Decay heat removal in a nuclear plant is very important, and the performance of a decay heat removal system in a plant is a critical factor for the plant safety. In designing the decay heat removal system, a passive-type system is usually more difficult than an active-type system, and there can be additional restrictions in designing plant systems for passive decay heat removal to secure a sufficient natural-circulation head. If one can devise a decay heat exchanger that can enhance buildup of the natural-circulation head during an accident, the restrictions on designing the systems related to the decay heat removal can be relaxed and a better plant design can be attained. To meet this necessity, a design concept of an improved decay heat removal heat exchanger, IDINHX, was devised for a pool-type liquid-metal reactor (LMR). Its performance was evaluated, and the physics related to the core cooling in a pool-type LMR was investigated. During an accident, the core exit temperature usually peaks twice. The first peaking reflects the early-phase cooling capacity of a system, and the second peaking reflects the late-phase or long-term cooling capacity. The physics of the first peaking are more complex than that of the second peaking and, consequently, designing against the first peaking is more difficult. Based on the investigation results, ways to control the first peaking are suggested.