An accurate and prompt prediction of the State of Health (SOH) of lithium ion batteries is a critical process for a Battery Management System (BMS). In this work, a single particle-based degradation model is developed by focusing on Solid Electrolyte Interface (SEI) layer formation and its evolution, which is coupled with the crack propagation due to the developed stress inside particles. The lithium ion loss estimated from the SEI layer evolution is incorporated with an advanced single particle model, where the electrolyte physics is included, for prompt calculation that is critical for online estimation. This model well predicts the measured capacity fade and voltage change as a function of cycle number and temperature. The lithium ion loss due to SEI layer formation, resistance rise, and operating voltage window are examined to explain the capacity fade. Besides its lower-order implementation in a single particle, the results provide quantitative information about the role of SEI layer growth and crack propagation, and corresponding capacity fade and power loss, which are critical for SOH and BMS.