Remaining service life assessment of a reinforced concrete quay structure

Sammanfattning

Harbour structures are critical parts of maritime trade networks, yet their reinforced concrete (RC) elements deteriorate under combined operational loads and aggressive marine environment. Extending their remaining service life (RSL) is of both economic and environmental interest, but no standardised methodology yet unites capacity assessment, reliability evaluation, and deterioration forecasting into a single RSL prediction framework. This thesis developed and applied such a framework to Quay 4 of Oxelösund Hamn, a heavily damaged RC quay, to investigate how the RSL of severely deteriorated RC structures can be reliably predicted. A six phase methodology is proposed that combines data collection and damage assessment, a corrosion model for damage prediction, and two parallel structural assessment tracks whose results are propagated probabilistically. In the numerical track, a 3D finite element (FE) model was calibrated for unknown parameters against eigenfrequencies from on-site measurements. Subsequently, resolved parameters are integrated into a 2D section model that was then run through a central composite design (CCD) simulation series to fit an analytical surrogate model (SM) representing section resistance. In the analytical track, shear and bending limit states based on Eurocode formulas were computed parametrically. Both were evaluated by Monte Carlo simulation (MCS) to obtain the time-dependent failure probabilities. For the representative secondary beam, the numerical track gives a failure probability of 1.24 % for 2026, while the analytical approach, treating shear and bending failure independently, yields 61.64 % for shear and 0.12 % for bending, identifying shear as governing. The time-dependent change in failure probability was determined only within the analytical track. The results show shear risk starting to rise significantly after roughly 40 years of service and predict a 20 percentage point increase starting 2026 over the next two decades. Given the calculation assumptions, these results are conservative, upper-bound limits. However, they motivate spatially resolved corrosion surveys, possible shear strengthening of the secondary beams, and continued monitoring of piles and rock anchors. Methodologically, the parallel-track design proved robust. The analytical approach is regarded as an efficient method for time-integrated analysis, provided the input data is of good quality and the limit states are clearly identifiable, while calibrated numerical SMs better capture interacting, redundant failure modes.