In　this　research,　a　method　of　developing　a　model　for　predicting　the　energy　saving　potential　of　regional　residential　buildings　has　been　presented.　This　model　is　predicated　on　a　bottom-up　methodology　in　representing　and　understanding　the　physical　properties　of　the　residential　buildings　in　the　region.　The　Residential　Energy　Consumption　Survey　(RECS)　database　is　taken　advantage　to　provide　modeling　inputs　for　the　building　modeling　of　the　five　developed　archetypical　building　types.　Different　secondary　energy　supply　systems　are　then　distinguished　and　modeled　by　a　developed　lightweight　building　simulation　tool　-　SimBldPy.　The　model　is　calibrated　and　validated　based　on　energy　use　data　from　1993　to　2015.　It　is　It　is　used　for　the　optimization　of　eleven　selected　energy　conservation　measures　(ECM)　to　find　the　best　potential　scenario　for　equivalent　primary　energy　for　residential　buildings　in　New　York　state　(NY).　Best　combinations　of　ECM　parameters　are　discerned　for　each　building　type　by　Sobol　sensitivity　analysis.　The　optimal　regional　energy　use　scenario　is　found　after　coupling　the　population　projection.　It　is　indicated　that　the　saving　potentials　of　the　state-level　primary　energy　use　for　different　building　types　range　from　48　%　to　62　%　compared　with　the　baseline　scenario　at　the　end　of　2040　in　NY.　©　2020　Elsevier　Ltd