We arrived at the Royal Eijkelkamp to discuss the SOCORRO findings. Gathered around a table for lunch, the reliability of different probes and sensors was at the core of the conversation, from the wiring being inverted to updates with the micro-chips of the sensors, slowly the conversation turned towards travelling and traffic and how difficult it was to get here.
Ready to start the presentations were Geert, Marjolein, Erik (from Eijkelkamp), Jeroen, Erik (C-Cube), Arne, Brent, Martin, and Wikke.
Geert opened the event with an overview of the scale of the corrosion issue which globally costs ca. 2.3€ trillion per year; with every 90 seconds, 1 ton of steel rusting away in the world. Marjolein discussed corrosion and climate change, and the impact it has on water quality, management and treatment. Diversifying water sources, increasing periods of droughts and instances of flooding as well as changing water demands will all affect corrosion behavior. You can have a look at the presentation here.
Whilst it's easier to collect data in the lab, to tackle the challenges ahead it's important to obtain data from different sensors and demonstration sites. Corrosion happens in the wild, and only when sensing real world scenarios can we truly learn to account for its variety and the uncertainties of the scenarios where corrosion happens. Also, we need to account for what grows on the sensors: cockles, clams, crabs, fouling (sludge, bacteria) for these have agentic capacities and can make the sensors stop giving accurate measurements or working altogether. Maintenance, cleaning, and recalibration isare essential to collect accurate data.
Erik from Eijkelkamp (Aquaread AP800 sensor) and Erik from C-Cube (LPR sensor) presented details of their respective sensing systems along with the research that they are currently doing. We sincerely thank them for their contribution to the SOCORRO project and for giving us access to their sensors to use in sea water, brackish water, boats, and other rather harsh environments where it was difficult to keep the sensors clean and working efficiently.
Jeroen explained the latest developments of the SOCORRO AI system and how they have been training the system using data from lab experiments and from "in the wild" demonstration sites, with an emphasis on how difficult it has been to get real world scenario data for this purpose, and what is still to be done to make the corrosion risk that is presented in the application aligned with what is happening in those types of sites.
Arne explains mathematically and statically that corrosion is both a forward and backwards reaction! Arne and their team are more interested in the kinetic connection than the thermodynamic, but it's important to note that all the different parameters affect the rate of corrosion, so for instance increase in temperature eventually tends towards a lower rate of corrosion because of the lack of dissolved oxygen in the water while the speciation of the iron present in the water can either increase or decrease the corrosion rate. The moral of the story is that things are so interrelated that we have to take into account how they influence and affect each other which reminds us of Callon's scallops (1986), and of course Latour's good old Actor-Network-Theory! The statistical analysis presented was related to lab experiments rather than real world scenarios yet the knock-on effect of parameters on other parameters is daunting. One thing is clear though; salinity and temperature are some of the most important factors influencing the corrosion rate.
Brent discussed also that when measuring corrosion in wastewater treatment environments, things start looking weird: Why are the sensor measurements different from the steel coupons which we have analysed for mass loss? Well, there probably are some little creatures happily living in that water. And what is also very clear, is that materials often used in these environments are susceptible to pitting.
Following from this, Martin carried on the discussion in this area by presenting findings from a port environment and how we can measure such sites. What can be measured so that the serious corrosion that is happening can be monitored and prevented to some degree? Perhaps infrared infra-spectroscopy FTIR and x-ray photoelectron spectroscopy can help us understand the issue better, in particular microbial-induced corrosion. What is the involvement of sulphate in the corrosion, and the bacteria that live off it? Fieldwork data (or in our case demonstration site data) is messy, never as neat and perfect as data from controlled lab environments, yet this is where the real impact of our research can be measured.
Now to wrap up the day of presentation, Wikke presented the very impressive and challenging demonstration site inside a ballast tank, which despite having several protective coatings and materials selected to counteract corrosion, still exhibits corrosion. After three months, we canould already see some creatures growing on and fouling the sensors, causing problems with the measurements and with gathering of the data.