Quantifying facility-level methane emission rates using satellites with fine spatial resolution has recently gained significant attention. However, the existing quantification algorithms usually require the methane column plume from a single point source as input. Such approaches are challenged with overlapping plumes from multiple point sources. To address these challenges, a multi-objective heuristic optimization algorithm is introduced to perform parameter estimations for the 2D multisource Gaussian plume model, which serves as the basis for the separation method. In addition, to improve the separation performance on relatively weaker sources, we proposed a metric called local binary pattern metric (LBPM), which is only sensitive to the sign of the gradient as a minimization objective. To verify the proposed separation method, observation system simulation experiments (OSSE) of various scenarios are performed, where the integrated mass enhancement (IME) is selected as a representative single-source quantization method. The result shows that plume overlapping will increase the quantifying error of IME as overlapping pixels may not be attributed correctly; compared to unseparated overlapping plumes, the proposed separation method decreases the quantification MAPE from 1.46 to 0.45 on synthetic observation over real targets. Our separation method can separate observation of overlapping plumes from multiple sources into several observations each with a plume from a single source, thereby extending single point source quantifying algorithms, such as IME, to be applicable within scenarios of multiple point sources.