Validation of RSM Predicted Optimum Scaling-Up Factors for Generating Electricity in a DCMFC: MATLAB Design and Simulation Model

Abstract

In this present study, the potential application of DCMFC for the treatment of three different sourced industrial wastewater streams: biorefinery, dairy and mixed streams was investigated. Operating conditions were optimised using the Box Behnken design in response surface methodology (RSM) with three validation experimental runs. The effect of process variables, i.e., HRT (48 h), catholyte dose (0.1 gmol/L) and electrode surface area (three carbon rods argumentation-m²) on the production of electricity as voltage yield (mV), power density (mW/m²), current density (mA/m²), Columbic efficiency (%) CE and Gibbs free energy correlation with the electromotive force of the DCMFC system. Experimental results obtained were a positive response towards the predictive values according to the DoE numerical optimisation sequence. At numerical optimum MFC conditions stated above, validation experimental responses of voltage yield by biorefinery wastewater were 645.2 mV, mixed wastewater was 549 mV, and dairy wastewater was 358 mV maximum yields. The power densities and current densities were attained, for biorefinery, mixed wastewater and dairy wastewater sources respectively as; 62 mW/m², 50 mW/m² and 27.2 mW/m², then current densities of 50 mA/m², 44,008 mA/m² and 18 mA/m². The coulombic efficiencies of 0.34%, 0.75% and 0.22%, respectively, were achieved. The validation of predicted optimum operating conditions was successfully attained, especially through the biorefinery wastewater organic substrate. This article articulates that it is highly imperative to choose the most suitable wastewater source as the viable electron donor towards scaling up and maximising the efficiency of generating electricity in the double chamber microbial fuel cell (DCMFC). Moreover, the findings of the current study demonstrate that the DCMFC can be further upscaled through a series connection in a fed-batch mode of operation using a well-designed and simulated process control system that has been computationally designed and modelled using first order MFC model bioenergy generating models MATLAB Simulink and Simscape electrical software. These findings of the simulations were successful and illustrated that an MFC power output can be successfully stepped to be a viable bio-electrochemical technology for both industrial wastewater (IWW) treatment and simultaneous sustainable power generation.

1. Introduction

Industrial wastewater treatment seems to be challenging for conventional wastewater treatment processes. Hence, different processes (i.e., biological, physical, chemical, electrochemical, and bio-electrochemical technologies) have been investigated; however, reports on parametric optimisation using statistical tools are scanty [1,2,3,4,5,6,7,8]. In the last couple of decades, microbial fuel cells (MFCs) have gained importance because of their ability to generate electricity from renewable and carbon-free energy sources, such as raw wastewater sources. The occurrence of simultaneous biological and electrochemical processes to facilitate the electron transfer mechanism increases the process thresholds and grey areas. Microbial fuel cells (MFCs) are a promising sustainable technology to address growing energy demands, as they can simultaneously treat and produce power from wastewater and can degrade organic pollutants strengths [9]. The integral components of a typical MFC are the anode, cathode, and proton exchange membrane (PEM)/cation exchange membrane (CEM). Active biocatalyst (bacterial species/active biomass consortia) oxidises the raw or organic complex substrates in the anodic chamber, thereby producing electrons and protons [10,11,12,13,14]. Protons selectively pass through the CEM/PEM, and the electrons flow through the external circuit to the cathode, normally a copper line or proper connection load electrical connection lines. These electrons flow through to the cathodic chamber, where they combine with a viable electron acceptor, most rudimentarily, oxygen, to form water, thus completing the bioelectrochemical technology of producing electricity [15,16].

This novel process is capable of not only reducing the dependency on fossil fuels and the adverse environmental impacts associated with it but also providing a sustainable, renewable, and reliable solution for wastewater treatment [4]. Moreover, there remains a need to overcome several limitations before its commercialisation and practical applications. The main bottleneck in the MFC operation is the low power voltage yields [17]. Furthermore, high start-up times and operating costs of materials, e.g., Nafion membrane as either CEM/PEM, electrodes, and exogenous materials, such as catalysts) in MFCs make it less economically convenient [18,19,20,21,22,23,24]. For improving the performance of the MFC system, it is vital to distinguish and identify the viable major operating factors that affect the MFC scaling-up efficacy. There are several factors that affect the performance of the MFC unit, such as the nature and population nature of microbial species, the anodic chamber configurations, the exogenous mediators, the type of substrate source and its organic concentrations or assays, the operating condition (pH, external resistance, temperature, ionic strength of catholyte, aeration dissolved oxygen dosages etc.) and the design and configuration of the MFC unit itself [12,13,25,26,27,28,29,30,31,32]. Optimisation of the MFC operating parameters will not only curtail the operational cost but also improve its overall voltage yield and viability.

It is challenging to experimentally establish the ideal conditions for power and energy generation and effective substrate removal in this hybrid type of reactor because of the complex interplay between many design and operating variables. Thus, mathematical models offer a straightforward method to explore the effects of various variables and enhance MFC performance. Since anode reactions are characteristics of MFCs and MFC performance may be partially anticipated from the growth condition of the anode-attached bacteria, the electrochemical reactions at the anode are highlighted in MFC modelling [24,33]. The current MFC anode models are typically built around purportedly present redox mediators in the medium [34,35]. Bacteria that are physically linked to the anode surface can operate as catalysts because research has already shown that their outer membrane-bound cytochromes and nanowires conduct electricity [36]. Thus, it may be inferred that the only bacteria that contribute to electricity production are those that are directly anode-attached or bacteria that form a monolayer [34,35]. The connected and suspended bacterial populations attain a dynamic equilibrium with one another, and both populations consume substrate for growth; hence, the suspended bacteria must also be accounted for in the model. When examining substrate consumption within populations, this inclusion is very crucial. This article explores the dynamics of the Freter model, which originally explains the dynamics of floating bacteria and bacteria adhered to the wall of a bioreactor [33,34,35,37].

In this article, significant experimental and modelling techniques have been carried out to identify and predict its relative impact on improved electricity generation and overall performance of an MFC unit towards a possibility for commercialization, practicality, and applicability purposes. Among the various modelling techniques, the Response Surface Methodology for a precise design of experiments (DoE) was used as a pragmatic tool to determine the correlation between operating parameters that influence the scaling up of the MFC unit. RSM has a few advantages over the one-factor-a-time (OFAT) methodology in terms of time and resource management as well as obtaining relevant outcomes. This article articulates the significance of validating the DoE-predicted optimum operating conditions combined. Moreover, the optimisation of the MFC unit is further carried out via MATLAB statistical software for designing, rebuilding, and simulating the best possible scaling-up outcomes via a precise scaling-up inverter-based model of the MFC unit, which is further compared to the reputable Freter-based MFC energy model and Conduction Based MFC model for its potential to generate new knowledge with a potential of future novel recommendations on MFC technology.

2. Materials and Methods

2.1. MFC Experimental Material

Figure 1 illustrates the lab view images of the DCMFC with its distinctive components utilised during the validation experimental runs. The MFC chamber is a dual-chamber design that was fabricated from 2 × Schutt blue cap 1 L bottles. The anode and cathode chamber were separated by Nafion 115^® Proton Exchange Membrane (PEM) (Lyntech, College Station, TX, USA) to transfer the proton (H⁺) from the anode chamber to the cathode chamber. A unique bi-electrode electrode (CCU-electrode) was proposed using copper strips that were 0.15 m long and 0.05 m wide. One to three carbon rods with a diameter of 0.0003 m and a length of approximately 0.15 m were used to mount the copper electrode. The name of this electrode was either 1-CCu or 3-CCu electrodes. The idea was to adjust the electrode surface area in accordance with the cross-sectional area of the 1-CCu-3-CCu electrode. The bi-electrode connection was then completed by an external circuit using professional electrical positive and negative wires to transfer electrons produced from organic matter oxidation from the anode chamber to the cathode chamber. An external load of 1000 Ω external resistance was added. The cathode compartment was filled with a sterile substrate solution, Potassium Permanganate (KMnO₄) solution, which was combined with slow oxygen sparging as electron acceptors.

The MFC performance was monitored by measuring electrical current (I) and voltage (V) using a multi-meter every 5 h for a precise 48 h optimum incubation period. The magnitude between current and voltage was authorised by the Equation [38]:

$$I \left(mA\right)=\raisebox{1ex}{$V$}\!\left/ \!\raisebox{-1ex}{$R_{ext}$}\right.$$

(1)

where R_ext was the resistance of an external resistor whose value was equal to 1000 Ω. V (mV) is the voltage capacity of the closed-circuit voltage of the MFC unit. The following sets of electrochemical equations will be taking place in both the anodic and cathodic chambers of the MFC unit as the fundamental principle of microbial electrochemistry taking place during the operation phases within the MFC, sourced [19,33,39,40,41,42,43,44,45,46,47,48,49,50,51]. Organic Matter is subsequently oxidised by microbes, electrons and protons and carbon dioxide is produced. Electrons are receipted by an anode biofilm surface and passed to the cathodic chamber by an external circuit to produce bioelectricity [52,53,54,55,56,57,58,59]. Equation (2) below depicts the anodic oxidizing bioelectrochemical reaction sequence.

Anodic Reaction:

$$\mathrm{organic} \mathrm{matter} (Raw Industrial Complex Substrates) \stackrel{(\frac{Proteobacter}{Bacteroidota})}{\Rightarrow } C{O}_2+H^{+}$$

(2)

Electron Acceptors, along with their concentration, have a significant effect on the generation of electricity in an MFC unit. In this study, unlike other previously completed work, we used Potassium permanganate (KMnO₄) as an electron acceptor. You et al. [25,60,61,62,63] presented a lab-scale MFC unit configuration with KMnO₄ and articulated the pH dependency of the system. This catholyte works on both acidic and basic configurations, as presented below:

$${MnO}_4^{-}+4H^{+}+3e^{-} \to Mn{O}_4+2H_{2}O \left(Acidic-Configuration\right)$$

(3)

$$Mn{O}_4+2H_{2}O+3e^{-}\to Mn{O}_2+4 O{H}^{-} (Basic-Configuration)$$

(4)

The findings of this journal are hypothesised to be the clear potential towards MFC technology’s practicality and applicability on a commercial scale operation. Below empirical models are fundamentals for calculating voltage generation in MFCs with partial thermodynamic models to mimic the biodegradation process favourability in exuding Gibbs free energy correlated to the overall MFC unit electromotive force. These were sourced from Logan et al. [38]

$$∆G_r \left(J\right)=∆G_r^0+RTln\left(\pi \right)$$

(5)

where $∆G_r \left(J\right)$ is the Gibbs free energy for specific conditions, $∆G_r^0$(J) is the Gibbs free energy under standard conditions, usually defined by 298.15 K, 1 bar- pressure and 1 M concentration for all species, R = (8.31447 $\raisebox{1ex}{$J$}\!\left/ \!\raisebox{-1ex}{$mol °K$}\right.$) is the basic universal gas constant and T (°K) is the absolute temperature calculated as the activities of the products divided by those reactants. The standard Gibbs free energy is calculated from tabulated energies of formation for organic compounds in water, sources available [37,55,64].

$$E_{emf} (mV)=-\frac{∆G_r}{nF}$$

(6)

where all biodegradation reactions transpire under standard conditions hence: π = 1, then the Equation simplifies further [38]:

$$E_{emf}^0(mV)=-\frac{{∆G}_r^0}{nF}$$

(7)

where $E_{emf}^0(mV)$ is the standard cell electromotive force. We can, therefore, use the above Equations to express the overall reaction in terms of potential as follows [38]:

$$E_{emf} \left(mV\right)=E_{cathode}-E_{anode}$$

(8)

where $E_{emf} \left(mV\right)$ is the cell electromotive force and the minus sign is a result of the definition of the anode potential as a reduction reaction (although an oxidation reaction is occurring). This is a thermodynamic value that does not consider internal losses, sourced from [18,19,20,22,23,24,65].

$$P \left(mW\right)=I{E}_{cell}$$

(9)

where $P \left(mW\right)$ is the overall power of the MFC unit. Normally, the voltage is measured across a fixed external resistor (R_ext, while the current is calculated from Ohm’s law: $I \left(mA\right)=\raisebox{1ex}{$E_{cell}$}\!\left/ \!\raisebox{-1ex}{$R_{ext}$}\right.$ (Logan et al. [18,19,20,22,23,24]).

$$P_{anode}(mW)=\frac{E_{cell}^0}{A_{andoe }{R}_{ext}}$$

(10)

The power output is usually normalised to the projected anode surface area because the anode is where the biological reactions transpire [12,18,19,20,21,22,23,24,28,65].

$${ϵ}_{Cb}\left(\%\right)=\frac{M{\int }_0^{t_b}Idt}{Fb{v}_{anode}∆COD}$$

(11)

where M = 32, the molecular weight of oxygen, F is Faraday’s constant =$9.6485 x {10}^4\raisebox{1ex}{$C$}\!\left/ \!\raisebox{-1ex}{$Mol$}\right.$, b = 4 is the number of electrons exchanged per mole of oxygen, V_An is the volume of the anodic chamber and $∆COD$ (mgCOD/L) is the change in COD over time t_b.

2.2. Industrial Wastewater Sample

Fresh IWW composite samples were collected from a local South African wastewater treatment plant at the effluent sampling point before the municipal dewatering point mix of the effluent open dam random sample method. Characterisation of the fresh sample was conducted in accordance with the Standard Method for the Examination of water and wastewater using a HANNA HI 9828 pH [66,67,68], oxidation-reduction potential (ORP), electrical-conductivity (EC), dissolved oxygen (DO), biological oxygen demand (BOD), total dissolved solids (TDS), salinity, resistivity, multiparameter HI 9298 and a HACH DR900 spectrophotometer for total organic carbon and total phosphates (PO₄⁻³). Turbidity was measured in Nephelometric Turbidity units (NTU). The sample composition is presented in

Table 1. Physico-chemical Properties of the three different wastewater sources used in this study.

Constituent	Clover-WW Present Study	Biorefinery-WW Present Study	Mixed-WW Present Study
pH	2.76–12.73	4–10.79	5.43–14.00
Temperature °C	18–28	18–28	18–28
COD (Chemical Oxygen Demand) mg/L	128–3455	65–1870	152–9965
TOC mg/L	63–1168.1	38–639.7	45.2–4070
BOD (Biological Oxygen Demand) mg/L	768–2073	39–1122	91.2–5979
COD/BOD ratio	1.667	1.665	1.666
VFA (Volatile Fatty Acids) mg/L	-	-	-
${PO}_4^{3-}$(Total-Phosphates) mg/L	75–674.5	13.40–161.0	36–1100
TKN (Total Kjedahl Nitrogen) mg/L	-	-	-
TS (Total Solids) mg/L	-	-	-
Turbidity NTU	59–2800	26–400	48–1866
TDS (Total Dissolved Solids) mg/L	804–6626	1405–8212	595–3824
EC (Electrical Conductivity) µS.cm	13.25–4562	2810–10,550	1191–7647
Salinity ppt	0.81–7.66	1.46–9.66	0.59–4.25
Resistivity Ω	0.0004	0.0002	0.0002
DO mg/L	2.99–20.28	5.05–45.06	11.51–47.03
pHmV	−359.8–57.30	−261.7–97.7	−428.5–19
Oxidation Reduction Potential (ORP)	−247.6–233.7	−212–46.30	−245.8–10.2

.

2.3. Design of Experiment

The Box Behnken design methods in Design-Expert version 11 were used to ascertain the number of experimental runs to be assayed for the optimization of four independent variables, i.e., Catholyte dosage $\left(X_1\right)$, HRT ($X_2)$, Temperature ($X_3)$ and Surface Area. This was performed to optimise the bioelectrical potential of a DCMFC unit in generating electricity using three purely organic industrial wastewater sources. The generation of overall voltage yield in the form of closed-circuit voltage (mV) ($Y_1)$, power density (mW/m²) ($Y_2)$, current density (mA/m²) ($Y_3)$ and coulombic efficiency (% CE) ($Y_4)$ was attained as the performance measures for this optimised technology. The removal of basic organic contaminants was also established as the measure of the DCMFC viability; the removal of chemical oxygen demand ($Y_5)$, total suspended solids ($Y_6)$ and total phosphates ($Y_7)$. The codes, mean ranges and standard deviation of the independent variables in the RSM design are shown in

Table 2. DoE on Response Surface Methodology design factors using the Box Behnken Design (BBD) Method.

Factor	Name	Units	Type	Minimum	Maximum	Coded Low	Coded High	Mean	Std. Dev.
A	Catholyte	mg/L	Numeric	0	0.1	−1 ↔ 0.00	+1 ↔ 0.10	0.0257	0.0425
B	HRT	Hours	Numeric	48	96	−1 ↔ 48.00	+1 ↔ 96.00	66.46	17.4
C	Temperature	°C	Numeric	10	40	−1 ↔ 10.00	+1 ↔ 40.00	26.46	9.13
D	Surface Area	cm2	Numeric	1	3	−1 ↔ 1.00	+1 ↔ 3.00	2.08	0.8623

.

The quadratic empirical model for predicting the response, i.e., $Y$ is derived as a function of the levels of the independent variables expressed according to Equation (1):

$$Y={\beta }_0+\sum _{i=1}^k{\beta }_i{x}_i+\sum _{i=1}^k{\beta }_{ii}{x}_i^2+\sum _{i=1}^{k=1}\sum _{j=i+1}^k{\beta }_{ij}{x}_i{x}_j+\epsilon$$

(12)

$$x_i=\frac{z_i-z_0}{{∆z}_i}$$

(13)

where $x_i$ and $x_j$ are coded independent variables, $i$ denotes the linear coefficient, $j$ is the quadratic coefficient, $\beta$ is the coefficient of regression, $k$ denotes the number of factors studied and optimised by the experiment, $\epsilon$ denotes the random error, $z_i$ and $z_0$ are the code and uncoded values of the $i\mathrm{t}\mathrm{h}$ independent variable, respectively, and ${∆z}_i$ is the step change value between the low level (−1) and high level (+1). It is imperative to note that independent variables may have units and orders of magnitude, and the codification ensures that all independent variables affect the specified responses evenly. The systems’ removal efficiency for the organic parameters precisely COD was calculated based on Equation (3):

$$\% Removal Efficiency \left(Y\right)=\frac{C_{initial}-C_{final}}{C_{initial}}\times 100$$

(14)

where (%) Y is the overall removal efficiency of that specific parameter, and C_initial and C_final, respectively, are the initial and final assays of the parameter in question.

2.4. MATLAB Design and Simulation Models

The experimental scaling-up attempts were carried out in the laboratory using the double chamber microbial fuel cell (MFC) unit made from the 1 Liter Schutt blue cap bottles, as detailed in Section 2.1 above. The emphasis of these experiments is precisely validating the optimum operating conditions that can have the possibility of attaining optimum voltage yield in this DCMFC system. Optimum DCMFC bioelectrochemical yields will then be incorporated into MATLAB-Simscape Electrical for further computation scaling-up simulations. This computer simulation work was performed in view of attaining an optimised DCMFC model that will suffice for commercial application and render reliable energy recovery efficiencies, thereby validating the hypothesis of this study, MFC commercialising and applicability.

MATLAB is a comprehensive software system with multiple empirical and statistical tools coupled with technical computing. It is an engineering-based tool and programming language that can be used to model, simulate, and analyse any physical data. In this study, precise simulations were carried out towards optimisation of the inverter circuit diagram design for scaling up the MFC unit’s overall voltage yield capacity towards commercial scale applications and other potential pragmatic real plant applications. Thus, concise circuit diagram design and drawing were carried out in MATLAB Simulink to model and analyse this bio-electrochemical technology process and achieve breakthrough scaling-up design for practicality and applicability purposes. Rigorous simulations were also carried out to optimise the possible and viable process controller that can integrate and stabilise the MFC system operation, considering first-order kinetics models and either the proportional, derivative, and integral controller or only the proportional and integral controller alone depending on the completeness of the simulation response curve. The following Figure 2, Figure 3 and Figure 4 were adopted as the basis of the inverter circuit design and the integrated controller system to the scaling up MFC design and configuration on both Simulink and Simscape electrical software (MATLAB version 2023b Software). These figures are sample rigorous design process block diagrams and inverter model circuit diagrams, hence further discussed in the results section further down.

3. Results and Discussion

3.1. Validation Experiments Based on RSM Optimised Operating Conditions for DCMFC Bioelectricity Scaling up Performance

Table of Results: Comparison between Three Wastewater Streams for Viable Composite Source

Three different wastewater sources were used to validate the RSM-predicted responses based on optimised DCMFC operating factors. The different sources were part of the experimental optimisation sequence since they all have different substrate strengths, which are imperative for tATP or organic content that essentially gets converted into bioelectrical energy. The results presented in

Table 3. RSM using the BBD method predicted values in the optimisation of the DCMFC unit.

Solution 1 of 92	Predicted	Predicted	Std Dev	n	SE Pred	95% PI Low	Data Mean	95% PI High
CCV	613.7122	613.7122	28.69175	1	69.59522	314.268119	626.96	913.156283
PD	80.15788	80.15788	2.379397	1	12.29463	27.2583705	70.92	133.05739
ID	72.384829	72.38483	1.671414	1	8.636397	35.2254129	61.083	109.544245
COD	35.933484	35.93348	9.807945	1	13.14292	3.7739074	48.97	68.0930597
${PO}_4^{3-}$	103.90157	103.9016	0.192223	1	0.644809	95.7084887	82.45	112.094644
CE	2.5386899	2.53869	0.372049	1	0.534304	1.16521674	1.64	3.91216297
TSS	74.200049	74.20005	9.435737	1	14.25721	37.5507118	69.6	110.849386
TOC	64.451472	64.45147	13.92211	1	22.693	8.92369021	5222	119.979254

Note: CCV = Closed Circuit Voltage; PD = Power Density; ID = Current Density; COD = Chemical Oxygen Demand; ${\mathrm{P}\mathrm{O}}_4^{3-}$ = Total Phosphorus; CE = % Coulombic Efficiency; TSS = Total Suspended Solids; TOC = Total Organic Carbon.

,

Table 4. RSM using the BBD method validation outcome in a DCMFC using a biorefinery wastewater stream as the feed source/anolyte.

HRT (Hours)	CCV (mV)	Anode Electrode (mV)	Cathode Electrode (mV)	Eemf (mV)	ΔGr (J)	CE (%)	Current (mA)	ID (mA)/m2)	Power (mW)	P(anode) (mW/m2)
0	385.5	492	200	692	−6.68 × 107	0.00	0.3855	29.9767	148.61	22.3642
5	375	460	195	655	−6.32 × 107	0.21	0.3750	29.1602	140.63	21.1625
10	324.7	395	192	587	−5.66 × 107	0.18	0.3247	25.2488	105.43	15.8661
15	349.9	380	180	560	−5.40 × 107	0.20	0.3499	27.2084	122.43	18.4244
20	365.4	375	155	530	−5.11 × 107	0.21	0.3654	28.4137	133.52	20.0929
24	390.2	372	165	537	−5.18 × 107	0.22	0.3902	30.3421	152.26	22.9129
30	465.3	369	173.2	542.2	−5.23 × 107	0.25	0.4653	36.1820	216.50	32.5815
35	595	368,8	173.2	542	−5.23 × 107	0.32	0.5950	46.2675	354.03	53.2769
40	637.2	367	172	539	−5.20 × 107	0.34	0.6372	49.5490	406.02	61.1022
48	640.2	364	169.3	533.3	−5.15 × 107	0.34	0.6402	49.7823	409.86	61.6789

,

Table 5. RSM using the BBD method validation outcome in a DCMFC using a mixed wastewater stream as the feed source/anolyte.

HRT (H)	CCV (mV)	Anode Electrode (mV)	Cathode Electrode (mV)	Eemf (mV)	ΔGr (J)	CE (%)	Current (mA)	ID (mA)/m2	Power (mW)	P(anode) (mW/m2)
0	305.1	345	20	365	−3.52 × 107	0.00	0.3051	23.7247	93.09	14.0084
5	319.6	355	22.2	377.2	−3.64 × 107	0.50	0.3196	24.8523	102.14	15.3716
10	330.9	362.3	27.3	389.6	−3.76 × 107	0.52	0.3309	25.7309	109.49	16.4778
15	373.8	361.3	100	461.3	−4.45 × 107	0.58	0.3738	29.0669	139.73	21.0273
20	405	385.4	100.9	486.3	−4.69 × 107	0.63	0.4050	31.4930	164.03	24.6840
24	403.7	405	130	535	−5.16 × 107	0.63	0.4037	31.3919	162.97	24.5258
30	503.4	360	143.8	503.8	−4.86 × 107	0.67	0.5034	39.1446	253.41	38.1357
35	534.8	364.5	105.2	469.7	−4.53 × 107	0.71	0.5348	41.5863	286.01	43.0415
40	566.9	373.5	143.8	517.3	−4.99 × 107	0.75	0.5669	44.0824	321.38	48.3635
48	534.1	363.2	147	510.2	−4.92 × 107	0.71	0.5341	41.5319	285.26	42.9289

and

Table 6. RSM using the BBD method validation outcome in a DCMFC using a dairy wastewater stream as the feed source/anolyte.

HRT (H)	CCV (mV)	Anode Electrode (mV)	Cathode Electrode (mV)	Eemf (mV)	Gr (J)	CE (%)	Current (mA)	ID (mA)/m2	Power (mW)	P(anode) (mW/m2)
0	290.2	351.6	30.6	382.2	−3.69 × 107	0.00	0.2902	22.5661	84.22	12.6736
5	320.8	355	50	405	−3.91 × 107	0.20	0.3208	24.9456	102.91	15.4872
10	337.8	353.3	20	373.3	−3.60 × 107	0.22	0.3378	26.2675	114.11	17.1721
15	345.2	339.5	23	362.5	−3.50 × 107	0.22	0.3452	26.8429	119.16	17.9327
20	339.5	333.8	22.5	356.3	−3.44 × 107	0.22	0.3395	26.3997	115.26	17.3454
24	335.8	326.6	22.4	349	−3.37 × 107	0.21	0.3358	26.1120	112.76	16.9694
30	342.7	328	23	351	−3.39 × 107	0.20	0.3427	26.6485	117.44	17.6739
35	350.7	327.3	40	367.3	−3.54 × 107	0.20	0.3507	27.2706	122.99	18.5087
40	340.2	325.6	35	360.6	−3.48 × 107	0.19	0.3402	26.4541	115.74	17.4170
48	337.7	324.8	26	350.8	−3.38 × 107	0.19	0.3377	26.2597	114.04	17.1620

were attained in conjunction with the predicted runs, as shown in Table 3. The clinical observation is that the most suitable source of wastewater source should experimentally generate the highest CCV-closed circuit voltage yield to show its superiority in terms of bio-electrochemical capacity and its thermodynamic favourability nature to produce high Faraday/coulombic efficiencies.

3.2. RSM MFC Output Predicted Optimum Values Experimental Validation Results

The results attained in the following graphical illustration were a validation of the predicted outputs using the design of experiments in RSM through the BBD model and finetuned by either the 2 FI and Quadratic models to validate empirical models’ significance and statistical validation towards being qualitative optimum operating conditions of the DCMFC unit as it has been numerically optimised. As stated in the previous materials and methods section, these runs were conducted between three sources of complex substrate, streams, or composite wastewater sources. The ideal motive is also to optimise the raw substrate source that will emerge a sustainable, reliable, and optimum voltage-yielding source of bioelectricity to a reliable and optimised power component unit.

3.2.1. Validation Results as Comparison Plots between Voltage vs. Power Density vs. Current Density of Curves in the DCMFC Process

In principle, MFCs are electrochemical devices that can directly transform the chemical energy from organic matter into electrical energy using microbial metabolic activity, so microbes play an essential role [38,50,55,69,70,71]. This study validates the effect of catholyte strength, HRT, Surface Area and Temperature at optimum operating conditions as viable towards attaining optimum bioelectrical capacity in the DCMFC as per predicted outcomes in RSM. Moreover, the effect of organic substrate type towards its effect in increasing the DCMFC voltage yield is in retrofit to derive a principled observation that attests to the electrochemical and bioelectrochemical characteristics of these raw industrial wastewater substrates towards its feasible conductivity conducive for electrical generation. Henceforth, this article validates the organic substrates as economically convenient exoelectrogens that are cost-effective and sustainable alternatives for reliable green electricity production in an MFC.

Figure 5 and Figure 6 clearly attained a precise outcome. Biorefinery organic substrate conductivity and electricity generation capacity are above the rest, reaching a high of 640.2 mV at the optimum operating MFC factors predicted on RSM using the BBD methodology. This experimental validation value presented in Figure 5 and Figure 6 for the overall voltage capacities, especially for the biorefinery wastewater, assures that the DoE was a success and that the MFC unit under the set operating parameters to attain optimum power production was clearly achieved. The mixed wastewater substrate follows closely as this stream possesses both raw wastewater source characteristics with 566 mV, again presented in Figure 5 and Figure 6.

3.2.2. Validation Results for Voltage Yield vs. Power Density and Current Densities in the DCMFC

Based on Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12, dairy wastewater attained a constant and low voltage yield of 350.7 mV: 27.2 mW/m² and 18 mA/m² for power density and current density of the dairy wastewater stream, respectively. Categorically, biorefinery exceeded the predicted overall voltage value of 613.7 mV at optimum operating conditions. The biorefinery stream also attained 62 mW/m² and 50 mA/m² for power density and current density, respectively, presented in Figure 7 and Figure 8. To allude to the effectiveness of these organic industrial wastewater streams, the mixed/blended wastewater stream presented in Figure 9 and Figure 10 also exuded close but lower capacities, which were below the predicted threshold values by RSM and BBD methodology. The power density and current density in this stream grazed at 50 mW/m² and 44.08 mA/m². At this point, using the mixed stream organic substrate, the MFC system recorded the highest Faraday efficiency (€_Cb) of 0.75%. This value was significantly lower than the RSM predicted coulombic efficiency but also presented an organically sustainable source enriched with chemical energy, which correlates with high values of the Faraday efficiency.

From a scientific perspective, the findings presented above validate the principles stated in the literature [18,19,20]. This scientific principle states that raw industrial organic substrates are good exoelectrogenic sources of chemical energy and have a good capacity of tATP and cATP that relates to the magnitude of the organic strength contained. As discussed by Logan et al. [18,19,23], the raw industrial substrates rich in organic bonds, such as glucose starch, etc., are a more viable electron donor and, subsequently, a good source of chemical energy for the MFC bioelectrical process. This observation is statistically and empirically validated by the strong R² mean correlation factors ranging from 0.8 ≤ R² ≤ 1 presented in general by Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12.

3.2.3. Gibbs Free Energy (ΔG_r) (J) Comparison from Three Different Wastewater Substrates for the Generation of Electricity in the DCMFC

Based on the illustration of Figure 13, Figure 14 and Figure 15, a clear and strong correlation significance is observed between Gibbs free energy and the overall cell electromotive force of the MFC unit. The R² values were attained as R² ≥ 1 across all substrate sources, particularly biorefinery, dairy and mixed wastewater composites. This strong correlation suffices the [18,19,20,21,22,23,24] principle that the perfect thermodynamic favourability of the MFC unit will be displayed by the overall magnitude of the MFC in terms of the E^o_emf. Moreover, a concise technical observation based on the above plots shows the sequence of decreasing overall Gibbs free thermodynamic energy with increasing overall MFC unit electromotive force. This attained mechanism correlates with the reports stated by previous authors [12,18,19,20,22,23,24,28,65].

3.2.4. Organic Contaminants Removal from the Three Wastewater Sources after Treatment in a DCMFC within a 48 h Retention Period

The organic removal capacity from industrial wastewater, precisely purely organic wastewater, is mostly a function of increased hydraulic retention periods in a typical microbial fuel cell (MFC), as stated by Logan et al. [18,19,20]. Presented below are the overall % Chemical Oxygen Demand (COD) removal efficiencies, % Total Suspended Solids (TSS) removal efficiencies and the % Phosphate removal efficiencies, respectively, in Figure 16, Figure 17 and Figure 18. COD is related to the overall DCMFC bioelectrochemical performance because it indicates the capacity of the high-strength organic content exuded by that specific wastewater source (Kurchania et al. [1,2,3]). In principle, the higher the COD removal, the higher the bioelectrochemical capacity of MFC should be achieved, as stated by Logan et al. [19,20] and Kurchania et al. [1,2,3]. A sequence of removal efficacies is precisely 62% for the biorefinery wastewater stream, 39% for the mixed wastewater stream and a convincing 94% COD removal efficiency from the dairy wastewater stream. The dairy wastewater stream achieved the expected efficiency, as reported by many researchers in MFC technology [18,19,20,21,22,23,24,25,26,27,28,29,30] and Kurchania et al. [1,2,3]. From a biological perspective, since the DCMFC was inoculated with well-acclimated active bacterial colonies identified as Proteobacteria, these were mostly acclimated with the mixed wastewater stream and, hence, have developed a growth yield sequence in the dairy medium. That can be attested to based on the other attained outputs of the total suspended solids presented in Figure 17: dairy wastewater with 81%, mixed with 61.4% and biorefinery with 60.5%, respectively. This scientifically relates to the active biomass ratio (ABR), which is the ratio of the TSS to the active volatiles contained in the sample or feed wastewater source. When observed from this biological aspect, the dairy wastewater seems a viable biological healthy background for the Proteobacteria used as the inoculum in this DCMFC unit and hence bears good organics or biodegradable removal efficiencies.

The nature of the dairy wastewater stream makes it a phosphate-enriched organic effluent system, as reported by [66,67,68]. This has also been observed from the dairy wastewater sample harvested from fresh from the treatment plant. Based on the characterisation aspects stated above, biorefinery wastewater contains less phosphates but reaches total dissolved solids, as reported by Shabangu et al. [39] and Metcalf and Eddie [21]. The biorefinery streams clearly attained the lowest phosphate result of 31.3% possible because the nature of the microbes and solids contained are less active in the removal of non-organic content or particulate matter, such as total phosphorus. A fair output of about 64% phosphate removal was recorded for the mixed wastewater stream, which has a huge physicochemical characteristic nature from dairy wastewater. Hence, dairy wastewater attained once again moderate total phosphorus removal of 58.1%, which correlates with reported studies by Kim et al. [22,28]. As reported by Logan et al. [20,21,22,23] and other studies, MFCs are deemed more viable in terms of the removal of biodegradable matter, at prolonged retention times. MFCs have mostly recorded low inorganic contaminants removal, such as phosphates, etc. This study recorded moderate findings on phosphate removal, which might be influenced by the active bacterial species inoculated as vital biocatalysts in this MFC unit.

3.3. Simulink Model Design and Simulation towards Scaling up a Double Chamber Microbial Fuel Cell (DCMFC)

According to the experimental data presented in the above sections, the MFC behaviour begins at low voltage and steadily increases until it reaches a steady state, which is caused by microbial activity in the grey water chamber. It lasts for eight hours, and when the microbial life decays or stops multiplying in the anodic chamber, the power generation begins to decline. This will continue until no microbial life is found in the grey anodic chamber. As a result, the voltage will fall to zero. Figure 19 depicts three MFC cell Simulink blocks. The first block is a sequence block that represents the MFC experiment sequence during a ten-hour period. The second block displays the microbial reproduction rate in the blue chamber. It took an hour for bacteria to multiply because of the warmth and oxygen present in the blue chamber. Figure 20 below graphically illustrates the results obtained using MATLAB Simulink. This computational response of voltage yield over a 10 h residence time precisely correlates with the experimental findings and voltage response curves as presented in the above experimental validation section. This sequence of bioelectricity and biochemical mechanism is a function of biomass activity rate as it decreases over time due to the depletion of the organic/chemical energy contained in the complex substrate sample in the form of total adenosine triphosphates, which are chemically converted into dissolved triphosphates and intra-cellular adenosine triphosphates to emulate the quantity of active biomass and living bacterial species at the time. The final phases of the voltage yield simulation curve also predict a very low biomass stress index due to the reduced active bacterial species to chemically and physically biodegrade the complex pollutants to produce bioenergy [11,12,13]. This phase easily aligns with significantly low active biomass ratios, which are a fraction of the active volatile suspended solids on the total suspended solids in the wastewater biodegradable sample [36].

3.3.1. Simscape Scaling up Design Configuration, Modelling and Simulation for Double Chamber Microbial Fuel Cell (DCMFC)

The block diagram-MFC configuration detailed in Section 2, Section 2.4 and Figure 3 depicts the coupling of MFC cells to make a battery for electronic applications. DPFM produces 700 mV and 700 mA per cell, while most applications require Section 2.1 Volts or higher. For Section 2.1 Volts to be feasible, the cell must be connected in a series and parallel configuration. This will increase battery capacity and duration while powering certain devices. The sequence block, the transfer function block and the summation block are used to model three MFC batteries in Figure 21. The results are illustrated in Figure 22, where the voltage peaks above 2 Volts between 30 min and 1.5 h.

3.3.2. Open Loop DC–DC Boost Converter

As illustrated in Figure 22, it was crucial for fragile electrical equipment that may require more than 1 Volt to step up voltage. In MATLAB Simulink, the proposed DC–DC converter in Figure 23 and Figure 24 is designed, modelled and simulated. It is composed of a 2 Volts input voltage source and a switching circuit composed of a transistor and PWM generator, and the energy is stored using an inductor and a capacitor for filtering and stabilising the output. An oscilloscope block is used to display the input and output characteristics. The results presented graphically by Figure 23 clinically reveal that the input voltage was increased from 2 to 12 Volts, showing that this research study is feasible to be utilised as a sustainable energy source to be coupled with reliable electrical scaling-up components that have been perfectly simulated to produce a pragmatic electrical voltage of 12 VAC sufficient to run a few electrical appliances, therefore, satisfying the hypothesis of applicability and practicality of this study’s output.

3.3.3. PID Control for DC–DC Boost Converter

The results of the DC–DC converter simulation reached 12 Volts as expected, although the graph in Figure 25 had some overshoot and ripples that lasted between 1 mS and 5 mS. This necessitated additional investigation into available controllers in MATLAB Simulink. The PID controller was added to the DC–DC converter to reduce voltage ripples and dampen output overshoot. The following model investigates the introduction of PID control on the DC–DC converter shown in Figure 26 and the graphical layout in Figure 27.

After several adjustments to the PID control, the results showed an improvement in the voltage output, as shown in Figure 27. The voltage input was held constant at 2.1 Volts, as indicated by the red curve, and the voltage output is labelled by the blue curve in Figure 27 above. The problem of overshoot was eliminated by using a soft start PWM module available in PID control, and voltage ripples were filtered and reduced to maintain a constant breakthrough voltage capacity of almost 25 Volts beyond 0.01 mS.

3.3.4. Inverter Model

As part of the validation of MFC experiment results for AC applications, the inverter was designed, modelled, and simulated in MATLAB Simulink. Figure 28 depicts a block diagram showing how an MFC battery can be connected to an inverter for devices that demand alternating current (AC). With a variety of inverter model designs available in MATLAB Simulink, the H-bridge was selected as the appropriate controller for the MFC battery. When compared to other electronic switching designs, it has various advantages, such as lower heat losses and dissipation, more efficiency and greater stability. The model in Figure 29 is an inverter design made up of the MFC battery input, PWM generator, H-bridge I GBT transistors, H-bridge filter, step-up transformer, voltage measurements and the scope. The boost converter’s DC voltage is applied across H-bridge I GBT transistors, and a PWM generator is incorporated for switching I GBT transistors. Using a PWM generator, the DC voltage across the I GBT transistors is pulsed at 12 Volts, 50 Hz alternating. This allows the transformer to convert the low voltage to high voltage for AC applications. The input and output voltage simulated results are shown in Figure 30.

4. Comparison with Previously Performed MFC Optimisation Studies

Table 7. Studies on MFC-modelling and optimisation towards enhanced bioelectricity production.

MFC Optimisation Model Attempt Methods
Treatment Process	Optimisation Model	Voltage Yield (VAC)	CE (%)	Power Density (mW/m2)	Current Density (mA/m2)	References
DCMFC	Inverter-based model	220	-	-	-	Current Study
MFC	Freter model	-	0.02	0.106	-	[34]
Stacked MFC	Conduction-based model	-	0.58	-	-	[35]

presents studies reported in the literature on the treatment of IWW using different microbial fuel cell (MFC) optimisation strategies to attain a scale-up production of electrical energy in an MFC unit [18,19,20]. The average % CE, voltage yields, power densities and current densities, as seen within magnitudes, reported in microbial fuel cells by Logan et al. [19,20,21,22,23], also reported by Kim et al. [49] and the rest of MFC research community attest to the fact that one for the predicaments of MFCs is the low % CE compared to other green technology strategies, such as hydrogen cell technology, which exudes more than a 50% energy efficacies, [34,35,36,37,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71]. This aspect can be attributed to the activity of microbial population and thermodynamic favourability in some of the MFCs reported in Table 7. From a scientific perspective, coulombic efficiency is low, describing the efficacy of the charge being successfully harnessed from the substrate source by microbial metabolism to the electrode surface film [53]. It is also reportedly reduced perhaps through alternative electron acceptors other than common electrodes; either the wastewater source or fraction of the diffused oxygen through the Proton Exchange Membrane (PEM) can inhibit a healthy % CE capacity [34,35]. Lin et al.’s [34] model study assumed that the anode attached to the bacterial monolayer served as a biocatalyst for MFC exoelectrogenesis. Through the improvement of the Freter model blended with the Butler–Volmer empirical equation, this model adequately describes the mechanism of power production, substrate utilisation and suspended and attached biomass growth in both batch and continuous operating mode [34]. Lin et al. [34] propose that MFC performances were impacted by various operating variables, such as initial substrate concentration, external resistor, influent substrate concentration and dilution rate and their interactions, which were revealed by data simulation to be complicated [34].

Based on the stackability of MFCs, Nasrabadi et al. [35] simulated a stacked system of microbial fuel cells in power plants. In this study, economic and environmental considerations were considered to generalise such systems as equipment in power plants [35]. Primarily, a conduction-based model was implemented to predict supplied current; then, it was used in energy, exergy, environmental and economic investigations [35]. An established programming code was exploited to calculate the model’s main outputs, and the code was justified according to an experimental study. Furthermore, a parametric analysis was executed to recognise each parameter’s impact on the energy and exergy functions [35]. The findings presented that the produced power from this stacked system is substantial compared to other energy production units, such as solid oxide fuel cells [35]. The energy and exergy efficiencies obtained 35% and 58%, respectively [35]. Hence, this suggests that the application of MFCs in IWW treatment seems to be a viable bio-electrochemical technology that is clean, reliable and renewable in terms of green energy and its capacity to be environmentally viable, socially safe and economically logical from the aspects of raw materials and minimal reliance to carbon enriched sources that emanates into frequent municipal sewer and emissions charges.

5. Conclusions and Future Perspectives

The potential application of an MFC for IWW treatment was investigated by validating the RSM predicted values for the scaling up of bioelectricity production while effectively removing the high-strength organic pollutants: COD, TOC, Turbidity, TSS and Phosphates, etc. The RSM methodology using the BBD design was implemented for experimental work to investigate and validate the interactive effects of the input variables (i.e., HRT, catholyte dose, electrode surface area and Temperature) on the responses or output variables (i.e., voltage yield, coulombic efficiency, power density, current density and Gibbs free energy (E_emf)). It is worth noting that the catholyte concentration (electron-acceptor KMnO₄, at 0.1 N), Temperature (at 27 °C) and Electrode surface area (at 3 C-Cu electrode configuration), were found to impact positively the improved power production in the DCMFC of 640.2 mV using biorefinery raw organic wastewater substrate as the viable source of electron donor and anolyte in the MFC. This observation is further attested to by the optimum values of coulombic efficiency, power density and current density that were validated experimentally as, at this optimum, predicted MFC operating factors. The findings of the current study suggest that the application and practicality of the MFC technology seem to be feasible for IWW treatment prior to being discharged into municipal wastewater receiving bodies whilst generating clean, reliable, renewable, and sustainable bioelectricity. Furthermore, the computation design and modelling of the MFC technology in MATLAB Simulink/Simscape electrical with rigorous simulations have, beyond doubt, validated the possibility of scaling up the sustainable bioelectricity source generated in the DCMFC unit using the perfect raw organic wastewater complex substrates as a viable electron donor. The above simulation and modelling findings validate the MFC unit for practicable and commercial application, considering its scale-up capacity and electrical power potential of up to 220 VAC after a successful integration with a stepping-up electrical components inverter. Precisely, the scaling up and application of an MFC at a commercial platform is proven feasible in this study, concurring with other reported studies. Hence, more MATLAB design and modelling research investigations to further validate the findings of this current study are recommended from a perspective of future ongoing research work having the potential of novel knowledge with future critical applications.