The　increasing　share　of　renewable　energy　in　energy　constituent　requires　the　development　of　large-scale　energy　storage　technologies　to　tackle　with　the　grid　connection　problem,　Carnot　Battery　is　a　possible　candidate.　In　this　study,　the　theoretical　models　of　two　Carnot　Battery　systems　with　four　different　working　fluid　pairs　were　established,　analyzed　and　compared.　Genetic　algorithm　was　introduced　in　the　optimization.　The　results　show　that　working　fluid　pairs　R245fa-HFO-1336mzz(Z)　(the　former　for　ORC,　the　latter　for　heat　pump)　presents　the　best　economic　performance,　reaching　the　lowest　levelized　cost　of　storage　(LCOS)　of　0.2631　$.kWh(-1)　when　the　thermal　storage　temperature　is　130　degrees　C.　The　power-to-power　efficiency　(P2P)　competes　with　LCOS,　with　the　increase　of　P2P　the　LCOS　decreases.　The　working　fluid　pairs　R245fa-HFO-1336mzz(Z)　and　HFO-1336mzz(Z)-HFO-1336mzz(Z)　show　better　exergoeconomic　performance,　the　obtained　optimum　value　of　P2P　is　42.63%.　The　variation　of　exergy　efficiency　(eta(ex))　with　thermal　storage　temperature　is　non-monotonic.　The　working　fluid　pairs　with　the　highest　eta(ex)　is　HFO-1336mzz(Z)-R245fa,　and　the　maximum　value　is　19.88%.　In　each　configuration,　the　most　costing　components　are　turbine　and　compressor.　In　addition,　the　throttling　valve,　compressor　and　heat　exchanger　takes　large　proportions　in　the　total　exergy　destruction.