In this research, an optimum equation is proposed for displacement estimation at the key point and key zone of powerhouse cavern sidewall in elasto-plastic condition based on the numerical experiments. Moreover, a new predictive equation is suggested to predict occurred displacement at the key point using the gene expression programming (GEP). Based on the numerical analyses (NA), the location of key point on the reciprocal longitudinal walls of the powerhouse cavern is studied. It was specified that for cavern with 33x52 m cross section, key point might be placed on the rock sidewall (left wall) or on the pillar side (right wall) of cavern. On the other hand, for cavern with 18x30 m cross section in various geo-engineering conditions, the key point was located on the rock sidewall. Verification of the proposed equations (from NA and GEP) were conducted using some case studies measurements which proved a good agreement of the new equations with the real values and showed the higher accuracy compared to the similar available equation. Finally, using warning levels of tunnel stability loss proposed by Sakurai, a new a system classification scheme is proposed for key point surrounding the powerhouse cavern based on the critical displacement.
