This　paper　proposes　a　stochastic　transmission　switching　integrated　interval　robust　chance-constrained　(TSIRC)　approach　to　assess　the　operation　of　a　wind　park-energy　storage　system　(WPES)　in　a　day　ahead　electricity　market　considering　the　system　technical　constraints.　To　this　end,　the　WPES　is　assumed　as　a　price-maker,　making　it　possible　to　alter　the　market　price　based　on　its　own　profit.　The　problem　formulation　is　constructed　as　a　tri-level　max-min-max　structure　during　a　24-h　time　horizon.　The　first　level　maximizes　the　hourly　profit　of　the　WPES.　In　the　second　level,　the　system　operation　cost　is　minimized　in　the　form　of　a　security　constrained　unit　commitment　(SCUC)　in　which　the　contingency　effect　associated　with　the　generation　units　and　transmission　lines　(TLs)　as　well　as　the　congestion　of　the　TLs　is　modelled.　Considering　the　independent　system　operator　(ISO)　preferences,　the　third　level　is　dedicated　to　the　robustness　of　the　WPES,　which　would　maximize　the　allowable　output　variation　of　the　wind　turbines.　The　proposed　tri-level　model　is　then　formulated　in　an　efficient　bi-level　structure　using　the　Karush-Kuhn-Tacker　(KKT)　conditions.　Since　the　operation　of　the　wind　turbine　is　inherently　associated　with　uncertainty,　the　TSIRC　approach　as　an　effective　tool　is　able　to　model　the　uncertainties　of　the　wind　units　by　increasing　the　WPES　profit　as　a　strategic　producer　while　reducing　the　system　operation　cost.　In　addition,　unscented　transform　(UT)　as　an　effective　tool　is　considered　to　model　the　uncertainty　of　the　stochastic　parameters.　The　performance　evaluation　of　this　work　is　assessed　on　an　IEEE　test　system.　Different　case　studies　are　provided　which　show　the　authenticity　and　effectiveness　of　the　proposed　model.　The　simulation　results　show　that　the　proposed　TSIRC　has　effectively　reduced　the　amount　of　lines’　switching,　and　increased　the　power　output　of　the　wind　park　by　95%　on　average　compared　to　the　conventional　TS-based　model.　In　addition,　the　robustness　of　the　system　has　increased,　and　the　TSIRC　has　led　the　WPES＇s　profit　to　increase　by　almost　11.5%　compared　to　the　TS-based　approach.　©　2020　Elsevier　Ltd