Proof load testing (PLT) offers a valuable and sustainable alternative to analytical approaches for improving knowledge on the safety level of existing bridges, providing an in-situ measurement of structural bearing capacity under actual traffic loads by reducing resistance uncertainties and associated probability of failure if the test is passed. The present paper investigates the influence of the PLT on the structural reliability of prestressed concrete I-type simply-supported decks representing the most common type of existing bridges in Italy. By supplying data on the lower-bound of the capacity distribution, the PLT turns into an updated estimation of the bridge reliability. A fully-probabilistic analysis is developed combining random uncertainties on both materials and load effects with epistemic uncertainties. A traffic load model variable based on Eurocode Load Model 1 effects is calibrated to provide consistent modelling with code-prescribed safety levels. Structural capacity of the edge girder is considered both in terms of ultimate limit state for flexure and shear and serviceability limit state in terms of cracking load which could affect long-term bridge durability. The manuscript main contribution lies in developing a reliability-based approach to PLT that accounts for both prior (before test) and posterior (after test) structural reliability, incorporating conditioning on the success of the test. A sensitivity analysis according to the partial safety factor method is presented to investigate the impact of different proof loads assuming different Capacity-to-Demand Ratios (CDR). A case-study bridge is investigated where a proof load was executed recently demonstrating the benefit of the PLT in case of CDR lower than unit. The case study also showcases the possibility to significantly reduce the failure probability during the test when the target level is imposed with a number of intermediate levels of load steps.

Structural assessment of existing bridges is one of the most challenging aspects for the correct and sustainable management of civil infrastructures that emerged in the last few years. Specifically, Italy has a substantial number of bridges over 50 years old, mainly represented by simply supported prestressed concrete I-girder-type with a cast-in-situ deck. Proof load testing represents an empirical alternative to the standard calculation methods commonly adopted for safety assessment of existing structures. This article investigates the reliability of a bridge in different configurations including damaged and undamaged scenario through sitespecific traffic data collected by Weigh-In-Motion (WIM) systems in the Netherlands. Assuming a target load corresponding to characteristic load combination according to Eurocode provisions, the proof load test resulted into slightly reduced reliability indexes during the test while achieving higher values under successful completion. Preliminary results contribute to demonstrate proof load test can represent a valuable method for assessment of existing bridges.