An often underestimated threat to our economy is steel corrosion: in the western world, damage by corrosion is estimated at 4% of the GNP, and approximately 5 tons of steel per second is lost through corrosion. In the Oil and Gas Industry (North Sea production platforms) 60% of all maintenance costs are related to corrosion, directly or indirectly (1993). 90% of ships failures are attributed to corrosion (Melchers, 1999).. Consequently, the costs pertaining to corrosion are sky high: a 2006 study indicates that the US Navy alone incurred 2.44 billion dollars’ worth of damage due to corrosion; for the U.S. marine shipping industry, the annual corrosion-related costs were estimated at $2.7 billion. For the whole U.S. economy, the 1998 cost of corrosion amounted to $275.7 billion in 1998 alone (see overview on the cost of corrosion in De Baere et al. 2013). With our economy relying strongly on maritime transport and the construction of offshore platforms for (deep) sea mining, damage to these structures should be limited as much as possible. But the consequences of corrosion are not limited to the purely economic aspect. Steel is a crucial resource for our technological society and even the most advanced present day steel producing technologies still rely on one of the most polluting fossil fuels – cokes. While it is imperative to continue to look for sustainable ways to produce new steel, prevention of steel losses by slowing down the corrosion process is equally important. Technologies to prevent corrosion are commonplace: anodic or cathodic protection, special coating layers, steel alloys that enhance corrosion resistance,… However, the missing step is a comprehensive, quick and easy method to estimate when the metal endures enhanced risk for accelerated corrosion. The identification of such periods helps in selecting the most appropriate preventive measures. In that way, periods of accelerated corrosion can be avoided in the future and the lifetime of metal can be prolonged. . In this project, we will focus on the corrosion of steel submerged in water such as wind turbines at sea, water desalination or sanitation installations, ship’s hulls, harbor pile sheets, bridges, quays, wharfs, jetties, etc. <!--break--> To assess and predict the occurrence of corrosion, we want to divert from the traditional methodology and avoid searching for the causal mechanisms. Corrosion is a complex phenomenon, driven by abiotic chemical causes (temperature, salinity, oxygen concentration, …) and modulated by the activity of a multitude of bacterial consortia. Trying to untangle this web of possible (co-)mechanisms has been a challenge for the last 2500 years (at least) and never resulted in a satisfying answer. Instead of investigating factitive factors, we will focus on identified environmental markers - parameters that are able to estimate accelerated corrosion. It is known that parameters such as the chloride concentration, oxygen concentration or redox potential of water affect in some way the corrosion rate of metal. One can assume that a sudden increase of one of these parameters, might result in an enhanced corrosion rate. Depending on the type of water (sea water, brackish water, drink water) the importance of these markers in the estimation of the enhanced risk can be different. This project will focus on the identification of these markers, their weights and the model to determine the risk for corrosion for different types of water. The first set of markers will be the type of the environmental water, and the concentration of each of the (major) constituents (chloride, oxygen, nitrate, phosphate, sulfate,…). Seawater, fresh- and brackish water, reusable wastewater, recycled condensates are all alternatives to be tested. Secondly, there will be salinity in itself, as well as pH, temperature, the concentration of oxygen dissolved in the water. In addition, we will change the redox potential of the surrounding water by adding oxidants (commonly used in chemical water treatment and disinfection). Other markers to follow comprise all forms of dissolved inorganic nitrogen (DIN), the bacterial load, the speed of the water, the presence of gases like H2S and methane. Evidently, the rate of corrosion itself will also be monitored. Current tests usually consist of coupons of different steel alloys that are brought into contact with the feed stream to track corrosion effects visually. However, these traditional tests provide time-averaged information for long immersion times. Therefore, we will use a combination of typical steel coupons (analyzed by the weight loss method) on the one hand, and the use of electronic resistance probes (e.g., from the company Cosasco) to monitor the continuous corrosion behavior. By applying data mining on the environmental markers in combination with the real time behavior of metals it should be possible to determine the impact of certain markers on the degradation rates and the relative weights of the markers. A similar approach has been successfully applied on estimating indoor air quality inside museums. Statistical analysis will reveal the relative importance of the different markers in relation to the onset or rate of corrosion. Using algorithms developed for a similar project by one of the researchers involved, the outcome of this analysis will be translated into an easily interpretable scorecard, useful for interested companies.