Search

International audience; Nowadays, the proton exchange membrane electrolyzer (PEMEL) is a promising and attractive technology when coupling with renewable energy sources (RES). Indeed, PEMEL can respond quickly to the dynamic operations of RES. Given that PEMEL must be supplied with a low DC voltage, DC-DC converters are mandatory. In this work, a stacked interleaved buck DC-DC converter is used. In this paper, the tuning of proportional, integral and derivative (PID) controllers, which are often used to enhance electrolyzer performances (efficient and reliable operation), is based on the dynamic behavior of the PEMEL. Before designing the PID controller, a model of the PEMEL is proposed based on input-output measured data, and this model is combined with the converter state-space description. Then, the obtained global model is used to tune two PID controller configurations: the first one to control the current and the second one to control the voltage of the PEMEL. Experimental tests are carried out to validate the effectiveness of the designed control laws.

Compared to alkaline electrolyzers, PEM electrolyzers offer high current densities and can be coupled to renewable energy sources because of their fast responses to dynamics. The modeling of PEM electrolyzers is a challenging issue to reproduce its behavior and to design properly power electronics and its control without damaging a real electrolyzer. The input current may have an impact on the dynamics of the electrolyzer and must be taken into consideration to make a model more reliable. In this work, an equivalent electrical circuit to replicate accurately the dynamic behavior of the PEM electrolyzer subject to fast current change is investigated. Based on the input current, the parameters of the model can not be considered as constant. Hence, to improve the accuracy of the model, an adaptive static-dynamic electrical model is proposed and takes into consideration the change of input current. This model is validated by using a commercial-400 W PEM electrolyzer. The obtained results demonstrate the effectiveness of the model to predict the PEM stack voltage.

We would like to acknowledge the cryogenic centre of the IJL institute (Institut Jean Lamour), and more especially Thomas Hauet for allowing us to carry out these experiments in the research centre, and Luc Moreau for his help during experimental tests.; International audience; In practice, the voltage of a hydrogen-oxygen fuel cell is around 1 V at open circuit and from 0.6 V to 0.7 V at full rated load and it can be considered as a low-voltage energy source. Moreover, preliminary investigations undertaken on a single proton exchange membrane fuel cell (PEMFC) highlighted its behavior as a DC current source, that can be directly controlled by the H2 flow rate when the operating point is at very low voltage. In this paper, we present an innovative application of PEMFC that relies on taking advantage of both low voltage level and current source operating mode to feed a high inductive superconducting coil. Such a coil has no resistance and among others, is very sensitive to current ripples. Thus, specific power supplies are designed to feed them but they exhibit in most cases a huge volume and/or a low energy yield. Connecting a superconducting coil to a PEMFC implies to operate in short-circuit, which is an unusual use of PEMFC. To this end, requirements of such an application are defined, by making use of a PEMFC electrical model based on a 1D analog representation of mass transport phenomena. This model, that enables to take into account the influence of gas supply conditions, notably diffusion limit operation, is directly implemented in a standard simulation software used in electrical engineering. Then, simulation results and experimental results obtained by supplying a 10 H superconducting coil cooled by liquid helium by means of a single 100 cm2 PEMFC are compared and discussed.

International audience; This article is focused on the development of an energy management algorithm applied to a multi-stack fuel cell (FC) system for DC microgrid applications. To guarantee the performance of the FC stacks, the current ripple is reduced by employing multiphase interleaved boost converters. A proposed advanced control technique of the multi-stack with multiphase converters for the proton exchange membrane (PEM) FCs is estimated based on a differential flatness approach, in which it can track the power demand in real-time. Furthermore, the differential flatness based-control can ensure the balance of the DC bus voltage of the DC microgrid when load disturbance occurs. The flatness-based energy management strategy is based on both inner current loops (control of the multi-stack PEMFC through their multiphase interleaved boost converters) and outer voltage loop (DC bus voltage regulation). Compared to classic PI controllers mainly based on the linearization of the system to obtain the transfer function (making complex its application), the flatness-based theory leans on time-domain making it easier its use for various applications while ensuring good performances. To validate the proposed control structure, an FC converter system (5 kW) is realized and validated in the laboratory. For hydrogen production, the methanol FC system has consisted of a reformer engine that changes water mixed methanol liquid into hydrogen to supply FC stacks (ME(2)Power Fuel Cell System: 50 V, 5 kW). The proposed control algorithm is tested experimentally by using a dSPACE controller board platform. Simulation and test bench results authenticate the excellent performance during load cycles in DC microgrid.

A buck-based, isolated, high-voltage-ratio DC/DC converter that allows supplying a proton exchange membrane (PEM) electrolyzer from a micro-wind energy conversion system ( μ WECS) has been recently presented. It exhibits low ripple at the switching frequency on the output voltage and current and represents an attractive solution for low-cost hydrogen production. In this paper, a more accurate mathematical model of such a converter is derived and discussed. Then, a model-based robust controller is designed in the frequency domain using the Internal Model Control structure and in the context of H 2 ∕ H ∞ optimal control. The controller satisfies the condition of robust stability and behavior, i.e., it guarantees stability and the desired behavior in the presence of parametric variations and unmodeled dynamics. In particular, the robustness in the presence of variations of DC-link voltage and buck input inductance is verified from the theoretical point of view. The validation of the controller is performed by integrating it into a detailed switching model of the DC/DC converter, which is implemented on a widely used circuit-oriented simulator. Good results are obtained in terms of dynamic and steady-state behavior, even in the presence of the above variations. A comparison is also carried out with the results obtained using an integral controller designed on the basis of the above mathematical model. Such a comparison shows the superiority of the robust controller over the integral controller in all the operating conditions, especially when the DC-link voltage is subject to significant variations and is affected by a non-negligible low-frequency ripple due to the presence of the diode rectifier at the output of the μ WECS.

This work investigates a general comparison of power loss between the neutral point clamped (NPC) and the multi-neutral points (MNP without clamping diodes) inverter models. Our goal in this study was to compare their respective efficiencies. For an NPC, four devices induce power losses: the switching devices (conduction and switching), the anti-parallel diodes, the clamping diodes and the DC link capacitors. For an MNP model, we have the same electrical devices for each phase except the clamping diodes. In this article, we discuss about two series of equations to estimate the power losses in NPC and MNP models. Finally, we present simulation results.

A multi-machine multi-converter system formalism has been proposed to describe systems composed of several electrical machines and converters. This description points out coupling elements, which have to distribute energy. Control structures have already been proposed for systems with downstream coupling. This paper is focused on control structures for systems with upstream coupling. Several solutions can be found by moving control blocks.