Controlled Battery Charger for Electric Vehicles - piers

PIERS ONLINE, VOL. 6, NO. 6, 2010

532

Controlled Battery Charger for Electric Vehicles M. Geske1 , T. Winkler2 , P. Komarnicki2 , and G. Heideck1 1

2

Otto-von-Guericke-University Magdeburg, Magdeburg, Germany Fraunhofer Institute of Factory Operation and Automation IFF, Magdeburg, Germany

Abstract— Due to rising fuel consumption, price of CO2 emissions and growing urban air

pollution, the global interest of the automobile industry, politics and scientists in electric mobility is increasing in the recent years worldwide. Thus, future challenges will be the integration of electric vehicles in distribution networks under the scope of balancing multiple charging processes while, at the same time, increasing dispersed generation. The development of a controlled battery charger for traction batteries aims to reduce the peak load and to shift energy demand.

1. INTRODUCTION

During charging, electrical energy is converted and stored in form of chemical energy. A conventional battery charger is supplied by mains voltage and feeds the battery with DC voltage and current. Therefore, a battery charger adapts the voltage and the charging current from the mains voltage to the specified battery parameters. For electric mobility, battery chargers of electric vehicles (EV) should fulfill additional future task. At one hand, exceeding of the maximum power of the grid and possible overloads in distribution networks by simultaneous charging of a fleet of EVs with large batteries must be avoided by means of controlled battery charging. At the other hand, batteries of EVs can be a possibility for better use of variable power sources like wind power [1] or photovoltaic [2, 3]. The more flexible use of traction batteries is a challenge for the future of electric mobility. The application of traction batteries as an energy storage for power grids is also moving into focus of research [4–7]. Hence, a charging process can be adapted to the total power consumption of a household to reduce or shift power consumption. Controlling the charging current makes it possible to keep load profiles inside defined ranges. A charger for the batteries of a small-sized experimental vehicle was developed and tested. The results of the operation of the charger will be presented and discussed. 2. CONCEPT OF THE BATTERY CHARGER

Requirements to electric vehicles and their charging environment, resulting from the grid parameters, should be considered in the charging concepts carefully [8]. The most comfortable option would be contact-less charging, it offers the possibility to charge batteries during stops along the route [9], but it demands adequate and cost-intensive charging spots. However, single-phase sockets will be the most available and public accessible connection points for electric vehicles. Due to a simple access to the electric grid and the lower complexity of the system components a concept for conductive battery charging was chosen for experimental setup. The battery charger was developed for a lithium iron phosphate battery with 48 V and a capacity of 15 Ah. The charging device (Figure 1) consists of an input rectifier and a d.c. link, an inverter bridge on basis of a commercial IGBT module, a medium frequency transformer for adapting the voltage level to the battery voltage and for insulation from the mains, an output rectifier and a control unit.

Figure 1: Topology of the developed battery charger.

PIERS ONLINE, VOL. 6, NO. 6, 2010

533

3. DESIGN OF THE TRANSFORMER

The frequency is set to 10 kHz because of the maximum frequency of the IGBT. The transformer is designed for a minimum frequency of 5 kHz and allows even higher frequencies. Therefore, the core of the transformer requires material with special frequency characteristics. The chosen core material is characterized by lower hysteresis losses and a steep hysteresis curve due to a ferrite material. The dimensioning was made for an U93 core. Material properties and magnetic parameters used for calculation of the windings are shown in Table 1. Table 1: Material properties and magnetic parameters of the core. Parameter Maximum flux density Length of magnetic path Magnetic cross-section

Symbol Bmax LF e AF e

Value ≥ 455 mT 355 mm 840 mm2

The voltage that is induced in a number of windings N by a magnetic flux Φ can be calculated by Eq. (1) dΦ V =N (1) dt Assuming a triangular current, resulting from the primary voltage V1 , which is approximately rectangular, the voltage induced by the current can be calculated by Eq. (2). ˆ ∆Φ 4·Φ ˆ V1 = N 1 = N1 =4·f ·Φ T Z ∆t Φ = B · dA ˆ = B ˆ · AF e Φ

(2) (3) (4)

ˆ using The number of turns N1 (Eq. (5)) can be obtained by substituting the magnetic flux Φ Eq. (4). V1 N1 = (5) ˆ · AF e 4·f ·B The transformer is dimensioned for the maximum of primary voltage V1 resulting from the rectified mains voltage VM V , Eq. (6). √ 2· 2ˆ V1 = VM V (6) π At the maximum mains voltage (nominal voltage of 230 V + 10% overvoltage), V1 equals 227.2 V. The number of turns of the primary winding N1 is set to 30. The number of turns of the secondary winding must enable a secondary voltage of 75 V also in case of 10% undervoltage. Therefore the secondary number of turns N2 is set to 11. The windings of the transformer consist of copper foil with thickness of 0.125 mm and width of 83 mm. The windings are electrically isolated by insulating foil with a thickness of 40 µm. For the selected material and a given current density of 3 A/mm2 the maximum current is 31.2 A for each winding. The transformer consists of two primary and two secondary windings. Using these windings in parallel allows higher currents and halves the value of leakage inductance, approximately. 4. IMPLEMENTATION OF CHARGING CHARACTERISTICS

Chargers for traction batteries can be subdivided into chargers with taper characteristic and chargers with regulated characteristic [10]. The taper characteristic charges in relation to the resistance but has the disadvantage that the charge time varies corresponding to the variation of the mains voltage. Regulated charging controls the current and voltage, achieves shorter charging times and enables the control of charging current for controlled charging. To investigate the developed controller a lithium iron phosphate battery with 15 Ah capacity was charged. This type of battery

PIERS ONLINE, VOL. 6, NO. 6, 2010

534

requires the constant current constant voltage (CCCV) charging characteristic. The first step of charging is characterized by constant charging current. When the battery voltage exceeds the charge termination voltage the charge control switches to constant voltage charging. The controller for charging can be made by integrated circuits [11] or by programming a microcontroller that enables quick adaptation of the reference variables. Thus, a simplified PI control algorithm has been implemented, that changes the reference variable from charging current to charging voltage in the event that the charge termination voltage is exceeded. The charge termination of the whole charging process is initiated through by fall below a minimum charging current of 50 mA. The current and the voltage transducer provide accuracies of 0.8 and 0.9%. The sensors of the battery charger were calibrated using a high precised measurement device. The battery voltage in Figure 2 increase weakly up to the charge termination voltage of 58.4 V and indicates the typical charging behavior of lithium iron phosphate batteries due to a strong reversible two phased reaction [12]. The further constant voltage charging is terminated after half an hour through the mentioned current criteria. 5. INVESTIGATION OF CONTROLLED CHARGING

The controlled charging modifies the reference variable of the PI control algorithm during the charging process. A scenario for a useful investigation of controlled charging is given by means of a shifted charging process adapted to a household. The investigations use the German standard load profile H0 [13]. This profile, shown in Figure 4(a), reflects an averaged energy consumption of households on working days in winter, normalized to 1000 kWh per year. The approach of controlled charging aims to keep the sum of energy consumption below a defined value, in the test below 150 W after 10 p.m. The measurement is carried out using the normalized value, the test results will be scalable. The experimental setup in Figure 3 uses simulation software to compare the normalized load profile and the current power consumption of battery charging with the limiting value of 150 W. Every five minutes the software calculates a new value for the reference variable and sends it via serial interface to the battery charger. The resolution of the reference variable was set to 0.1 A. To achieve comparable results the battery capacity was discharged to 10.2 Ah before the charging process starts. The charging processes have been initiated at 10 p.m. (Figure 4(a)). The uncontrolled charging

Figure 2: Measured charging characteristic of the developed battery charger.

Figure 3: Experimental setup.

PIERS ONLINE, VOL. 6, NO. 6, 2010

535

process in Figure 4(a) is similar to the characteristic in Figure 2 and causes a high peak of 288 W during the charging process. The load profile relating to the controlled charging shows maximal total power consumption that does not exceed the defined limit of 150 W during battery charging. The graph of charging current in Figure 4(b) points out the implemented resolution of the reference variable and time steps. The controlled charging extends the charging time up to 4 hours, while the corresponding uncontrolled charging time is 1 hour and 48 minutes.

(a)

(b)

Figure 4: (a) Comparison of different charging concepts based on normalized load profile H0 and (b) resulting charging characteristic for controlled battery charging. 6. CONCLUSIONS

The functionality of the developed battery charger for an experimental electric vehicle has been proved by charging a lithium iron phosphate battery with a nominal capacity of 15 Ah. A proposal for controlled battery charging of electric vehicles has been successfully implemented. Reducing the peak load of households within controlling the charging current has been demonstrated. The total power consumption has been kept below a defined limit by temporally modifying the internal reference variable of the battery charger. The calculation of the reference variable was made by means of simulation software that processes present measurement data and a standard load profile for households. Batteries of real EVs are some times bigger than the one of the experimental setup. The control of battery charging on the basis of a given constant or varying maximum load can be an effective instrument to reduce peak loads in the electric power grid as compared to uncontrolled charging. Further investigations are necessary to generate the peak load signals depending from the load status of the distribution networks and to transfer these signals to real households and to the chargers of the EVs. REFERENCES

1. Kiviluoma, J. and P. Meilborn, “Influence of wind power, plug-in electric vehicles, and heat storages on power system investments,” Energy, 2009. 2. Dunbar III, P. B., “Solar-to-vehicle (S2V) systems for powering commuters of the future,” Journal of Power Sources, Vol. 186, No. 2, 2009. 3. Broderick, R. J. and J. J. Bzura, “Investigation of Plug-in Hybrid EVs (PHEVs) and their effects on power distribution systems in the NE and SW united states: Summary of a panel session presentation,” IEEE, 2009. 4. Kempton, W. and J. Tomic, “Vehicle-to-grid power fundamentals: Calculating capacity and net revenue,” Journal of Power Sources, Vol. 144, No. 1, 268–279, 2005. 5. Turton, H. and F. Moura, “Vehicle-to-grid systems for sustainable development: An integrated energy analysis,” Technological Forecasting and Social Change, Vol. 75, No. 8, 1091–1108, 2008. 6. Sauer, D. U., M. Kleimaier, and W. Glaunsinger, “Relevance of energy storage in future distribution networks with high penetration of renewable energy sources,” Proceedings of 20th International Conference on Electricity Distribution (CIRED), Prague, Czech Republic, 2009. 7. Heuer, M., G. Heideck, Z. A. Styczynski, T. Winkler, P. Komarnicki, and G. Mller, “Electric vehicle charging stations in Magdeburg,” Proceedings of 5th International IEEE Vehicle Power and Propulsion Conference, 60–67, Dearborn, USA, September 2009.

PIERS ONLINE, VOL. 6, NO. 6, 2010

536

8. Kroitzsch, J. and T. Winkler, “Intelligentes batteriemanagementsystem f¨ ur elektrofahrzeuge,” Proceedings of 6th IFF-Kolloquium, 72–76, Magdeburg, Germany, November 2008. 9. Villa, J. L., J. Sall´an, A. Llombart, and J. F. Sanz, “Design of a high frequency inductively coupled power transfer system for electric vehicle battery charge,” Applied Energy, Vol. 86, No. 3, 355–363, 2009. 10. Kiehne, H. A., Battery Technology Handbook, Second Edition, Marcel Dekker Inc., Basel, New York, 2003. 11. Tsang, K. M. and W. L. Chan, “A simple and low-cost charger for lithiumion batteries,” Journal of Power Sources, Vol. 191, No. 2, 633–635, 2009. 12. Jossen, A. and W. Weydanz, Moderne Akkumulatoren Richtig Einsetzen, Reichhardt Verlag, Untermeitingen, 2006. 13. Schieferdecker, B., C. F¨ unfgeld, H. Meier, and T. Adam, “Repr¨asentative VDEW-Lastprofile,” VDEW-Materialien M-28/99, 2009.