Dynamic modeling, monitoring and control of staged microbial fuel cell

Aiming to evaluate the impact of charge storage on MFC performance, the first contribution of this thesis consists in the development of a combined bioelectrochemical--electrical (CBE) model of an MFC. In addition to charge storage, the CBE model is able to describe the multi-scale nonlinear ix dynamics of MFCs by merging mass and electron balances with equations describing an equivalent electrical circuit. Experimental validation and parameter estimation were performed using results of MFC operation with PWM connection of the external resistance. The CBE model shows an acceptable accuracy when describing both fast and slow dynamic behavior observed in the electric voltage produced by the MFC, while also being able to adequately predict the output substrate concentration. Furthermore, the CBE model is used to qualitatively study the effect of duty cycle and switching frequency on MFC performance. Increasing the switching frequency favors the exoelectricigenic over the methanogenic population for higher values of the apparent external resistance. Thus, maximum power presents a plateau ranging from 100 % to lower duty cycles (around 90 %) which corroborates the positive effect of the intermittent operation of the external resistance on the electricity production even during mismatch with the internal resistance value.;The second contribution of this PhD thesis is the development of strategies for MFC effluent quality monitoring from the real time measurement of the electrical variables. A first approach simplifies the CBE model to a single equation that incorporates the electric current as an input into the substrate mass balance. This dynamic approach is valid in the whole range of effluent concentrations and requires the estimation of a single parameter relating the current produced to the rate of substrate consumed. Yet, it requires for the influent substrate concentration to be known. The second approach overcomes such limitation by describing the electric current as a function of the effluent substrate concentration. Kinetics of the CBE model suggest a Monod expression with the addition of a limiting term at low effluent concentrations. In this case, three parameters need to be estimated and therefore, it necessitates information from historical records. Approximations at low effluent concentrations are also suggested that only require the estimation of a single parameter. Unfortunately, because all the strategies are based on open-loop estimations, convergence to the true value is not guaranteed.;Finally, the third contribution of this thesis proposes a centralized control configuration suitable for effluent quality control in two staged MFCs. Reactor staging is a technique widely used in the context of wastewater treatment. It is used to increase carbon source consumption rates and reduce carbon source limitations related to microbial kinetics. The centralized control configuration consists of a PID to control the flow rate and an ON/OFF controller to adjust the connection of the external resistance in the first MFC. A comparison is made with a decentralized control configuration that uses PIDs in cascade for the control of the flow rate. In both cases, the electrical x operation of the staged MFCs is kept independent between the cells. Experimental results show a big overshoot in the manipulated flow rate when using the decentralized cascade PIDs. Meanwhile simulations of the centralized PID-ON/OFF control configuration result in a quicker flow rate control with no overshoot and the ability to cope with disturbances in the influent concentration whilst keeping the effluent concentration within acceptable tolerances. Basically, the ON/OFF controller uses the external resistance as an electrical equivalent to hydraulically bypassing the first cell when substrate deplete conditions occur in the second cell.