Durability management and planning for concrete structures in the Hong Kong–Zhuhai–Macau link: data-driven and model-based approach

As the world’s longest sea-crossing project, the Hong Kong–Zhuhai–Macau (HZM) bridge is designed for a service life of 120 years—a milestone in infrastructure engineering that presents unprecedented challenges in long-term durability assurance. To safeguard the structural integrity of this large-scale concrete infrastructure against aggressive marine environmental exposure over its whole design life, a dedicated research team has developed an innovative, data-informed framework for intelligent durability management and life-cycle planning.

Their research was published in Civil Engineering Sciences in September 2025.

The objective of this work is to mitigate chloride-induced corrosion of embedded steel reinforcement—a dominant degradation mechanism governing the service life of marine concrete infrastructure. In contrast to conventional approaches, which depend on periodic, reactive inspection and localized repair interventions, the proposed framework enables proactive, prediction-enabled structural health management grounded in real-time condition assessment and durability forecasting.

“To achieve a 120-year service life, we must move beyond traditional maintenance models. We have developed an interactive, intelligent framework that systematically integrates multi-source data acquisition, mechanistic durability modeling, and life-cycle integrated maintenance planning. This system supports proactive prediction and performance-based management of long-term structural behavior—shifting the paradigm from reactive defect remediation to anticipatory durability assurance.” said Kefei Li, corresponding author and Professor in the Department of Civil Engineering at Tsinghua University.

The framework comprises three integrated components. First, an intelligent, self-cleaning and self-predictive database that fuses heterogeneous data from long-term exposure testing, continuous sensor-based monitoring, scheduled visual inspections, and meteorological records. Second, a modular durability model library—calibrated for diverse marine exposure scenarios—that enables precise prediction of chloride ion ingress kinetics under variable environmental conditions (e.g., tidal zone vs. splash zone) and structural configurations (e.g., cracked concrete, coated surfaces, stainless bars and corrosion inhibitors). Third, a tiered, performance-driven maintenance strategy, systematically derived from model-based degradation predictions and life-cycle cost optimization.

Quanke Su, Engineer-in-chief of the HZM bridge, said,“The 120-year design life has been one of the most central technical challenges since the project’s inception. This data-driven management system is a crucial innovation. It transforms long-term maintenance from a set of fixed rules into a dynamic, intelligent process, providing a robust guarantee for achieving this ambitious goal and setting a new benchmark for future mega-projects.”

According to the system’s assessment, the failure probability of the bridge’s key structural elements is currently far below the 1% intervention threshold, indicating ample safety reserves for durability. The proposed basic maintenance plan, such as applying silane impregnation in the 60th and 90th years of service, will cost-effectively ensure its 120-year design life. This systematic solution not only provides a solid guarantee for the HZM Bridge’s century-long service but also offers a replicable paradigm for durability management in similar marine engineering projects worldwide.

The research was published in Civil Engineering Sciences, a high-quality journal under the Chinese Science and Technology Journal Excellence Action Plan. The team will continue to optimize the model library and database, promoting smarter and more precise durability management for long-span marine concrete structures.

Other contributors include Ze Yuan, Pengfei Ma, Quanwang Li, Yi Zhang, Le Li, Junjie Wang from the Department of Civil Engineering, Tsinghua University, Beijing, China, Qingfa Wu from the Hong Kong–Zhuhai–Macau Bridge Administration (HZMBA), Zhuhai, China, and Quanke Su from the Hong Kong University of Science and Technology, Guangzhou Campus, China.

This work was supported by the National Natural Science Foundation of China (52038004) and the National Key R&D Program of China (2019YFB1600702).