Simulation of a small wind fuel cell hybrid energy system

Transcription

1 Renewable Energy 28 (2003) Simulation of a small wind fuel cell hybrid energy system M.T. Iqbal Faculty of Engineering, MUN, St John s, NF, Canada A1B3X5 Received 28 May 2002; accepted 11 June 2002 Abstract This paper describes simulation results of a small 500 W wind fuel cell hybrid energy system. The system consists of a Southwest Wind Power Inc. AIR 403 wind turbine, a Proton Exchange Membrane Fuel Cell (PEMFC) and an electrolyzer. Dynamic modeling of various components of this small isolated system is presented. Simulink is used for the dynamic simulation of this nonlinear 48 V hybrid energy system. Transient responses of the system to a step change in the load current and wind speed in a number of possible situations are presented. Analysis of simulation results and limitations of a wind fuel cell hybrid energy system are discussed Elsevier Science Ltd. All rights reserved. Keywords: Hybrid energy systems; Wind energy; Fuel cell; Dynamics of energy system; Simulation and control; Distributed power generation 1. Introduction After many technological advances Proton Exchange Membrane Fuel Cells (PEMFCs) technology has now reached test and demonstration phase [1]. The recent commercial availability of small PEMFC units has created many new opportunities to design hybrid energy systems for remote applications. For example Hpower.com stationary power unit EPAC500 is a 500 W hydrogen fueled PEMFC for backup power in homes. Other examples of commercially available 500 W PEMFC includes Tel.: address: tariq@engr.mun.ca (M.T. Iqbal) /03/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S (02)

2 512 M.T. Iqbal / Renewable Energy 28 (2003) SR-12 from Avistalabs.com and HyPORT power generator from Hydrogenics.com. Such commercially available fuel cell units can be combined with other renewable energy sources such as wind energy. Production of hydrogen using energy from a wind turbine for later use in a fuel cell is being studied at the Hydrogen Research Institute [2]. A detailed study [3] done for the Department of Energy indicates that a wind fuel cell based integrated system will produce power at a much lower cost and will contribute a much lower amount of CO 2 to the environment. An experimental study of daily variations in the bus bar voltage of a hybrid energy system consisting of solar panels, wind turbine, fuel cells and batteries has been recently completed at the Desert Research Institute [4]. Many aspects of such a hybrid energy system need to be investigated e.g. cost, efficiency, reliability, and dynamic response of the electrolyzer, fuel cell and hydrogen storage components. One important aspect of a wind fuel cell hybrid energy system that needs further investigation is design and simulation of the control system. A block diagram of the hybrid energy system is shown in Fig. 1. The load can be supplied from the wind turbine and/or fuel cell. If the wind turbine is producing enough power, the load will be supplied entirely from wind energy. In case of low wind a share of power can be supplied from the fuel cell. If the output power from the wind turbine exceeds the demand, the excess power may be used to produce hydrogen for later use in the fuel cell. Most of the time, in this system, the fuel cell stack is to be operated in the variable current output mode. While in most applications fuel cells are operated at a constant current. A wind fuel cell hybrid energy system may be based on a reversible fuel cell system [5,6]. Reversible fuel cells are still not commercially available. Therefore, in this study a more practical fuel cell and an electrolyzer-based system is considered. The system description, modeling, Simulink based simulation and an analysis of system dynamics are presented below. 2. Description of the system The system consists of a Southwest Wind Power Inc. AIR 403 wind turbine, a Proton Exchange Membrane Fuel Cell (PEMFC) such as Hpower.com EPAC500, an electrolyzer ( a wind mast, a dump load, a personal computer Fig. 1. Wind fuel cell hybrid energy system.

3 M.T. Iqbal / Renewable Energy 28 (2003) acting as controller and data acquisition system. Fig. 2 shows the details of system interconnections. For the study and analysis of the system the following parameters of this 48 V hybrid energy system are recorded: (a) Wind speed; (b) wind turbine current; (c) fuel cell voltage; (d) fuel cell current; (e) fuel cell temperature; (f) fuel cell pressure; (g) fuel flow rate; (h) wind direction; and (j) load current. The inbuilt wind turbine controller runs the turbine in variable speed mode while extracting maximum power. Fuel cell system consists of a PEM fuel cell stack and an electrolyzer. A fuel cell stack consists of 65 individual fuel cells connected in series. The output current can vary between 0 and 25 A. Fuel cell delivers the current difference between the load current and the wind turbine current. If the output voltage of the fuel cell stack drops below 46 V its controller switches on. A PC based PID type fuel cell controller adjusts the fuel and oxygen flow rates to maintain a constant stack output voltage. Controller action compensates the drop in the fuel cell stack voltage caused by the load current variations. If the wind turbine generates more current than required by the load then the excess current is diverted towards an electrolyzer. The electrolyzer-produced hydrogen is stored in a tank for later use in the fuel cell stack. 3. System model The main components of the system are wind turbine, fuel cell stack, electrolyzer and controller. A standard classical method of representing the system by a set of differential equations and PID controller by a transfer function is used [7]. Wind turbine rotor diameter is 1.14 m. This self-regulating permanent magnet alternator based variable speed wind turbine produces 400 W at a wind speed of 12.5 m/s. Self-regulation (stall controlled) is achieved by the twisting of the blades. A wind Fig. 2. Proposed wind fuel cell hybrid energy system.

4 514 M.T. Iqbal / Renewable Energy 28 (2003) turbine power curve is shown in Fig. 3. This small wind turbine is capable of extracting maximum power until a wind speed of 40 mph. Above a wind speed of 40 mph (17.9 m/s) the wind turbine quickly enters stall mode by blades twisting avoiding any over speed. This small wind turbine is well suited for roof top installation. Power curve of this wind turbine is nonlinear. It is digitized and the resulting table is used for simulation. Dynamics of the wind turbine are added by considering the wind turbine response as a second order slightly under damped system [7]. This is true when we take a first order moment of inertia (J) and friction based dynamic model for the wind turbine rotor and a first order model for the permanent magnet generator. Using this simple approach wind turbine dynamics are modelled as [7]. Y(s)/X(s) 1/(s 2 s 1) (1) Where input is power obtained from the power curve for a known wind speed and output is actual power of the wind turbine. Fuel cells are electrochemical devices that convert the chemical energy of a reaction directly into electrical energy. The basic building block of a fuel cell consists of an electrolyte layer in contact with a porous anode and cathode on either side. A schematic representation of a fuel cell with the reactant/product gases and the ion conduction flow directions through the cell is shown in Fig. 4 [8]. In a typical fuel cell, gaseous fuels are fed continuously to the anode (negative electrode) compartment and an oxidant (i.e. oxygen from air) is fed continuously to the cathode (positive electrode) compartment; the electrochemical reactions take place at the electrodes to produce an electric current. Fuel cells are classified by the type of electrolyte used in the cells and includes: (1) proton exchange membrane (polymer) electrolyte fuel cell (PEMPC); (2) alkaline fuel cell (AFC); (3) phosphoric acid fuel cell (PAFC); (4) molten carbonate fuel cell (MCPC); and (5) solid oxide fuel cell (SOFC). These fuel cells are listed in the order of approximate operating temperature, ranging from 80 C for PEMFC to 1000 C for SOFC. Fuel cells can operate on natural gas/propane using a reformer. PEMFC running on hydrogen for stationary and portable applications are commercially available from a number of sources [1]. A number Fig. 3. Wind turbine power curve.

5 M.T. Iqbal / Renewable Energy 28 (2003) Fig. 4. Schematic of individual fuel cell. of approaches have been used to model a PEMFC behavior [9]. Parametric model of PEMFC developed by Amphlett [10,11] using mechanistic approach and a number of group parameters is used. The fuel cell voltage V, isdefined as the sum of three terms: the thermodynamic potential, the activation over voltage and ohmic over voltage. The ideal standard potential of an H 2 /O 2 fuel cell (E 0 ) is V with liquid water product. The actual cell potential is decreased from its equilibrium potential because of irreversible losses. Several sources contribute to irreversible losses in a practical fuel cell. The losses, which are often called polarization over potential or over voltage, originate primarily from three sources: (1) activation polarization; (2) ohmic polarization; and (3) concentration polarization. These losses result in a cell voltage, V for a fuel cell that is less than its ideal potential, E (V=E Losses). The overall reaction in a fuel cell can be described as H O 2 H 2 O (2) The thermodynamic potential E is defined via a Nernst equation in expanded form as [11] E (T ) T.(lnP H2 (3) 0.5 lnp O2 ) where P is the effective pressure in atm and T is temperature in Kelvin. The concentration of dissolved oxygen at the gas/liquid interface can be defined by Henry s Law expression of the form [10,11] c O2 P O2 /( exp( 498/T) (4)

6 516 M.T. Iqbal / Renewable Energy 28 (2003) The parametric equation for the over voltage due to activation and internal resistance developed from the empirical analysis [11] are given as h act T T[ln(i) T ln(c O2 )] (5) R int T i (6) where i is the fuel cell current and the activation resistance is determined as R a h act /i (7) The combined effect of thermodynamics, mass transport, kinetics, and ohmic resistance determines the output voltage of the cell as defined by [10,11] V E v act h ohmic (8) The model described above indicates the current drawn, cell temperature, H 2 pressure and O 2 pressure will affect the fuel cell voltage. A drop in fuel cell voltage can be compensated by an increase in fuel pressure. Dynamics of fuel cell voltage can be modelled by an addition of a capacitor C, to the steady state model [12]. The effect of double charge layer is modelled by a capacitor C connected in parallel with activation resistance as shown in the Fig. 5. A differential equation describes the voltage of a fuel cell as dv act /dt i/c v act /R a /C (9) The ohmic voltage loss in the fuel cell is given by h ohmic ir int (10) The fuel cell system consists of a stack of 65 similar cells connected in series. Therefore the total stack voltage is given by V stack 65 V (11) Amount of hydrogen and oxygen consumed in the fuel cell depends upon the flow rates and the current drawn out of the fuel cell. It also depends upon the volume of electrodes. If incoming and outgoing flow rates (m ) are known then pressure can be determined using the mole conservation principle. For the fuel cell anode we can write V a dp H2 m RT dt H 2 in (r H2 UA) out i (12) 2F where V a is anode volume in litres, R is universal gas constant ( L- Fig. 5. Model of a fuel cell.

7 M.T. Iqbal / Renewable Energy 28 (2003) atm/(mol.k)), T is fuel cell temperature (K), ρ is mole density, U is fuel velocity, A is channel flow area and F is Faraday constant (96,500 C). Similarly, an equation for cathode is V c dp O2 m O RT dt 2 in (r O2 UA) out i (13) 4F where V c is the cathode volume in litres. Eqs. (3) (13) describe the fuel cell dynamic behaviour. A bank of switched resistors is used as a load. If the wind turbine produces more current than required by the load then excess current is diverted towards an electrolyzer. Electrolyzer converts abundant chemicals into more valuable ones by the passage of electricity, normally by breaking down compounds into elements or simpler products. Electrolysis of liquid water [H 2 O] into hydrogen gas [H 2 ] and oxygen gas [O 2 ] is the classic example of electrochemistry. Hydrogen generated by the electrolyzer is estimated using the equation [13]. V h2 5.18e 6.i electrolyzer (moles/s) (13) A PID type controller is used to control the fuel cell voltage by varying the H 2 and O 2 flow rates. Controller is activated when stack voltage drops below 46 V. A general transfer function of a PID may be written as [7] G r (s) K p (s T d s 2 1/T i )/s (14) The Ziegler Nichols methods for determining the parameters of a PID controller are used [7]. Suitable controller parameters are K p =1.0, T i =2, T d = Simulation results Simulink is an interactive tool for simulating, and analyzing dynamic systems. It enables one to build graphical block diagrams, evaluate system performance, and refine designs [14]. Simulink is the tool of choice for control system design and other simulation. The Simulink simulation engine offers numerous features for simulating large, challenging systems. The mathematical model of the system described above is simulated in Simulink version 4.1. Fig. 6 shows the Simulink simulation module. Basically it consists of four sections i.e. wind turbine, fuel cell system, controller and electrolyzer. The temperature of fuel cell is assumed to be constant at 80 C. A step input in the wind speed or demand current can be applied in the simulation. Controller is activated when fuel cell voltage drops below 46 V. Current value of wind power, fuel cell current, fuel cell voltage and hydrogen produced by the electrolyzer is displayed on the digital meters. An oscilloscope is used to record the system transients. Simulation results are shown in Figs. 7, 8, 9 and 10. Fig. 7 shows the response of the system to a step change in the load current from 10 to 15 A at t=5 s. Wind speed is assumed constant at 12 m/s i.e wind turbine is producing 346 W. In this situation load current is supplied by the wind turbine as well

8 518 M.T. Iqbal / Renewable Energy 28 (2003) Fig. 6. Simulation of wind fuel cell energy system in Simulink. as fuel cell. A step increase in the load current by 5 A results in more contribution of fuel cell. Expected transient in the fuel cell voltage and wind turbine current are shown in the Fig. 7. Fuel cell voltage drops in one second as its current contribution increases. During this transient wind turbine current first increase but later it settles down to a constant value. Fig. 8 shows the transient response of the system when wind speed increases from 12 to 16 m/s. During this transient load current is assumed constant at 15 A. As wind turbine starts producing more power its current contribution to the load increases. This results in a higher fuel cell voltage as shown in the Fig. 8. Fuel cell voltage increases from 52 to 59 V within 4 s. There is a small overshoot indicating that system is slightly under damped. During this transient fuel cell current drops 8.34 A to 2.43 A. Fig. 9 shows response of the system when there is a step increase in the load current while wind speed is only 5 m/s. In this situation most of the power is supplied by the fuel cell as wind turbine produces only 51 W. As fuel cell currents jumps from to A its voltage drops below 46 V limit. Fuel cell controller acts and allows more fuel flow to the fuel cell. Fuel cell pressure increases from 2 to 5.6 atm. An increase in the fuel cell pressure, as shown in Fig. 9, results in an

9 M.T. Iqbal / Renewable Energy 28 (2003) Fig. 7. Response of the system to a step change in the load current. Fig. 8. Response of the system to a step change in the wind speed. increase in its voltage back to 46 V. For this small hybrid system expected transient duration is only 1 s. Fig. 10 shows transient response of the system when load current suddenly drops while wind speed is 10 m/s. Change in load current is from 10 to 5Aatt=5 s. In this situation load is partially supplied by the fuel cell. As current demand drops, the fuel cell current decreases from 5.88 to 1.34 A. This drop in fuel cell current leads to two changes in its voltage. A sudden increase in voltage due to a drop in ohmic losses and a gradual increase in its voltage due to fuel cell capacitance. Total increase in the fuel cell voltage is from 55 to V. As shown in Fig. 10 response of the system in this situation is over damped.

10 520 M.T. Iqbal / Renewable Energy 28 (2003) Fig. 9. Response of the system to a step change in load while wind speed is 5 m/s. Fig. 10. Response of the system to a step change in the load while wind speed is 10 m/s. Simulation results presented above indicate the transient expected in a small wind fuel cell hybrid energy system. This information is vital for designing the system and selecting power electronics for such a system. Simulation predicts expected variations in the system voltages and currents. Duration of the transients in different possible situations is estimated. Design of controller is also finalized in the simulation before actual testing of such a system. Simulink based simulation is found to be a very useful method for design and analysis of the wind fuel cell hybrid system. The

11 M.T. Iqbal / Renewable Energy 28 (2003) system described above is under development at Memorial University of Newfoundland and experimental test results will be reported in later publications. 5. Conclusions Hybrid energy systems are best suited for isolated communities. A small 500 W wind fuel cell hybrid energy system is proposed in this paper. Design and analysis of this demonstration type ultra low emission energy system is presented. System modeling, simulation and design of controller are presented. Hybrid energy system is simulated using Simulink to determine its controllability and expected transients. Simulation results indicate the expected transients in the system in a number of possible situations. After a step change in the wind speed or load current expected voltage variations in this 48 V system are found to range between 43 and 65 V. Transient s duration is between 1 and 5 s. In most of the transient situations system behaves like an over damped system. Simulink is also found to be a very useful tool for analysis and design of such a nonlinear hybrid energy system. References [1] Costamagna P, Srinivasan S. Quantum jump in the PEMFC science and technology from the 1960s to the year 2000 Part II. Engineering, technology development and application aspects. Journal of Power Sources 2001;102: [2] Agbossuo K, Chahine R, Hamelin J, Laurencelle F, Hamelin J. Renewable energy systems based on hydrogen for remote applications. Journal of Power Sources 2001;96: [3] Ruhl RC. Fuel cell and reversible fuel cell modules for grid independent electric power systems, Final report DOE No. DE-FC36-99G , September 30,2000. [4] Jacobson R, Purcell R, Wermers D, Wood B, Donovan D, Lane D, Matheous M. Renewable hydrogen system integration and performance modeling, Proceedings of the DOE Hydrogen program Review, 2001 (NREL/CP ). [5] Mitlitsky F, Myers B, Weisberg AH, Molter TM, Smith WF. Reversible (Unitized) PEM Fuel Cell Devices. Portable Fuel Cells Conference. Lucerne, Switzerland. June 21-24, [6] Riezenman MJ. Metal Fuel Cells, IEEE Spectrum, pp , June [7] Kuo B.C., Automatic Control Systems. 7th ed. Englewood Cliffs, NJ: Prentice Hall; [8] EG&G Services Inc. Fuel Cell Handbook. 5th ed. (DE-AM26-99FT40575) US Department of Energy. October [9] Rowe A, Li X. Mathematical modeling of proton exchange membrane fuel cells. Journal of Power Sources 2001;102: [10] Mann RF, Amphlett JC, Hooper M, Jensen HM, Peppley BA, Roberge PR. Development and application of a generalized steady state electrochemical model of a PEM fuel cell. Journal of Power Sources 2000;86: [11] Amphlett JC, Baumert RM, Mann RF, Peppley BA, Roberge PR, Harries TJ. Performance modelling of the Ballard mark-iv solid polymer electrolyte fuel cell. Journal of Electrochemical Society 1995;1421:9 15. [12] Larminie J., Dicks A., Fuel cell systems explained. London: Wiley; [13] Sapru K, Stetson NT, Ovshinsky SR. Development of a small scale hydrogen production storage system for hydrogen applications. Proceedings of the Intersociety Energy Conversion Engineering Conference

12 522 M.T. Iqbal / Renewable Energy 28 (2003) [14] Simulink is a product of The MathWorks, 3 Apple Hill Drive, Natick, MA , USA. (