Download or read book Development of Reduced-order Models for Fire Hazards in Battery Energy Storage Systems written by Serhat Bilyaz and published by . This book was released on 2020 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Lithium-ion batteries are being increasingly used in many applications, including Energy Storage Systems (ESS). Although there are many planned installations for these systems in the near future, the failure paths and consequences of these systems are still not well understood. Understanding and developing protection systems for fire hazards in large scale ESS applications through experiments is either very expensive or infeasible. Modeling of large scale systems by using high-fidelity tools such as CFD or FEM, while useful, may be computationally infeasible when large numbers of predictions are required as might occur in some design or characterization studies. This work aims to develop physics-based, reduced-order models that can supplement more expensive characterization approaches to analyze fire hazard scenarios that might occur in ESS applications. The modeling effort is performed at four physical scales: building, room, cell array, and single cell. First, a transient resistance network model is developed to characterize fire smoke transport in high-rise buildings. If a fire, whether from a battery or otherwise, occurs in a high-rise building, decisions must be made on the application of engineered smoke management system operation and guidance for occupants on evacuation. Models that can predict the transport of the toxic products of fire are critical to making better decisions. In general, a building scale fire starts in a room. When lithium-ion batteries fail, they vent flammable gases. For a fire to start, these gases must encounter an ignition source before gas sensors in the room or HVAC system detect the failure and initiate the fire and explosion protection system. At the room-scale, the dispersion and mixing of buoyant gases in a closed room is considered. This problem is critical to estimate the probability of ignition after a battery system fails. A reduced-order model is developed, which combines the integral non-Boussinesq plume model with 1D advection-diffusion transport of species in the room. The model uses the method of moments to predict the distribution of gas concentration in the room. Proceeding to smaller scales of failure, it is useful to understand how a given cell failure may result in the cascading failure of an array of cells. The volume rate of production of flammable gases is proportional to the cascading failure rate. Towards understanding this problem, cascading thermal runaway propagation between the cells is analyzed. A 1D finite-difference model is developed to solve the coupled reaction kinetics and heat transfer problems. The simplicity of the model enabled the examination of many thermal runaway reaction mechanisms that are proposed in the literature. A global one-step reaction kinetic mechanism was calibrated using experimental results and laminar flame propagation theory. Thermophysical parameters for the multiple cell failure were determined using single cell failure data. Reduced-order models for a single cell were useful in the calibration work. The reduced-order models are validated by either using the available experimental data or developing higher-fidelity model alternatives. For the building scale model, experimental data on stack effect was taken from the literature, and the fire investigations are validated by using an open-source fire computational fluid dynamics code, Fire Dynamics Simulator (FDS). For the room-scale dispersion problem, an FDS model is developed as a benchmark case for the reduced-order model. For the cell array scale problem, experimental data were used for validation. In summary, this dissertation develops and evaluates fast reduced-order models that can be used to characterize fire hazard scenarios over a range of scales in battery energy storage systems