kt values of zero order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 R² = 0.7008 -0.000002 0.000000 0.000002 0.000004 0.000006 0.000008 0 5 10 15 20 25 30 35 Kt Time (days) Kt vs time-3rd order Kt vs time-3rd order Linear (Kt vs time-3rd order) R² = 0.9731 0.000000 200.000000 400.000000 600.000000 800.000000 1000.000000 0 2 4 6 8 10 12 14 16 K t Time (days) Kt vs time-0 order 62 Figure 4.19 Comparison between Kt values of first order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 Figure 4.20 Comparison between Kt values of second order reaction and linearized Kt values for COD behavior in MFCs 10, 11 and 12 2nd order kinetic reaction is the most suitable one for MFC4. 4.3. Relationship between output voltage and substrate COD Data obtained from the second part of this study are provided in Appendix D. It can be noticed that there is a relationship between COD and output voltage. Figure 4.21 shows the plot of all data points obtained. R² = 0.9728 0.000000 0.500000 1.000000 1.500000 2.000000 0 5 10 15 20 25 30 35 K t Time (days) Kt vs time-1st order R² = 0.9871 -0.000100 0.000000 0.000100 0.000200 0.000300 0.000400 0.000500 0.000600 0.000700 0.000800 0.000900 0 2 4 6 8 10 12 14 16 K t Time (days) Kt vs time-2nd order 63 Figure 4.21 Relationship between COD and output voltage Since that this is a nature process, the best mathematical model represents this data is natural logarithmic model. COD (mg/L) = 229.85 Ln (V)-1039.6; where V is output voltage (mV). This model can be used to indicate COD of a certain wastewater sample, by measuring the output voltage of a MFC operated by that sample, and considering all conditions of the experiments performed to obtain this model. It worth to mention that COD removal behavior for the four different MFCs was ranged between 1st and 2nd order kinetic reaction. Actually, COD behavior is 1st order kinetic. In the analysis of COD data with time, it was found that MFC 1 and 4 are following 2nd order with a minor difference from 1st order. Where in MFC 2 and 3 are following1st order reaction. These results indicate that our data, to somehow, are logical. y = 229.85ln(x) - 1039.6 R² = 0.8773 -200 -100 0 100 200 300 400 500 600 700 800 0 200 400 600 800 1000 1200 1400 1600 O u tp u t V o lt ag e (m V ) COD (mg/L) COD vs output voltage 64 4.4. Achieved output power and feasibility estimation Output power can be calculated as: P = I X V Where: P: output power, watt, I: Electrical current through the load (resistance), Amber, V: output voltage, Volt Electrical current can be calculated as: I=V/R Where R is the applied resistance, Ohm Used Resistances were 1000 ohm. Maximum output voltage achieved in this study was 0.684 volt. Calculating power for the maximum voltage: I=0.684/1000=0.684 mA=0.000684 A. P=0.000684 X 0.684=0.00046786 Watt. Normalize the output power to the anode surface area, anode surface area was 28 cm2= 0.0028 m2, resulting: Normalized power = P/ Anode surface area = 0.00046786/ 0.0028 = 0.71 W/m2 Normalized power to the used volume of WW= P/volume = 0.00046786/ 0.0008 = 0.585 W/m3. 65 During 30 days of operation, average output voltage is 448 mV; so average P= 0.000201 W and volume normalized power is 0.2509 W/m3, calculate output power for 30 days: Output power= 0.2509 W/m3 X 30 days X 24 hr/day = 140.4 Wh =180.65 WH/m3. Comparing these results to the data presented in Table 1 and Table 2, found that maximum output power in this research is higher than several previ