Lithium-Ion Battery Safety Study Using Multi-Physics Internal Short-Circuit Model The 5th Intl. Symposium on Large Lithium-Ion Battery Technology and Application in Conjunction with AABC09 June 9-10, 2009, Long Beach, CA Gi-Heon Kim, Kandler Smith, Ahmad Pesaran National Renewable Energy Laboratory Golden, Colorado gi-heon.kim@nrel.gov This research activity is funded by US DOE s ABRT program (Dave Howell) NREL/PR-540-45856 NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.
2 Background NREL s Li-ion thermal abuse modeling study was started under the Advanced Technology Development (ATD) program; it is currently funded by Advanced Battery Research for Transportation (ABRT) program. NREL s previous model study focused on understanding the interaction between heat transfer and exothermic abuse reaction propagation for a particular cell/module design, and provided insight on how thermal characteristics and conditions can impact safety events of lithium-ion batteries. Total Volumetric Heat Release from Component Reactions 4 8 12 16 20 24 28 32 36 40 sec
Focus Here: Internal Short-Circuit Li-Ion thermal runaway due to internal short-circuit is a major safety concern. Other safety concerns may be controlled by electrical and mechanical methods. Initial latent defects leading to later internal shorts may not be easily controlled, and evolve into a hard short through various mechanisms: separator wear-out, metal dissolution and deposition on electrode surface, or extraneous metal debris penetration, etc. Thermal behavior of a lithium-ion battery system for an internal shortcircuit depends on various factors such as nature of the short, cell capacity, electrochemical characteristics of a cell, electrical and thermal designs, system load, etc. Internal short-circuit is a multi-physics, 3-dimentional problem related to the electrochemical, electro-thermal, and thermal abuse reaction kinetics response of a cell. National Renewable Energy Laboratory 3
Approach: National Renewable Energy Laboratory 4 Understanding of Internal Short Circuit Through Modeling Perform 3D multi-physics internal short simulation study to characterize an internal short and its evolution over time Expand understanding of internal shorts by linking and integrating NREL s electrochemical cell, electro-thermal, and abuse reaction kinetics models Electro-Thermal Model Temperature Current Density Abuse Kinetics Model T Electrochemical Model soc [%] 62 61.5 soc 61 60.5 60 59.5 150 100 Y(mm) 50 0 0 50 100 X(mm) 150 200 Internal Short Model Study
5 Research Focus Is on Understanding electrochemical response for short Understanding heat release for short event Understanding exothermic reaction propagations Understanding function and response of mitigation technology designs and strategies
6 Heating Pattern Change A multi-physics model simulation demonstrates that heating patterns at short events depend on the nature of the short, cell characteristics such as capacity and rate capability. Heating Pattern at Different Resistance-Shorts 7 Ω Short 30 mω Short Can slow down the evolution? Can suppress? 3 min 6min 9 min Affected by external thermal conditions 4s 8s 12s Undetectable from external probes
7 Understanding Heating Response Differences Heating from Short Circuit = Heat from Cell Discharge + Joule Heat at at Short R o V ocv R s Short Heating = + Global Local Heat for Cell Discharge [ o C/sec] 15 10 5 cell size increase 0 10-4 10-2 10 0 10 2 Short Resistance [Ω] R o = 100mΩ (0.4Ah) R o = 1mΩ (40Ah) Qd M = V R 2 ( R + R ) M o 2 ocv s o Small cell Large cell
8 Understanding Heating Response Differences Heating from Short Circuit = Heat from Cell Discharge + Joule Heat at at Short R o R s V ocv Short Heating = + Global Local Qualitative Representation for Heating Pattern 10 mω Short 10 Ω Short Heat for Cell Discharge [ o C/sec] 15 10 5 cell size increase R o = 100mΩ (0.4Ah) R o = 1mΩ (40Ah) Qd M 0 10-4 10-2 10 0 10 2 = Short Resistance [Ω] V 2 ( R + R ) M o 2 ocv R s o Small cell Large cell Heat at Short [W] 10 4 10 2 10 0 Q s = V 2 ocv ( R + R ) 2 o R s s 10-2 10-4 10-2 10 0 10 2 Short Resistance [Ω]
9 Understanding Heating Response Differences Heating from Short Circuit = Heat from Cell Discharge + Joule Heat at at Short R o R s V ocv Short Heating = + Global Local Qualitative Representation for Heating Pattern 10 mω Short 10 Ω Short Heat for Cell Discharge [ o C/sec] 15 10 5 cell size increase R o = 100mΩ (0.4Ah) R o = 1mΩ (40Ah) Qd M 0 10-4 10-2 10 0 10 2 = Short Resistance [Ω] V 2 ( R + R ) M o 2 ocv R s o Small cell Large cell Heat at Short [W] 10 4 10 2 10 0 Q s = V 2 ocv ( R + R ) 2 o R s s Large cell: localized heating Small cell: global heating 10-2 10-4 10-2 10 0 10 2 Short Resistance [Ω]
10 Understanding Heating Response Differences Heating from Short Circuit = Heat from Cell Discharge + Joule Heat at at Short R o R s V ocv Short Heating = + Global Local Qualitative Representation for Heating Pattern 10 mω Short 10 Ω Short Heat for Cell Discharge [ o C/sec] 15 10 5 cell size increase R o = 100mΩ (0.4Ah) R o = 1mΩ (40Ah) Qd M 0 10-4 10-2 10 0 10 2 = Short Resistance [Ω] V 2 ( R + R ) M o 2 ocv R s o Small cell Large cell Heat at Short [W] 10 4 10 2 10 0 Q s = V 2 ocv ( R + R ) 2 o R s s Large cell: localized heating Small cell: global heating Same local heating, and Negligible global heating for both 10-2 10-4 10-2 10 0 10 2 Short Resistance [Ω]
11 Observations from Literature: Various Short Resistances small resistance short (metal to metal) medium resistance short (bypassing cathode) large resistance short (flow through cathode) John Zhang, Celgard, AABC08 Heat for Discharge Joule Heat at Short *Small Cell with Shut-Down Separator cell shut-down Short Resistance
Observations from Literature: National Renewable Energy Laboratory 12 Various Short Resistances small resistance short (metal to metal) medium resistance short (bypassing cathode) large resistance short (flow through cathode) John Zhang, Celgard, AABC08 Heat for Discharge Joule Heat at Short cell shut-down Short Resistance large resistance short (>5Ω) Temperature (C) 200 180 160 140 120 100 80 60 40 20 0 *Small Cell with Shut-Down Separator Peter Roth, SNL, ATD presentation Delayed Separator Breakdown short With observed No Cell Runaway without (1 thermal amp current) runaway Gen3 Electrodes Celgard Separator EC:EMC\LiPF6 0 500 1000 1500 2000 Time (s) Pos. Term. Temp OCV Voltage 30 25 20 15 10 5 0 Cell Voltage (V) Literature cases with wide range of internal short resistances are observed
13 Prismatic Stack Cell Short Simulation Cu Cu Cu Al Al 20 Ah P/E ~ 10 h -1 Stacked prismatic Form factor: 140 mm x 200 mm x 7.5 mm Layer thickness: (Al-Cathode-Separator-Anode-Cu) 15 µm-120 µm-20 µm-135 µm-10 µm Multi-physics model parameters Electrochemistry model: a set evaluated at NREL Exothermic kinetics: Hatchard and Dahn (1999) Electronic conductivity: Srinivasan and Wang (2003) Modeling Objectives Characterize short natures Predict cell responses Predict onset of thermal runaway Negative Tab Shorted Spot 140 mm Assumptions Short remains same Structurally intact No venting and no combustion Positive Tab 200 mm 7.5 mm
14 Short Between Al & Cu Current Collector Foils Shorted area: 1 mm x 1 mm e.g., metal debris penetration through electrode & separator layers contact between outermost bare Al foil and negative-bias can Cu Al R short ~ 10 mω I short ~ 300 A (15 C-rate) current density field near short electric potential distribution at shorted metal foil layers diverging current at Cu foil converging current at Al foil
15 Short Between Al & Cu Current Collector Foils Joule Heat for Short Temperature @10 sec after short 800 o C surface temperature internal temperature Joule heat release is localized for converging current near short Localized temperature rise is observed. Temperature of Al tab appears to reach its melting temperatures (~600 o C) 25 o C
16 Short Between Al & Cu Foils: Reaction Propagation Exothermic Reaction Heat [kw] 12 10 8 6 4 2 0 0 10 20 30 40 50 60 Time [sec] 10 sec T Q T : Temperature Q: Reaction Heat 20 sec T Q 30 sec 40 sec 50 sec 60 sec T T T T Q Q Q Q
Short Between Cathode and Anode Electrodes Shorted area: 1 mm x 1 mm e.g., separator puncture separator wearout under electrochemical environment potential near short Anode Cathode R short ~ 20 Ω I short ~ 0.16 A (< 0.01 C-rate) Anode Cathode current density field near short Electron current is still carried mostly by metal current collectors Short current should get through the resistive electrode layers Potential drop occurs mostly across positive electrode National Renewable Energy Laboratory 17
18 Short Between Electrodes: Temperature Temperature ( o C) Temperature at Short Cell Average Temperature Temperatures at 20 min after Short 43.1 C 0 200 400 600 800 1000 1200 Time (sec) surface temperature internal temperature 25.5 C Thermal signature of the short is hard to detect from the surface The short for simple separator puncture is not likely to lead to an immediate thermal runaway
19 Observations: Simple Separator Puncture Celina Mikolajczak, Exponent, NASA Aerospace Battery Workshop 2008 Wear and puncture or degradation of separator Local degradation of electrode materials? Issue on Structural Integrity of Separator
Short Between Electrodes: Impact of Short Area National Renewable Energy Laboratory 20 Impact of Separator Structural Integrity Separator Hole Propagation 1 mm x 1 mm 3 cm x3 cm R short ~ 20 Ω I short ~ 0.16 A (< 0.01 C-rate) R short ~ 30 mω I short ~ 100 A (5 C) Myung Hwan Kim, LG Chem, AABC08 3cm x 3cm Separator Hole Exothermic Heat [W] Joule Heat @ cathode layer Joule Heat @ foils Temperature at 1min after short Time [sec]
21 Observations: Structurally Reinforced Separator Y. Baba, Sanyo Electric, PRIME 2008 (214 th ECS) Ceramic coated (one-side) functional separator was tested. Improvements in safety were NOT observed clearly against typical abuse tests. Slight performance improvements were reported. Myung Hwan Kim, LG Chem, AABC08 Mn-spinel based cathode, ceramic coated separator, and laminated packaging provide good abuse-tolerance against typical abuse tests. Nail Penetration Test vs NOTE: An abuse test such as nail-penetration is not likely to represent the process of formation and evolution of internal short-circuits.
22 Rationale: metal plating lowering R s Celina Mikolajczak, Exponent, NASA Aerospace Battery Workshop 2008 Li plating on Anode Surface Short Resistance (R s ) Decrease With increase in plating area and depth Metal plating provides a potential site for low resistance short formation.
Short Between Anode and Al Foil Shorted area: 1 mm x 1 mm e.g., metal particle inclusion in cathode slurry deep copper deposition on cathode during overdischarge Anode R short ~ 2Ω I short ~ 1.8A (< 0.1C) Al Temperatures at 1 hr after short 250 C Temperature ( o C) Temperature at Short Cell Average Temperature surface temperatures Time ( x 10sec) Temperature at short quickly reaches over 200 o C. This type of short is likely to evolve into a hard short in relatively short time. National Renewable Energy Laboratory 23 23 internal temperatures 37 C
24 Observations: Cathode Layer Bypassing Takefumi Inoue, GS YUASA, NASA Aerospace Battery Workshop 2008 Explanation published by SONY and presented by GS YUASA The mechanism of the fire accident on the DELL PC / SONY Lithium-Ion battery published by SONY in Nikkei Electronics, Nov. 6, 2006. Metal particle enters into the triangle zone at the edge of positive electrode where bare Al foil is exposed. Metal particle dissolves and deposits on anode surfaces. Lithium dendrite grows back on the deposited nickel. Low resistance short forms. NOTE: A short formed through or bypassing a resistive cathode layer would result in relatively low resistance short and,highly likely, evolve quickly into a more severe short leading to a safety incident.
Small Cell (0.4 Ah) Short: metal to metal Cu 0.4 Ah cell Shorted area: 1 mm x 1 mm R short ~ 7 mω I short ~ 34 A (85 C-rate) Al Shorted Spot Large cell (20Ah) R short ~ 10 mω I short ~ 300 A (15 C-rate) 40 mm 35 mm 3 mm 140 surface temperature 130 120 110 100 National Renewable Energy Laboratory 25 volume fraction volume fraction 0.2 0.1 0.2 0.1 0 100 110 120 130 140 0 100 110 120 130 140 temperature [ o C]
Shut-Down Separator Li + STOP Large Cell Representation Thermally triggered Block the ion current in circuit Li + STOP Difficult to apply in Large capacity system High voltage system Li + e - National Renewable Energy Laboratory 26
27 Thermal Behavior of a Small Cell 800 244 Even with a small cell, in some conditions, shutdown separator may not function 213 555 359 600 400 200 149 123 without shut-down shut-down functioned 0 0 2 4 6 8 10 12 14 16 18 In some conditions, shutdown separator will function 131 130 126 126
28 Summary NREL performed an internal short model simulation study to characterize an internal short and its evolution over time by linking and integrating NREL s electrochemical cell, electro-thermal, and abuse reaction kinetics models. Initial heating pattern at short events depends on various physical parameters such as nature of short, cell size, rate capability. Temperature rise for short is localized in large-format cells. Electron short current is carried mostly by metal collectors. A simple puncture in the separator is not likely to lead to an immediate thermal runaway of a cell. Maintaining the integrity of the separator seems critical to delay short evolution. Electrical, thermal, and electrochemical responses of a shorted cell change significantly for different types of internal shorts.
29 Future Work Perform in depth analysis for evaluating recommended safety designs such as structurally intact separators and shutdown featured device/strategy in relation to cell design parameters (materials, electrode thickness, cell capacity, etc.) Design experimental apparatus for model validation through the collaboration with other national labs (Sandia National Laboratory) Partner with cell manufacturers and auto industries to help them design safer lithium-ion battery system that appears critical to realize technologies for green mobility
Acknowledgments Vehicle Technology Program at DOE Advanced Battery Research for Transportation program Dave Howell Gary Henriksen National Renewable Energy Laboratory 30