Flare minimization under normal and abnormal operating conditions in large-scale industrial process, especially for refineries and petrochemical plants, is a double-win practice, which simultaneously benefits industrial and environmental sustainability. Unfortunately proactive and cost-effective flare minimization (PCFM) approaches under abnormal situations are still lacking. In this study, a PCFM approach for an ethylene plant is presented for its start-up operations. This approach employs rigorous steady-state and dynamic models of a front-end deethanizer ethylene plant to serve as a foundation to explore flare root causes during plant start-ups and subsequently a test bed to support both design and operational strategies for start-up flare minimizations. It has been demonstrated that the charge gas compressor (CGC) start-up is the most critical operation, which results in the largest amount of flaring sources. Several start-up strategies at different CGC feed rates and compositions, including scenarios recycling off-spec C 2s and C 3s from downstream recovery units to the CGC inlet as a substitute of a portion of charge gas, are virtually examined to identify the least flaring one. This work contributes to evaluating new start-up strategies, predicting abnormal process dynamic behaviors, and acquiring more precise estimation of flare emission sources. The study shows that the plant flaring can be significantly reduced, while the start-up time is shortened