Some Constraints on the Reuse of Li-ion Batteries in Data Centers

DOI: 10.1515/SBEEF-2016-0010 SOME CONSTRAINTS ON THE REUSE OF LI-ION BATTERIES IN DATA CENTERS 1 C. CEPISCA1, G.C. SERITAN1, S.D. GRIGORESCU1, SABINA POTLOG1, S. GANATSIOS2 Departament of Measurements, Apparatus and Static Converters, University Politehnica, Splaiul Independentei 313, Bucharest, Romania 2 TEI West Macedonia, Kozani, Greece E-mail: [email protected], ganatsios.teikoz.gr Abstract. The development of rechargeable electrochemical sources, especially those based on Li-ion technology, has opened the way for their use in various fields ranging from electronics and IT to electric cars. New ideas appear, such as using replaced the batteries from electrical vehicles in the development of energy storage devices for critical consumers such as Data Centers. New uses impose detecting the unfavorable operating conditions which may endanger the power supply of Data Centers. This paper, by modeling and simulation, analyzes some particular problems of Li-ion batteries that appear due to the differences in state of charge of the cells. Keywords: Circuit simulation; Li-ion batteries modeling; State of charge; Data Center 1. INTRODUCTION Different economic sectors have important needs related to the use of rechargeable batteries. For example, consumer electronics use relatively small batteries with high energy density and low power. Other sectors, such as electric car industry, appeal to rechargeable sources with high capacity, high voltages that can produce high electrical currents [14], [20-21]. An overview of the opportunities for energy storage is presented in [1]; being considered lead batteries, nickel metal hydride, lithium batteries and capacitors – Figure 1. Figure 1. Energy storage technologies - specific power and energy In applications, more than half of the turnover coming from the trade with batteries was made with the Li-ion battery type [2]. This battery requires little maintenance, has no "memory", self-discharge is lower than in other devices and it can power digital equipment directly due to the nominal voltage of 3.6 V cell, offering simplifications and thus cost savings. The cells of a battery can be connected in series (high voltage), parallel (high current) or mixed. The lifetime of these batteries used for electric vehicles is at least five years. Lately there are new solutions for further use of Lithium batteries after removal from duty cycle to power electric vehicles and one of the proposed directions is to use them in complex systems for powering Data Centers, replacing the UPSs [3]. By using a large number of partially degraded batteries used for electric vehicles, energy from renewable sources (photovoltaic, wind) can be stored in the Data Center, which helps reduce energy consumption from conventional networks. It is considered that this concept can be widely extended until 2020 and Li-ion batteries can find a new utility in Data Centers and also in household applications based on renewable energy [12], [16]. The use of these rechargeable batteries, partially worn, must take into account the following aspects: - In the process of charging the Li-ion cells the voltage must not exceed 4.2V; otherwise, damage may occur due to thermal effects and may even lead to fires [15]; - Exceeding the electrical values particular to the charging procedure leads to accelerated degradation of electrochemical phenomena and drastically reduces the lifetime [4]. For example, by increasing the charging voltage from 4.2 V to 4.25V, the degradation of the cell is 30%; - Partial use of energy from the battery elements. If one of the cells reaches the required discharge limit, the protection system interrupts the discharge, although the other cells are not discharged to the extent imposed [17]. The occurrence of voltage differences between the cells of a Li-ion battery or between the batteries that make up a power storage medium is a situation that can create difficulties. OCV (open circuit voltage) and internal resistance of the battery cells depend on the State of Charge (SoC) [5-6], [13], [18-19]. Battery State of Charge (SOC) is defined as the ratio of the remaining capacity and the total capacity (Qn) of the battery: ܵ‫ܥ݋‬ሺΨሻ ൌ ೟ ொబ ି‫׬‬బ ௜ሺ௧ሻௗ௧ ொ೙ (1)  Here, Q0 is the initial capacity of the battery and i(t) is the discharge current. The management of charging/discharging of these devices uses a continuous measurement of the voltage across the battery components. The values are an indication of the SoC of the batteries [7]. This paper aims to realize modeling and simulations of Liion batteries in different states of charge to show the possible use of these batteries in the applications listed above and indicate the technical measures for the realization of the circuits. 2. MODELING OF LI-ION BATTERY The equivalent circuit for the Li-ion cell is done with resistors and capacitors – Figure 2. Resistor Ri models internal resistance and transient effects can be taken into account by the RC network [8]. Because our study analyzes the permanent working regime, the used model does not include the RC group [9]. (b) Figure 3. Variation of cell parameters: (a) OCV =f(SoC), (b) Ri = f(SoC) 3. SIMULATION RESULTS The battery taken into account in the considered study consists of eight parallel branches, each branch having 12 identical cells connected in series. Figure 4 shows battery model used and the possibility of changing the cell parameters with the values presented in Figure 3. Figure 2. Thevenin model for Li-ion battery According to technical data [10], [11], Figure 3 indicates the OCV values and values of internal resistance Ri of the Li-ion cell at various states of charge. These values were used in the simulations described in this paper. Figure 4. Battery modeling Were studied following circuits made with varying degrees of load cells. a) All the cells charged SoC=100% In this situation, the cell parameters considered were the following: OCVcell =4.2V, Ri=0.0345Ω. Simulation result is shown in Figure 5. The voltage of the battery at no load operation was determined to be OCV= 50.4V. This circuit provides the highest voltage that can be obtained from the battery so configured. (a) Parameters of the cell, with SoC=10%, are: OCVcell =3.3V, Ri=0.06 Ω. Figure 7(a) shows the new structure of the battery. The value of the current through a branch is 39.83A (Figure 7(b)) and the total current in the external circuit is 318.64A (Figure 7(c)). Figure 5. Battery voltage at no-load operation By connecting a resistor Rext= 0,1Ω to the battery terminals (Figure 6 a), the branch currents are 41,38A (Figure 6 b). The total current, given by the eight branches in parallel through the external resistor is 331,04A (Figure 6 c). (a) (b) (c) Figure 7. Voltage and currents for a cell with SoC=10% on each branch (a) (b) (c) Figure 6. Voltage and currents in the circuit for all cells SoC=100% The powers in this case are: - Debited by the battery elements: Pt = 16.684 kW; - In the external circuit: Pu =10.958 kW. Only 65.7 % of the energy produced by the battery is transferred to the external circuit. b) A cell with SoC =10% on each branch The presence of a discharged cell on each branch leads to a decrease in the total current from 331.04A to 318.64A and for output voltage from 33.10V to 31.86V. The powers in this case are: - Debited by the battery elements: Pt = 15.758 kW; - In the external circuit: Pu =10.153 kW. One can see a decrease of 6% in energy produced by the battery to the situation in which all cells were fully charged. Only 64.4 % of the energy produced by the battery in this case is taken of external consumer, increasing the heating effect of the cells. c) A branch with SoC=10% cells, seven branches with SoC=100% cells The battery structure is shown in Figure 8(a). In this situation, in which entire branch consists of discharged cells, the currents are presented in Figure 8(b). The current in the external circuit has a value 319.6A – Figure 8(c), the current in the branch with SoC=10% is 10.6A and in the regular branch the current is 44.14A. The current through the branch with discharged cells is over four times lower than in branches containing cells fully charged. Similarly, the power sent to an external circuit by the branch with discharged cells Pd =388.7W is much smaller than that produced by a branch with fully charged cells Pf= 1410.7W. (a) (a) (b) (c) Figure 9. Battery structure and currents for various SoC cells The powers in this case are: - Debited by the battery cells: Pt = 13.808 kW; - On the external circuit: Pu =8.311 kW. Only 63.5 % of the energy produced by the battery,, in this case is taken of external consumer, increasing the heating effect of the cells. The power delivered to the external circuit drops to 90% of the nominal value of the battery with fully charged cells. (b) (c) Figure 8. Battery structure and currents for a branch with SoC=10% cells The powers in this case are: - Debited by the battery elements: Pt = 15.992 kW; - In the external circuit: Pu =10.214 kW. Only 63.8 % of the energy produced by the battery, in this case, is taken of external consumer, increasing the heating effect of the cells. d) The battery has branches with different SoC Figure 9 presents the results in the case of a battery containing branches with completely discharged SoC = 10% cells and SoC = 50% cells. The current in the external circuit is 288.3 A – Figure 9(b). The current in the branch with SoC=10% is 14.46A, in the branch with SoC=50% the current is 25.9A and in the regular branch the current is 51.65A. The existence of a discharged cell on the branches leads to a decrease in the total current from 331.04A to 288.3A and for output voltage from 33.10V to 28.83V. 4. CONCLUSIONS Developing a Li-ion multiple cell batteries involve continuous monitoring of the electrical parameters, in order to assess its state of charge. Modeling and simulation results show that the difference in the state of charge of the component cells leads to a drop of the current in the external circuit and of the power. At the same time, heat dissipation on the internal resistance increases, leading to effects that can damage the battery. A more complicated solution is envisaged for installations composed of used Li-ion batteries. In these cases, in addition to the differences between the battery cells, significant differences between the batteries that make up the construction may also occur. In addition to a thorough study regarding the electrochemical phenomena in these situations, the need for sophisticated monitoring circuits, which ensure the operation of the battery assembly at optimal parameters, is taken into account. 5. ACKNOWLEDGMENT The work has been funded by the Sectorial Operational Programme Human Resources Development 2007-2013 of the Ministry of European Funds through the Financial Agreement POSDRU/159/1.5/S/134398 and by Grant no.30_PCCA_2012 of the Romanian National Authority for Scientific Research CNDI– UEFISCDI. 6. REFERENCES [1] F. Conte, Battery and Battery Management For Hybrid Electric Vehicle: A Review, Elektrotechnik & Informationstechnik, 123, 10(10), pp. 424-431. [2] A Guide to Understanding Battery Specifications MIT Electric Vehicle Team. [Online] Available at: http://mit.edu/evt/summary_battery_specifications. pdf [3] P. Buckley, Second life EV batteries to power data centers?, Technology News, March 19, 2014. [4] J.G. Lozano, M.I.M Montero, M.A.G. Martinez and E.R. Cadaval, Electric Vehicle Battery Charger for Smart Grids, Electric Power Systems Research, Vol.90, 2012. [5] V. Pop, H.J. Bergveld, P.P. Notten and P. Regtien, State-of-Art Of Battery State-of- Charge Determination, Measurement Science and Technology, 16(12), 2005, pp. 12. [6] F.Trinh, A Method for Evaluating Battery State of Charge Estimation Accuracy, Master Thesis EX080/2012, Chalmers University of Technology , Gothenburg. [7] T. Wu, M. Wang, Q. Xiao, X. Wang, The SOC Estimation of Power Li-Ion Battery Based on ANFIS Model, Smart Grid and Renewable Energy, 2012, 3, pp.51-55. [8] H. Culcu, B. Verbrugge, N. Omar, P. Van Den Bossche, J. Van Mierlo, Internal resistance of cells of lithium battery modules with FreedomCAR model, in Proc. EVS24, Stavanger, Norway, May 13-16, 2009. [9] V. Cimpoeru, C. Cepisca, Analysis of the state of charge of rechargeable dc sources to optimize their embedded control systems”, U.P.B. Sci. Bull., Series C, Vol. 76, Iss.1, 2014. [10] Lithium Ion Rechargeable Batteries -Technical Handbook, Sony Corporation. [Online] Available at:www.esupport.sony.com. [11] M. Kultgen, J. Munson, Battery Stack Monitor Extends Life of Li-Ion Batteries in Hybrid Electric Vehicles”, Linear Technology, vol. XIX, no.1, 2009 [12] R. Tipley, PUE:A Comprehensive Examination of the Metric, The Green Grid EMEA Forum 2012, [Online] Available at: www.thegreengrid.org [13] R.A. Ugle, Effects of Internal Resistance on Performance of Batteries for Electric Vehicles, Master Thesis, University of WisconsinMilwauke, august 2013, [Online] Available at: http://dc.uwm.edu/etd [14] B. Enache, E. Lefter, C. Cepisca, Batteries for Electrical Vehicles: A Review, Chapter 15. Autonomous Vehicles: Intelligent Transport Systems and Smart Technologies , Nova Science Publishers, 2014. [15] S. D. Grigorescu, C. Cepisca, V. Cimpoeru, Virtual Instrumentation Based Laboratory for Batteries Testing, in Proc. 8th International Symposium on ADVANCED TOPICS IN ELECTRICAL ENGINEERING, Bucharest, 2013, pp.1-4. [16] D. P. Iracleous, N. Papadakis, C. Cepisca, Administration and control of autonomous agents in distributed environment. Application to DoDaF, in Proc. 2nd International Conference on Applied and Computational Mathematics (ICACM '13), Vouliagmeni, Athens, Greece, 2013, pp. 166-170. [17] J. Borgersen, Multicell Battery monitoring and balancing with AVR, Master Thesis, Norwegian University of Science and Technology, Depy. Electronics and Telecommunications, 2009 [18] AVR453: Smart Battery Reference Design, Atmel Norway AS, 2559D-AVR-02/06, 2006, [Online] Available at: www.atmel.com.avr [19] J. Ye, J. Fallin. K.K. Rushil, Extract maximum power from supply when charging a battery, Power Management, Analog Application Journal, Texas Instruments, [Online] Available at: www.ti.com/aaj [20] D. Tuite, Vehicle Charging Standards: A Quick Primer, Electronic Design, 05,2014, [Online] Available at: http://electronicdesign.com/print/power [21] Porumb, R.; Postolache, P.; Golovanov, N.; Toader, C. - Continuity of supply assessment in electric distribution systems with distributed generation, Proceedings of Universities Power Engineering Conference (UPEC), 2010 45th International, Aug. 31 2010-Sept. 3 2010, Pag. 1- 5, Cardiff, UK,