Under the standard procedure for causal discovery using Structural Causal Model, instrumental variables (IV) are mandatory to avoid the problems of endogeneity, bias, and asymptotic inconsistency. However, finding such appropriate IVs is at many times limited, if not impossible, under natural circumstances. 
This paper attempts to provide a practical solution to such problem via introducing an alternative approach, namely, Elastic Segment Allocation-based Two-Stage Least Squares Structural Causal Model (ESA-2SCM) using synthetic instrumental variables for causal discovery.
Together with illustrations on the design and procedures of the ESA-2SCM, empirical verifications are provided with simulations using artificial datasets to test for the effectiveness of the model.