2005 Progress Report: "Collaborative Proposal" On Optimal-Control Strategies Based On Comprehensive Modeling And System-Interaction Analyses For Energy-Efficient And Reduced-Emission Fuel-Cell-Energy-Conversion Technologies For Hybrid Stationary And Non-Stationary Applications
EPA Grant Number: R831514Title: "Collaborative Proposal" On Optimal-Control Strategies Based On Comprehensive Modeling And System-Interaction Analyses For Energy-Efficient And Reduced-Emission Fuel-Cell-Energy-Conversion Technologies For Hybrid Stationary And Non-Stationary Applications
Investigators: von Spakovsky, Michael R. , Herbison, Dan , Leo, Donald J. , Mazumder, Sudip K. , McIntyre, Chuck
Institution:
EPA Project Officer: Bauer, Diana
Project Period: June 1, 2004 through May 31, 2007 (Extended to December 31, 2007)
Project Period Covered by this Report: June 1, 2004 through May 31, 2005
Project Amount: $200,000
RFA: Technology for a Sustainable Environment (2003)
Research Category: Pollution Prevention/Sustainable Development
Description:
Objective:Fuel cell systems (FCSs) are seen as more environmentally friendly and energy efficient than their primary fossil-fuel counterparts and have a high potential for playing a significant role in the energy economy. Proton Exchange Membrane Fuel Cell (PEMFC) systems are one of the primary fuel-cell candidates for transportation and distributed-power applications, whereas PEMFC based total-energy systems (TESs) are being considered for distributed cogeneration. Such systems typically consist of a fuel reformer (which in a transportation application may be on- or off-board), a fuel-cell stack with appropriate water-, air-, and thermal-management subsystems, power-electronics and power-conditioning subsystems, and some type of energy buffering both for electricity, fuel, and air.
Most conventional fossil fuel based energy-conversion systems, such as the internal combustion engine (ICE), may exhibit relatively high efficiencies at full load, but their efficiencies drop off precipitously at part load. In contrast, FCSs exhibit relatively high efficiencies both at full and partial loads, which lead to a consequent positive environmental impact both from a fuel and emissions standpoint. The significant energy gain leads to a considerable reduction in CO2 emissions. Furthermore, emissions of SO2 are effectively nonexistent as the reforming of a fossil fuel to produce the H2 used by the fuel-cell stack requires that the fuel be desulphurized to less than 200 ppbm of H2S prior to reforming. Even 1-2 ppm of H2S can deactivate the electrochemical catalyst used in a PEMFC, drastically reducing performance (Paul and Frank, 2005). In addition, because the reforming process occurs in a temperature range of 450°C to 850°C (depending on the fuel), the production of NOx is insignificant as most NOx formation occurs at temperatures above 1,000°C. Finally, CO emissions also are insignificant because sufficient cleanup is provided to ensure that less than 10 ppm of CO is ever present in the PEMFC stack, even during transient operation. Even small amounts of CO can poison the catalytic electrochemical process within the fuel cell itself.
The gains both in terms of energy savings and pollutant emissions greatly depend on whether or not the FCS is optimally synthesized and designed and whether or not appropriate control strategies have been developed to meet the varying load profiles for a variety of applications. To facilitate this process, detailed and very realistic steady-state and transient models are required that reproduce accurately the electrical, electrochemical, thermodynamic, kinetic, chemical, and geometric characteristics of the FCS and its subsystems: the power electronics subsystem (PES), the PEMFC stack subsystem (SS), and the balance-of-plant subsystem (BOPS). Such models once developed are used to design optimal control architectures and strategies, which alleviate the feedbacks from the PES and application loads (stationary/nonstationary) that negatively affect the reliability and performance of the PEMFC system. Furthermore, optimally and simultaneously synthesizing and designing the system as well as its control architecture leads to higher system efficiencies and lower emissions across the entire load spectrum (Rancruel, 2005) and does so in part by managing each subsystem’s response to effectively deal with load-fluctuations and load-following issues. It also does so by ensuring that the optimal component and subsystem characteristics match those of the optimal control architecture.
The goal of this research project is to develop an enabling design for the environment tool for use in the development and control of PEMFC technologies. The specific objectives of this research project are to: (1) develop a fully transient, nonlinear, general model of PEMFC stacks and a variety of BOPSs; (2) develop realistic load profile scenarios for various types of stationary and nonstationary applications for uses in generating optimal control strategies for PEMFC systems; and (3) implement bottomup and topdown approaches for the development of optimal-control strategies.
Progress Summary:Figure 1 shows the proposed PEMFC system configuration, which consists of fuel processing subsystem (FPS), SS, and work recovery and air supply subsystem (WRAS). This was developed taking into account all the equipment and recovery loops necessary for maximizing total system efficiency.
The main objective of the FPS is to convert the natural gas consisting primarily of methane to the hydrogen-rich reformat gas that will provide the hydrogen fuel required for the operation of the PEMFC stack. Organic sulphur contained in natural gas is removed by the desulphuriser. The preferential oxidation (PrOx) reactor reduces the amount of CO in the mixture to less than 10 ppm.
The SS consists of a PEMFC stack, a deionizer subsystem, and a cooling cycle that controls the operating temperature of the stack. In the BOPS and SS, the temperatures of a number of critical components have to be controlled carefully, and the flow and utilization of energy from several sources within the configuration via heat interactions have to be managed effectively to achieve high overall efficiency. Therefore, thermal management plays a significant role in the operation of the PEMFC system. Its major functions include maintaining the stack operating temperature in the appropriate range, bringing the hydrogen-rich reformed gas and compressed air to the desired anode/cathode inlet conditions before exiting the FPS and the WRAS, respectively, and controlling the steam-methane reformer’s as well as the shift and PrOx reactors’ operating conditions and the generation of the steam required for the FPS. A number of high performance heat exchangers are used within the configuration to meet these objectives.
Figure 1. Proposed Nonoptimized PEMFC System Configuration for Maximum Efficiency (Virginia Tech).
Modeling (Virginia Tech and UIC)
Fuel Processing Subsystem (Virginia Tech). The FPS includes a combustor, a steam generator, a desulphuizer, a steam-methane reformer (SR), a high and a low temperature water gas shift reactor (HSR, LSR), and a PrOx reactor. The mathematical model of the FPS consists of a set of transient governing equations (conservation of mass and energy) and closure relations that describe all component characteristics (geometries, both part- and full-load performance, chemical kinetics, control strategies, etc.) as well as the mass and associated energy flows in each of the lines of the subsystem.
The steam generator model developed and implemented consists of an economizer, an evaporator, and a superheater. These three integrated component parts have been modeled as a crossflow, shell-and-tube heat exchanger with a single-pass shell and one tube pass. The heat transfer characteristics of the steam generator involve three different heat transfer models because of the fact that different convection heat transfer coefficients, as well as methods, are applicable to the design/analysis of each of the three parts.
A steam-methane reformer is modeled based on the chemical kinetic model equations of Xu and Froment (1989a,b). To more accurately model the characteristics of a commercial reformer in which large catalyst pellets are used, the reformer’s intrinsic reaction kinetics are combined with diffusional limitation terms.
High- and low-temperature water-gas shift reactors are modeled by using the reaction rates of the water gas-shift reaction studied by Ettouney, Shaban, and Nayfeh (1995). Their theoretical analysis of the catalytic water-gas shift reaction of CO has been validated with data collected from industry.
Removing CO before the hydrogen-rich gases enter the fuel cell is of critical importance. A PEMFC system requires an extremely low concentration of CO of under 10 PPM, even under transient conditions. The kinetics of the CO preferential oxidation reaction have been modeled and implemented based on Choi and Stenger (2004).
Table 1 describes an example calculation based on the models developed and already implemented of a steam-methane mixture converted to a hydrogen-rich reformate gas in the train of reactors depicted in Figure 1. This FPS simulation shows that 55 percent-by-volume of hydrogen-rich gas is generated with 99.8 percent of the methane converted (for this example, the natural gas is assumed to consist only of methane), and the concentration of CO is controlled to less than 10 ppm by means of two water-gas shift reactors and a PrOx reactor.
Table 1. Mole Fraction of Components at Each Reactor in the FPS (Virginia Tech)
Chemical species |
SR |
HSR |
LSR |
PrOx |
|
Operating conditions |
|||||
~ 800°C & 3 bar (S/C = 3.6) |
~ 500°C & 3 bar |
~ 250°C & 3 bar |
~ 110°C & 3 bar |
||
Mole fractions |
|||||
Inlet |
Outlet |
Outlet |
Outlet |
Outlet |
|
CH4 |
0.33 |
0.0019 |
0.00190 |
0.00190 |
0.00181 |
H2O |
0.67 |
0.3355 |
0.31518 |
0.30915 |
0.29047 |
CO |
0.00 |
0.0885 |
0.04343 |
0.00767 |
8 ppm |
CO2 |
0.00 |
0.0617 |
0.09591 |
0.13193 |
0.13376 |
H2 |
0.00 |
0.5124 |
0.54358 |
0.54935 |
0.54890 |
O2 |
5 % of O2 fed before PrOx |
0.02506 |
|||
SOx |
Under 200 ppm exiting the desulphuriser |
||||
NOx |
Essentially free from NOx generation (NOx = parts/10-15, at equilibrium) |
Stack Subsystem (Virginia Tech). The SS consists of a PEMFC stack and a cooling loop with a deionizer. Several mathematical models of PEMFC stacks appear in the literature. The majority of them, however, only are able to simulate a stack’s steady state behavior. For this work, a semiempirical, one-dimensional approach based on Ceralo, Miulli, and Pozio (2002) has been chosen to model both the steady state and transient behavior of the PEMFC stack. This model considers a cell that utilizes H2 as the fuel and air as the oxidant, both humidified. The temperature of the fuel cell stack is assumed to be uniform (a good assumption as the temperature variation across the stack is relatively small) and is used to determine the flow rate of the stack cooling water used to maintain this stack temperature. The water vapor contained in the reactant mixtures in the pores of the cathode-side gas diffusion and electrode-catalysts layers is assumed to be in equilibrium with the surrounding liquid phase so that, consequently, the partial water pressure is uniform throughout these layers. Furthermore, the membrane electrolyte is assumed to be saturated completely with water so that its conductivity is only a function of temperature.
In Figure 2, a comparison of simulated and experimental cell voltage and cell power density for different current densities is reported. Figure 3 shows a comparison of simulated and experimental cell voltage dynamic response to a sudden change of cell current density. These comparison graphs show that the model simulation corresponds well with experimental data.
Figure 2. Comparison of Simulated and Experimental Cell Voltage and Cell Power Density at a Cell Temperature of 70°C and an Inlet Air Flow Rate of 1,500 scc/minute (Virginia Tech)
Figure 3. Comparison of Simulated and Experimental Cell Dynamic Response at a Cell Temperature of 70°C and Air and Hydrogen Inlet Pressures at 3 and 2 Bar With Flow Rates of 1,500 and 1,400 scc/minute, Respectively (Virginia Tech)
Work Recovery and Air Supply Subsystem (Virginia Tech). WRAS plays a significant role in the thermal management of the whole system. As shown in Figure 1, it consists of a compressor, an expander, a motor, and a transient air tank. The compressor provides compressed air to the fuel cell stack and other thermal management systems and is driven by an expander and motor. To run the expander, energy is recovered from the combustion gases exiting the steam generator at between 55 and 75°C to offset some of the parasitic power required by the compressor. The compressed air exiting the compressor then is stored in the air tank, which provides air to the stack and the combustor. An electric motor is used to supply additional power to the compressor in case the power extracted from the expander is not enough to run the compressor.
For the WRAS model, a blushless DC (BLDC) motor and an integrated twin screw compressor-expander unit are modeled. Some advantages of BLDC motors are their compact size, controllability, and high efficiency. In addition, screw compressors supply oil-free compressed air, are compact, and can compress highly at low flow rate. A twin screw compressor can obtain a high efficiency and charging pressure in a small size by means of combining two screws in one unit.
Figure 4 depicts the predicted work required by the compressor and extracted work from the expander at different air charging rates. At 0.7 mol/second of air exiting the compressor, 36 percent of the power load is recovered from the expander and the recovered work to load ratio increases with increases in air charging rate. These numbers compare well with the work recovery ratios found for these types of compressor-expander units in the commercial sector.
Figure 4. Required Work to the Compressor and Extracted Work From the Expander at Different Air Charging Rates (Virginia Tech).
Switching Power Electronics Model (UIC). The PEMFC stack voltage varies with current drawn by the load, decreasing significantly at higher load current. Therefore, a PES is used to process the raw output power from the stack and provide power to the load at constant DC or AC voltage. Therefore, the first stage of the PES is chosen essentially as a DC-DC converter, which would provide constant DC output voltage irrespective of the input stack voltage. To provide the AC loads of a residential power system, this intermediate DC voltage is to be converted to sinusoidal AC (50 or 60 Hz) using DC-AC converter (inverter).
The topological model of PES for a residential power system is shown in Figure 5. The model consists of two DC-DC boost converters to step-up the PEMFC output voltage and the battery output as well to a higher intermediate DC bus voltage. The DC-DC boost converter for the stack has to be unidirectional (because current should not flow into the fuel cell) while regulating the output bus voltage. We chose a simple DC-DC boost converter, as shown in Figure 5, to meet these requirements. For the battery, however, a bidirectional boost DC-DC converter is needed to provide additional required current to the load during a transient and to recharge the battery. A DC-AC converter (inverter) is further used to convert the DC bus voltage, the output of the DC-DC boost converter, to feed the AC load. For this purpose, a pulse-width-modulated voltage source inverter (VSI) is used because of its simpler control scheme. The high-frequency harmonic content of the output of the VSI is eliminated by the output filter.
Figure 5. An Architecture of the PES for the Residential PCS (UIC)
The operation and control of DC-DC boost converter is defined and explained in detail elsewhere (Erickson and Maksimovic, 2004). When the switch S of the unidirectional boost converter is turned ON, the energy content in the inductor L increases, and when the switch S is turned OFF, the stored energy in the inductor is transferred to the output capacitor C of the converter through the diode D. The output capacitor C eliminates the high-frequency switching ripple of the DC bus voltage and provides required output current. The turn ON time of the switch S is calculated to obtain the required bus voltage. The operation of the bidirectional converter is exactly similar to unidirectional converter except that the switches S2, which replaces the diode, turns ON and OFF complementarily with switch S1.
The operation and control of full-bridge VSI is defined and explained in detail elsewhere (Lee, 1990). The switch pairs SW1-SW3 and SW2-SW4 switch in complement with a very small time delay between the switching of SW1 (SW3) and SW2 (SW4). A sinusoidally-modulated switching sequence is generated for SW1 (SW3) and SW2 (SW4) to obtain an averaged sine wave AC at the output of the VSI. The L-C filter, consisting of Lf and Cf eliminates the high frequency ripple from the VSI output.
Average Model (UIC). Because the PES switching model comprises discontinuous-differential equations, a stiff solver and fast sampling is required for convergence and numerical stability. For example, to solve the switching model of a PES converter operating at 20 kHz (i.e., a switching period of 50 µsec), sampling time as low as hundreds of nanoseconds may be required. Comparing that to the vastly different (typical) sampling times of PEMFC and BOPS models (which are around few milliseconds and hundreds of milliseconds, respectively), we realize that the PES switching model is one of the key hindrances to the fast computation of PEMFC PCS model. To increase the speed of the PCS simulation, we need to avoid discontinuity, thereby reducing the sampling rate and bring it as close as possible to that required by the PSOFC and BOPS models without significantly compromising the accuracy.
To analyze the behavior of the PES without significant computational overhead, we adopted an averaged-modeling technique (Middlebrook and Cuk, 1977; Lee, 1990) to analyze the behavior of the PES. Figure 6 shows the averaged models of the PES converters in equivalent-circuit form (Lee, 1990). Unlike the switching models, the averaged models do not have any discontinuous elements. Further, the averaged (circuit) models enable the usage of built-in circuit modules in Simulink. We note that an averaged model ideally provides the averaged dynamics of the PES, and, as such, it transforms the PES model from a discontinuous to a smooth form. The accuracy of the averaged model, as shown in Figure 7, typically varies with the switching frequency, with higher accuracy at higher operating frequencies. At the operating switching frequency (20 kHz) the accuracy of the average model is 0.945 percent, whereas the simulation time is approximately 1 percent of that of the switching model.
Figure 6. Average Model of the Residential PES for the PSOFC PCS (UIC)
Figure 7. Comparison of Response of the Average Model to the Switching Model During a Sharp Transient (UIC)
System Performance and Emission Levels (Virginia Tech)
Performance. As already discussed above, FCSs are a promising technology for small-scale generation systems because of their potential for achieving high efficiencies and environmental benefits. In Figure 8, the full and partial load efficiencies of the PEM fuel cell system modeled here are compared to a phosphoric acid fuel cell (PAFC) system (Energy Nexus Group, 2002) and to conventional power generation technologies, (i.e., lean burn diesel engines [Energy Nexus Group, 2002]), reciprocating internal combustion Diesel engines (Wilbur, 1985) and micro-turbines (Energy Nexus Group, 2002). Note that the full-load efficiencies of conventional electrical power generation systems are dependent on system capacity and the type of fuel used. In general, this full-load efficiency increases as the size of the system increases at lower output ranges (< 1 MW).
In Figure 8, the full-load efficiencies of the PAFC system, the typical lean burn engine, and reciprocating internal combustion Diesel engine have similar values of about 35 to 37 percent, whereas that of PEM fuel cell system is 5 percent higher than the others. The partial load efficiency of fuel cell systems, however, remains fairly constant down to about 30 percent of full-load, whereas that of reciprocating internal combustion Diesel engine begins to drop rapidly after 60 percent of full-load. The graph shows that the micro-turbine has the lowest full-load and partial-load efficiency behavior.
Figure 8. Comparison of Full and Partial Load Efficiencies of Power Generation Systems (Virginia Tech)
Thus, in general the PEM fuel cell system not only exhibits a higher full-load efficiency than any of the conventional power generation systems but is more efficient over a wider operating range than these other technologies. Furthermore, it should be noted that the PEM FCS used in this comparison has as of yet not been optimized and, therefore, may come in at even higher efficiencies.
Emissions. Table 2 shows a comparison of the emission levels associated with the burning of fossil fuels in conventional power generation systems with those that result from the PEMFC system using methane. Note that the PEMFC system includes a combustor utilizing methane as a fuel and a steam methane reformer producing hydrogen that is consumed in the fuel cell stack. As can be seen, the pollutant emissions of the PEMFC system for CO, NOx, SOx, particulates, and Hg are significantly less than that for any of the conventional systems. In addition, even though the CO2 numbers are comparable in this table between the PEMFC system and conventional systems using natural gas combustion, this is misleading because CO2 production increases and decreases directly with efficiency, and the PEMFC system as shown in the previous section has a definite advantage in terms of efficiency at both full and partial loads.
Table 2. Comparison of Fossil Fuels and Predicted PEMFC System Emission Levels (Virginia Tech)
Pollutant |
Natural gas combustion |
Oil |
Coal |
PEM FC system |
CO2 |
50303.61 |
70511.05 |
89428.64 |
49549.55 |
CO |
17.19782 |
14.1882 |
89.42864 |
~0 |
NOx |
39.55498 |
192.6155 |
196.485 |
~0 |
SOx |
0.429945 |
482.3987 |
1113.989 |
Under 0.2 ppm |
Particulates |
3.009618 |
36.11541 |
1179.77 |
~0 |
Mercury |
~0 |
0.00301 |
0.006879 |
~0 |
EIA – Natural gas issues and trends, 1998 |
To illustrate this point, Figure 9 provides a comparison of the CO2 emission levels of the power generation systems appearing in Figure 8, referenced to the data given in Table 2. The PAFC, microturbine, and lean-burn diesel engine use natural gas and the PEMFC methane, whereas the reciprocating IC diesel engine runs on diesel. This graph shows that CO2 emissions increase almost linearly as the percent load increases and that the CO2 emissions of the reciprocating IC diesel engine and microturbine are higher than that of the other power generation systems by about 60 percent throughout the entire load range. Although the microturbine uses natural gas, it produces more CO2 emissions than that of reciprocating IC diesel engine because of its overall low part- and full-load efficiency. Moreover, both the reciprocating IC diesel engine and microturbine produce higher NOx emissions at high temperature so that additional system complexity is required to reduce these emissions.
As to a comparison between the two FCSs and the lean-burn diesel engine, the as-of-yet-not-optimized PEMFC system has, nonetheless, the lowest CO2 emissions across the entire load range with the lean-burn diesel engine falling somewhere between the two FCSs. In fact, the PEMFC system’s CO2 emissions are consistently about 10 to 13 percent lower than those for either the PAFC system or the lean-burn diesel engine.
Figure 9. Comparison of CO2 Emission Levels of the Power Generation Systems Appearing in Figure 8 (Virginia Tech)
WRAS, FPS, and SS Controls (Virginia Tech).
Not only is the optimal synthesis/design of the PEMFC system’s overall configuration and component capacities and performances of importance for achieving the goals of low emissions and high efficiencies at minimum cost but so is controlling the system’s static and dynamic operation, which is critically dependent on the control strategies used and the control architectures developed. At present multi-input/output controls are being developed for the WRAS in terms of their overall architecture. The WRAS includes an air compressor with speed control features and is used to supply different amounts of air for a wide range of power demands to reduce parasitic power losses. In addition, to recover energy from the gases exiting the FPS, an expander with flow rate control has been integrated with the compressor, as has an electrical motor to make up the difference in the work required by the compressor and that recovered by the expander (see Figure 4). At the moment, a comparison between classical and state space control theory design for this subsystem is being studied and the results will be applied not only to the WRAS but to the FPS and SS, as well.
PES Controls (UIC)
The FCS static and dynamic performance and energy efficiency are dependent critically on the intelligent design of the control system and control strategies. In conventional fuel cell power systems, each individual subsystem is controlled individually and tries to achieve the individual desired objective with the best possible time. The power converters are controlled to provide a constant output voltage (DC or AC) in the steady state and should have minimal overshoot and undershoot during a transient. The BOPS determines the required flow rate of the air and fuel into the FCS based on the power requirements of the load and maintains the temperature inside the stack. Therefore, it is designed to provide the required amount of fuel and air to the stack in the fastest possible time and maintain the temperature in the stack with minimum variations. The control for the battery or the BMS is designed to provide the additional current during the transient with very fast response.
To alleviate the computational complexity of a centralized controller, hierarchical control architecture for the fuel cell power system, as shown in Figure 10, is conceived. This hierarchical control enables mutually independent control design, leading to higher flexibility, and enables easier and faster calculation of stable operating points for the individual subsystems; this optimizes the system performance and reliability.
Figure 10. Comprehensive Hierarchical Controller Architecture for the PEMFC-Based Power System (UIC)
Battery-PEMFC Hybrid System Control. The output voltage of the PEMFC drops as the current drawn from the stack increases, because of the polarization loss. During a sudden load current transient, the input fuel-flow rates remain constant because of the sluggishness of the balance of the plant subsystem. Therefore, because of a no-load to full-load transient, the fuel utilization increases abruptly, which may reoxidize the anode of the fuel cell. To alleviate this, a battery needs to provide the additional power during the load transient and needs to sink the surplus power during a drop in the load current.
We designed a power electronics topology for the hybrid power management system as shown in Figure 11, in which one or multiple DC-DC converters are connected at the output of the stack and the battery, which are then connected in parallel to share the load. The number of parallel converter modules, N, would depend upon the rated power of the stack and the battery module. The switches R1… RN activate and deactivate the converter modules, and the number of active DC-DC converter modules at any time depends on the amount power drawn from the stack and/or the battery and the efficiency of the converter modules at particular power drawn by each of the active converter module. The switches are used to enable access of the battery for discharging as well as the charging through multiple converter modules. Now, the current drawn by each of the active converters connected to the stack and/or the battery at any time is calculated based on the total required current and the efficiency of the converter modules, which maximizes the overall efficiency of the PES.
Figure 11. Topology of Multiconverter-Based Hybrid Power Management System (UIC)
Efficient Power Sharing Strategy. The efficiency of any DC-DC converter depends upon the power delivered by the converter. At higher output power, the losses increases with increase in the converter current, leading to a drop in the efficiency. On the other hand, at very low output power, the losses become a larger percentage of the output power, which again decreases the converter efficiency.
Mathematically, for each of the k converter modules the efficiency map, experimentally obtained for a converter module, is given as
ηi = f(pi) i = 1,2,...k (1)
where pi the fraction of the rated output power, Prated of the converter. Now assuming, Pi to be the input power for the ith converter, such that, we have,
.
Now equation (1) can be rewritten as
ηi = f(ηi, mi) i = 1,2,...k where, mi= Pi /Prated (2)
The iterative equation (2) can be solved to obtain the equation of the ηi terms of mi only
ηi = fm(mi) i = 1,2,...k (3)
Now the efficiency of the system is given as,
(4)
To maximize the efficiency of the converter system, the values of mi need to be obtained, which maximize the function, for a given input power, and satisfy the constraint
∑ mi fm
, and mi ≤ 1.
Figure 12 shows the effect of power sharing strategy on the overall converter system efficiency for a five module DC-DC boost converter using the efficient power sharing strategy. The fraction of the input powers drawn by each of the converter modules Px /Pin shown in the range of the input power. This shows that the converter system efficiency is improved significantly at lower loads.
Figure 12. Comparison of the Converter System Efficiency Based on Power Sharing Strategy (UIC)
Conclusions (Virginia Tech and UIC)
UIC
We designed a stationary PES based on switching discontinuous model. To reduce the computational complexity of the discontinuous model, a highly accurate average model is deduced based on averaging the switching discontinuity, leading to a reduction in the simulation time by a hundredth.
We designed an optimal multiconverter topology and control for the hybrid power system, based on the efficient power sharing strategy. This strategy not only increases the efficiency of the converter system but also eliminates the need for the dedicated converters for the battery and the PEMFC, leading to decreased cost of power electronics subsystem.
Virginia Tech
We developed a fully transient, nonlinear, general model of PEMFC stacks and a variety of BOPS (FPS and WRAS) models. The FPS simulation shows high methane conversion and very low emission, while the cell voltage dynamic response of the SS to a sudden change of cell current density has been studied. The WRAS model shows that around 35 percent of the power load required by the compressor can be recovered by the expander in the fuel cell system. Furthermore, comparisons have been made between the FCS (WRAS, FPS, and SS) and a number of other power generation systems with respect to system performance and emission levels.
Finally, the multi-input/output control architecture is being developed for the WRAS and the results will be used to develop those for the FPS and SS.
Future Activities:Virginia Tech—Detailed, fully transient, nonlinear models have been developed, implemented, and validated for all of the PEMFC system components shown in Figure 1. System integration is presently underway; once complete, various optimization and control strategies will be applied to determine the optimal dynamic synthesis/design and operational/control of the system based on a variety of realistic load profiles. A new control algorithm based on state-space will be evaluated with the optimization strategy and it will be compared to results based on conventional control strategies. To facilitate solving the extremely complex and difficult problem of integrated system synthesis/design/operation for variable loads and/or environmental conditions, two different overall strategies will be used. The first, a bottom-up approach, will utilize a very effective decomposition strategy called Dynamic Iterative Local-Global Optimization, which was developed by our Center for Energy Systems Research to facilitate the type of large-scale optimization problem that presents itself here. The use of a top-down approach based on game theory and a probability analysis of the uncertainties present will also be explored in collaboration with our partner at the University of Illinois at Chicago (UIC) to determine its appropriateness for use in tackling the optimal operational/ control problem. If applicable, the results of this approach will be compared with those of the bottom-up approach to verify or modify these results and provide further insights. Virginia Tech also will collaborate with UIC in integrating our FPS, WRAS, and SS with the PES that they are developing.
So far, an optimal power electronics topology for the PEMFC residential power electronics subsystem has been developed based on a fully transient nonlinear model. System integration will be implemented and various optimization and control strategies will be applied to the system integration and simulation based on various realistic load profiles for developing optimal system synthesis/design/operation strategy.
References:
Ceraolo M, Miulli C, Pozio A. Modelling static and dynamic behaviour of proton exchange membrane fuel cells on the basis of electro-chemical description. Journal of Power Sources 2003;113(1):131-144.
Choi Y, Stenger HG. Kinetics, Simulation and insights for CO selective oxidation in fuel cell applications. Journal of Power Source 2004;129:246-254.
de Bruijn FA, Stobbe ER, Wakker A. Catalysis in low temperature fuel cell system. Presented at the Fifth Netherlands Catalysis and Chemistry Conference, June 2004.
Energy Nexus Group. Technology characterization-micro-turbines, 2002.
Energy Nexus Group. Technology characterization-fuel cells, 2002.
Ettouney HM, Shaban HI, Nayfeh LJ. Theoretical analysis of high and low temperature shift concerters. Chemical Engineering Community 1995;134:1-16.
Rancruel-Acre DF. Dynamic synthesis/design and operation/control optimization approach applied to a solid oxide fuel cell base auxiliary. Ph.D. Dissertation. Virginia Polytechnic Institute and State University, Blacksburg, VA, 2005.
Van den Oosterkamp P, de Bruijn F. Fuel specification for fuel cells. Presented at the EU Workshop on Regulations, Codes, and Standards for H2/FC Technologies, 2005.
Wilbur LC. Handbook of Energy Systems Engineering: Production and Utilization. Wiley, Sommerset, NJ. 1985/10/01/:1-1792.
Xu J, Froment GF. Methane steam reforming and water gas shift: I. Intrinsic kinetics. AIChE Journal 1989;35(1):88-96.
Xu J, Froment GF. Methane steam reforming: II. Differential limitations and reactor simulation. AIChE Journal 1989;35(1):97-103.
Erickson RW, Maksimovic D. Fundamentals of Power Electronics, Second Edition. Kluwer Academic Publishers, 2004.
Lee FC. Modeling, Analysis, and Design of PWM Converters. Virginia Polytechnic Institute and State University, Blacksburg, VA.. 1990:1-555.
Middlebrook RD, Cuk C. A general unified approach to modeling switching-converter power stages. In: Proceedings of the IEEE Power Electronics Specialists Conference, 1977, pp. 521-550.
Journal Articles:No journal articles submitted with this report: View all 1 publications for this project
Supplemental Keywords:Proton exchange membrane fuel cells (PEMFCs), total energy systems, distributed cogeneration, fuel reforming, energy integration, detailed transient nonlinear models, large-scale optimization, decomposition strategies, game theory, probabilistic uncertainty analysis, optimal control, state space based control, reduced emissions, high part and full load efficiencies, PEM fuel cell system, balance of plant subsystem, work recovery and air supply subsystem, power electronics subsystem, optimization, switching nonlinear power electronics models, average models, synthesis/design, top-down and bottom-up approaches,
,
TREATMENT/CONTROL, Sustainable Industry/Business, Scientific Discipline, RFA, POLLUTION PREVENTION, Technology for Sustainable Environment, Sustainable Environment, Technology, Energy, Environmental Engineering, energy conservation, environmental sustainability, clean technologies, green design, environmentally friendly fuel cell power system, alternative energy source, air pollution control, energy efficiency, innovative technology, modeling, environmentally conscious design, emission controls
Progress and Final Reports:
Original Abstract