Service Life Assessment of Harbor Structures - Case studies of chloride ingress into concrete and sheet piling corrosion rates

Abstract: The two most used building materials in harbor structures are undoubtedly steel and concrete. These two materials are often combined in the structures of wharfs and quays where the steel sheet pile walls often have cap beams of reinforced concrete. The degradation processes of these structures must be taken into account both when designing new structures and when inspecting existing harbor structures with the purpose of determining their remaining service life. Since the environmental loads on these structures differ substantially depending on their location, both globally and locally, the degradation processes of the structures also differ. Assessment of these types of structures must therefore combine general knowledge about degradation processes with knowledge about the local conditions. This PhD-project has focused on degradation of steel and concrete structures in harbor environments. The purpose of the work has been to increase the understanding of the degradation processes in order to optimize the design of new structures in marine environment, and to make better predictions of the remaining service life for existing load carrying structures. The results presented in this thesis come from both laboratory studies and from field studies in three Swedish harbors together with a large inventory of earlier performed ultrasonic thickness measurements on sheet pile quays along the Swedish coast. The degradation of concrete structures in the marine environment mainly consists of corrosion of the reinforcement due to chloride ingress. When the chlorides have passed through the concrete cover and reached the reinforcement, the passivating protection of the rebars is lost and the corrosion processes starts. When designing new concrete structures in marine environments, the only ways to increase the length of the expected service life of reinforced concrete structures are to either increase the thickness of the concrete cover and/or to use high quality concrete with a low water cement ratio that makes the concrete less permeable. Another way to secure non corroding rebars is, of course, to use rebars of stainless steel but this is expensive and is rarely used in ordinary construction works today. Field studies in sampling concrete for chloride content analysis was performed in two harbors on the Swedish south coast. The sampling was done by dry drilling and dust sampling. In these studies the local climate, with a focus on the dominating wind direction, has also been studied. The results from the chloride analysis shows that if a surface on a concrete structure is exposed to open sea without any sheltering barriers in front of it, the chloride content tends to be much higher than in sheltered parts of the structure or in structures facing other directions. This is true irrespective of the dominating wind direction in the area. Concrete slabs were exposed to saline water in laboratory. When the slabs had been exposed for about seven months, concrete sampling was performed with the purpose of testing different sampling methods. Both core drilling with a 100 mm core and dry drilling with different drill diameters was performed on the slabs. The cores were grinded and the dust from both grinding and dust sampling by dry drilling were analyzed with respect to chloride content. The results showed that dry drilling with small diameters collecting mixed samples from several nearby drill holes gave almost the same result as grinding the 100 mm core with respect to chloride content in percent by mass of CaO. The chloride content in a sample from one of the exposed slabs was mapped with EPMA (Electron Probe Micro Analyzer) with the purpose of testing another sampling method. The result from the EPMA analysis was used to simulate drilling with different drill diameters and to compare these results with the results from the sampling made in the laboratory. The results from both dry drilling and the EPMA measurements showed that if the chloride content is presented as percent by mass of CaO instead of as percent by mass of concrete, the variations in chloride concrete were substantially decreased. Chloride data should therefore be presented per mass of CaO or per mass of cement. The requested service life of steel structures in marine environments such as harbors is usually achieved either by over-dimensioning the steel thickness and assuming a certain even and constant corrosion rate in mm/year, or by applying a protective coating such as a paint on the structures to prevent corrosion. Car ramps and other steel structures that are not in direct contact with salt water are often coated with protective paint to prevent corrosion, while structures like sheet pile walls which are in direct contact with sea water most commonly are unprotected. A third way to protect wharfs from corrosion is to use cathodic protection with sacrificial anodes or by an impressed current cathodic protection system. Cathodic protection is common on ships and sheet pile walls. A field study measuring the remaining steel thickness on existing quays was performed in the harbor of Halmstad on the Swedish west coast. An inventory of data from earlier measurements of steel thicknesses in harbors along the Swedish coast was also performed in this work. The purpose of these studies was to investigate whether the recommended design values on corrosion rates in the existing construction codes are accurate. In this study, corrosion rate design values from USA, Australia, Europe (Eurocode) and Sweden were compared with the results from measurements in Swedish harbors. The results from the corrosion rate measurements show that the measured corrosion rate is generally lower than the recommended Swedish design values, but in the same order as the recommended design values given in the European design code for sheet piles in marine environment. With the purpose of investigating the corrosion loss on steel in a marine environment in a controlled way, steel plates were exposed to a marine environment for almost one year. The plates where mounted on ropes at different depths at three sites in the harbor of Halmstad. One of the exposure sites was located in the river of Nissan (fresh water) which has its outflow in the sea of Kattegatt. The two other sites were located inside the docks. The purpose of this study was also to determine if salinity affects the corrosion rate since the salinity in the fresh water river is generally lower than in the docks. The result from this study showed that salinity (and pH) did not influence the corrosion loss in the short time perspective (about 50 weeks). The corrosion losses on the exposed plates was almost three times as high as the measured corrosion losses on the existing sheet pile walls in the same harbor. This result suggests that the corrosion rate is non-linear in the short time perspective with the highest losses in the beginning of the exposure. This agrees with recent models of steel corrosion in marine environments. However, for harbor structures with 100 year service life, the thickness decrease can be modelled as being proportional to time. The results from the present studies have implications both for the design of new harbor structures and for the assessment and maintenance of aged harbor structures. Increased knowledge of sampling procedures and improved degradation models make it possible to make more precise predictions of remaining service life for existing structures and will lead to more accurate design values for new steel and concrete structures in marine environments. Periodical inspections could be used for more accurate predictions of remaining service life of a given structure, for example using the method of Bayesian updating.