The development of robust damage detection methods for offshore structures is crucial to prevent catastrophes caused by structural failures. In this research, we developed an Improved Modal Strain Energy(IMSE) method for detecting damage in offshore platform structures based on a traditional modal strain energy method(the Stubbs index method). The most significant difference from the Stubbs index method was the application of modal frequencies. The goal was to improve the robustness of the traditional method. To demonstrate the effectiveness and practicality of the proposed IMSE method, both numerical and experimental studies were conducted for different damage scenarios using a jacket platform structure. The results demonstrated the effectiveness of the IMSE method in damage location when only limited, spatially incomplete, and noise-polluted modal data is available. Comparative studies showed that the IMSE index outperformed the Stubbs index and exhibited stronger robustness, confirming the superiority of the proposed approach. The development of robust damage detection methods for offshore structures is crucial to prevent catastrophes caused by structural failures. In this research, we developed an Improved Modal Strain Energy (IMSE) method for detecting damage in offshore platform structures based on a traditional modal strain energy method The most significant difference from the Stubbs index method was the application of modal frequencies. The goal was to improve the robustness of the traditional method. To demonstrate the effectiveness and practicality of the proposed IMSE method, both numerical and experimental studies were conducted for different damage scenarios using a jacket platform structure. The results demonstrated the effectiveness of the IMSE method in damage location when only only, spatially incomplete, and noise-polluted modal data is available. Comparative studies showed that the IMSE index outperformed the Stubbs index and showed stronger robustness, confirm ing the superiority of the proposed approach.