AABC 2014 LLIBTA Symposium - Large Lithium Ion Battery Technology and Application - Track B: Battery Engineering and Application - Session 2

Lithium ion batteries continue to expand in vehicle usage with increasing levels of energy density and the potential for increased safety risk. High profile field events, regardless of the severity of outcome, continue to keep lithium ion battery safety at the forefront. Global standards and regulations are currently undergoing modification and development as the industry works to reach consensus on the most appropriate methods to assure battery safety. The system approach used by General Motors, and others, continues to provide structure to this issue and allow for efficient development of safe battery systems. Some of the latest developments in these areas and how a system approach can be effectively applied will be covered in this presentation.

Accelerated aging protocols are critical for the successful screening of energy storage devices for automotive and other applications. Idaho National Laboratory and Argonne National Laboratory are collaborating to compare the effects of the US- and China-based accelerated aging protocols for electric vehicle (EV) applications using commercially-available lithium-ion cells. The US test protocols are based on the methods established by the US Advanced Battery Consortium (USABC), where cycle-life testing consists of repeating a power-based discharge profile (also called dynamic stress test or DST), followed by constant-current charging. In contrast, the China test procedures employ constant current charge and discharge cycles. Life testing in both the US- and China-based protocols is periodically interrupted for reference performance tests (RPTs) to gauge degradation as a function of aging. For the US-based testing, RPTs are conducted every 50 cycles and include two constant current discharges, a DST capacity measurement, and a peak power test consisting of a 30-second discharge pulse at each 10% depth of discharge (DOD) increment with a constant current discharge in between. For China-based aging, life testing is interrupted every 25 cycles. The RPTs include one constant current discharge and, in some cases, a 10-second discharge pulse at 50% DOD. Results from this study indicate that the US-based aging protocol tended to yield a slightly lower capacity fade but a greater increase in cell resistance. The comparison between US and China-based testing protocols and results will be presented and discussed.

This work was performed under the auspices of the U.S. Department of Energy, Office of Vehicle Technologies, Hybrid and Electric Systems, under Contract Nos. DE-AC07-99ID13727 (Idaho) and DE-AC02-06CH11357 (Argonne).

To date, most serious xEVs have used modestly sized lithium ion cells. One of the reasons is the relative uncertainty of larger cells when subjected to abuse conditions. The attraction of the lower cost per Wh or kg will inevitably encourage some risk taking and use of larger Ah cells. In preparation for wider use of large cells, TUV SUD, a global third party provider of product testing and certification services, has supported cell manufacturers, battery assemblers and OEMs since 2009 with a wide variety of cell abuse testing.  Over this time, TUV SUD has accumulated a considerable of store of data and knowledge about the response of over 1,300 cells to abuse test conditions. This presentation draws from those test results to show the following:

• Trends over time of response severity to selected abuse tests
• Sensitivity of the response to key test and performance parameters
• Tests parameters which have limited influence on response
• Abuse tests which influence cell response most and least

TÜV SÜD, has developed the necessary capabilities to perform all of the various electrical, thermal and mechanical abuse tests identified in all of the current 20 abuse test standards. In this presentation, changes in test methods are also identified which can help to improve the test relevance and produce a more robust product.

Predictive models of capacity and power fade must consider a multiplicity of degradation modes experienced by Li-ion batteries in the automotive environment. Lacking accurate models and tests, lifetime uncertainty must presently be absorbed by overdesign and warranty costs. To reduce these costs and extend life, degradation models are needed that predict lifetime more accurately and with less test data. Models should also provide engineering feedback for cell, pack and system control designs.

Advanced life predictive models are presented in several categories:

• Cell physics life models. At present, physics models are mostly available for only individual mechanisms, however a holistic approach is underway to couple these models for lifetime prediction. New 3D computational models of  electrochemical-thermal-mechanical degradation mechanisms are highlighted
• Cell surrogate life models. Surrogate lifetime models provide a practical approach for warranty analysis, pack sizing and control development. A modeling framework created at NREL has been successfully applied to multiple Li-ion chemistries with good results
• System life models. To predict system lifetime, methods to characterize sources of cell performance and aging variability in a multi-cell pack are presented:

• Cell performance & aging process variation
• Cell-to-cell thermal imbalance in pack
• Pack and passenger cabin thermal response in varying environments

Based on life/degradation models, the presentation also summarizes recent and ongoing studies in low-cost opportunities for battery system lifetime extension:

• Thermal control – passive, active, standby
• Charge control – time of day, rate tradeoffs, varied end-of-charge voltage
• Active cell balancing – accommodating cell performance and aging variability
• Cell electrochemical control –  dynamic maximum voltage limits, optimized charge trajectories
• Prognostic-based xEV supervisory control – real-time optimization of battery performance/lifetime trade-offs

Internal Short Circuit in Lithium ion batteries lead to guild failures that could dampen their widespread use for battery electric vehicles. NREL, in conjunction with NASA, has developed an internal short circuit (ISC) device that can be placed anywhere within the battery and may be used with both spirally wound and flat-plate cells. The internal short device is small compared to other shorting techniques being developed by industry and does not rely on mechanical pressure deforming the battery to activate the short as do most of the other “internal shorts” being developed. The battery can be used and cycled within normal operating conditions without activating the internal short device. This allows for the battery to be aged prior to activation of the internal short. Another unique feature of NREL’s internal short device is that the resistance of the short can be tuned to simulate a hard (more energetic) or soft (less energetic) short. Once the short is activated, the positive and negative components of the battery are internally connected within the cell and internal short circuit begins. NREL’s ISC can simulate all four types of shorts within a cell – collector to collector, collector to anode, collector to cathode and collector to collector. Over the past two years, NREL has implanted their ISC into cylindrical 18650 LiCoO₂ cells to determine the effectiveness of a shutdown separator with regard to ISC type – in particular, we compared a collector to collector ISC versus a collector to aluminum ISC. NREL has also been using the ISC to determine the effectiveness of a non-flammable electrolyte in LiMnO₄ pouch cell. In this presentation, we will provide and update on the effectiveness of the ISC in understanding the behavior of these safety devices incorporated into these two cell types and chemistries.

The emergence of Plug-in hybrid electric vehicles (PHEVs) and electric vehicles (EVs) as a viable means of transportation has been coincident with the development of lithium-ion battery technology and electronics that have enabled the storage and use of large amounts of energy that were previously only possible with internal combustion engines. However, the safety aspects of using these large energy storage battery packs are a significant challenge to address. In addition, advances in battery electrochemistry and electrolyte technologies have made direct comparisons and rankings of safety parameters between cell types and manufacturers a difficult and unscientific process. In this work, we outline several newly developed methods for quantifying the safety aspects of Li-ion battery technologies through experimental tools as well as using sophisticated analytical and computational models.

One potentially hazardous failure mode seen in large lithium-ion systems occurs when flammable and toxic gases are emitted from failing lithium-ion cells, and are distributed into the confined areas within the vehicle where other energized systems or occupants may be present. We have developed a test method that captures and quantifies the composition of gases vented from such batteries. The released gases are then tested in a combustion sphere to quantify their explosivity. Using this information, explosion prevention or protection systems can be designed and optimized for use on battery compartments in PHEVs and EVs. The compositional analysis also aids in quantifying the health effects of human exposure to the vent gases.

Developing thermal management systems for upset conditions in battery packs requires a clear understanding of the heat generation mechanisms and kinetics associated with the failures of li-ion batteries. Using a cone calorimeter, we have developed, for the first time, tests to precisely quantify the heat release rate and the total energy released from cells when they undergo a combustion event. Using these tests, certain safety related parameters of cells with different chemistries can be compared directly in a quantifiable and repeatable manner. Data generated from these tests can also be used in developing and improving computational and analytical models for analyzing thermal runaway and cascading thermal failures in modules and packs.

Thermal management systems can also be designed to detect, prevent and control thermal runaway conditions in battery modules and packs. In particular, the system needs to be able to detect an anomalous temperature rise and be able to cool the failing cells rapidly to prevent the upset condition. Using experimental data from Accelerated Rate Calorimetry (ARC), the exothermic reaction that precedes the thermal runaway process was quantified and used to develop a thermal model of batteries. The resulting model allows for the computational analysis (CFD) of the efficacy of thermal management systems in avoiding upset conditions.

High capacity automotive battery packs used in hybrid and electric cars require robust packaging for both safety and performance.  The battery pack must be properly sealed to protect the battery cells from contamination and loss of electrolyte, leading to premature battery degradation.  Battery leakage also poses risk of damaging components or presenting other safety hazards.

Seal-Scan® Ultrasound Inspection Technology has proven to be a reliable and effective method to inspect high capacity battery packs non-destructively to verify seal and package integrity.  Seal-Scan is a unique air-coupled ultrasound technology that inspects the bonding of package seal materials, detecting channel defects and weak seals.  VeriPac inspection systems deploy a high vacuum pressure decay technology to detect micro leaks, cracks and pinholes in battery cell walls.

Producing reliable and robust high capacity automotive batteries requires testing solutions that identify long term issues with seal quality and micro leaks in the battery cell barrier. PTI’s non-destructive test systems provide unique solutions for assuring integrity and quality of high capacity automotive batteries.

• Measure seal quality for extended battery life cycle assurance
• Detect micro leaks, preventing critical battery failure
• Identify localized seal defects (weaker bond, inclusion, etc.)
• Generate quantitative results with opto-accoustic images
• 100% on-line inspection capability

Both ultrasound and vacuum decay testing methods have proven to be highly accurate and repeatable and offer a non-destructive approach to inspection of both seal quality and barrier integrity of battery cells.  The applications of both technologies extend into other package formats, yet provide crucial answers to the challenging application of battery cell inspection.

Battery performance and life testing is an ongoing program at Argonne National Laboratory. Batteries from U. S. Advanced Battery Consortium and U. S. Department of Energy projects are objectively evaluated according to well defined protocols. Testing provides useful information about battery performance changes with time and, combined with statistical analysis and accelerated aging test conditions, allows forecasting performance over the expected lifetime of the devices. This approach, however powerful, does not provide insights into physicochemical causes of device degradation at materials level. To address this issue, we have established a post-test analysis facility to elucidate physicochemical changes in Lithium-ion batteries responsible for performance degradation.

Electrochemical performance is often limited by surface and interfacial reactions at the electrodes. However, routine handling of samples can alter the very surfaces that are the object of study. Our approach combines standardized testing of batteries with sample harvesting under inert atmosphere conditions. Cells of different formats are disassembled inside an Argon glove box with controlled water and oxygen concentrations below 2 ppm. Cell components are characterized in situ, guaranteeing that observed changes in physicochemical state are due to electrochemical operation, rather than sample manipulation.

We employ a complementary set of spectroscopic, microscopic, electrochemical and metallographic characterization to obtain a complete picture of cell degradation mechanisms. The resulting information about observed degradation mechanisms is provided to materials developers, both academic and industrial, to suggest new strategies and speed up the Research & Development cycle of Li-ion and related technologies.

These poster and talk will describe Argonne’s Post-Test Facility, with an emphasis on capabilities and opportunities for collaboration. Cell disassembly, sample harvesting procedures and recent results will be discussed. In particular, SEM and XPS characterization of Li and Mn rich NMC and Ni-Mn spinel cathodes, both pristine and aged in graphite-anode cells, revealing changes in chemical composition of evolved surface films.

This work was performed under the auspices of the U.S. Department of Energy, Office of Vehicle Technologies, Hybrid and Electric Systems, under Contract No. DE-AC02-06CH11357.

The production of large scale electric vehicle fleets for market is already well established across Europe, North America and increasingly in other areas. One of the most complicated and most costly parts of the production process is the lithium-ion battery fabrication. Individual cells are produced and packaged and are then assembled into battery modules or packs containing many hundreds of cells. Large scale battery cell production, like any production process, has a finite yield of acceptable end products. If a single cell in a battery pack fails then the entire pack must be disassembled and the failed cell removed and replaced. A single weak cell in a module of hundreds seriously affects the output power rendering the whole module useless. Furthermore, the failure mechanisms for Li-ion technology could pose potential hazards. There is therefore a strong safety and economic case to perform quality control testing on all cells before the battery module assembly stage.

For batch production in the chemical industry quality control (QC) is usually achieved via taking a sample from each batch which is tested to ascertain the acceptability of the batch. Since each cell is effectively a batch in the cell production plant the natural QC method would be to sample something in each cell to ascertain whether it has passed QC. Apart from the obvious complexity of a li-ion cell being an impeding issue with chemical analysis QC of a single cell, the cell arrives off of the production line in a sealed packaged state. It is therefore impossible to chemically sample the cell about the cell without materially damaging it. The delicate nature of the cell chemistry also means that such attempts may in fact contaminate the chemistry and thus void the cells QC pass by the very action of performing the QC test.

Clearly a method to perform the QC testing that is non-invasive, does not involve contaminating the chemistry or mechanically altering or interfering with the cell is required. A process of non-destructive testing must be developed whereby the QC test can be achieved without possibility of the test causing the cell to fail yet with strong enough detection rates as to prevent cells likely to fail, either further down the production line or in use after distribution from passing undetected. Equally, since the volume of battery production is large, and is likely to increase with as penetration of EV/HEVs increases, the test must not only be accurate but also timely and economically viable.

This paper describes a targeted method of non-destructive electrochemical analysis using a “black box” approach to characterise cells which are viable and which are not without significant knowledge of the underlying chemistry of the cell. The analysis technique Electrochemical Impedance Spectroscopy (EIS) is used to measure the impedance of the cells and a bespoke processing algorithm is used to analyse the data. The testing method uses empirical data to sort between viable and non-viable cells and a frequency targeted approach to reduce the testing time.

This work was performed under the auspices of the U.S. Department of Energy, Office of Vehicle Technologies, Hybrid and Electric Systems, under Contract No. DE-AC02-06CH11357.