Development of a Computational Model for the Corrosion of Steel in Nigerian

Coastal Region and Offshore Environments.

1.0 Introduction
The development of materials has always been to improve the material properties for better
durability in application. Important properties of materials that are usually the target of
materials engineers include mechanical properties like strength, toughness, etc., physical
properties such as densities and natural frequencies, and chemical properties like corrosion
resistance. The corrosion resistance of a material is of particular importance because of the
catastrophic damage it can cause to structures in industries such as in oil and gas, chemical
processing and nuclear fields, with the cost of corrosion in industrialized economies like the
US and China exceeding $500 billion per annum (Hou et al., 2017; Koch, 2017).
The method of developing materials with improved properties has mostly been through
empirical effort, resulting from a high degree of either trial and error, laboratory testing,
serendipity, incremental advances, and field experience (Taylor et al., 2018). The possibility
to predict the properties and performance of materials from developed computational models
has grown dramatically with the advances made in the computational powers of hardware and
software systems. Unknown material properties can be predicted using models developed from
materials and constituent materials properties as basic as their atomic numbers, dipole
moments, ionization potentials and other quantum chemical parameters (Dong et al., 2021).
This has become a powerful tool for establishing numerical principles and the discovery of
advanced materials. The use of these techniques will greatly help in developing corrosion
resistant alloys through the ability to predict the nature of the response of the material in
different corrosion environments by testing millions of material compositions in dramatically
minimal time and cost.
Many computational techniques, tools and systems for predicting corrosion have emerged such
as Questeck, Materials Project, AFlow, MatNavi, Open Quantum, Materials Database, etc.
(Dong et al., 2021). These tools and systems provide a means for innovating materials
composition and structure based on calculated E-pH diagrams, work function value or band
gap data (Jain et al., 2013; Saal et al., 2013; Calderon et al., 2015).
It has been shown by Gong et al. (2020), Shi et al. (2018), Zhi et al. (2020) and Ser et al. (2020)
that Artificial Intelligence can be used to mine data from research works that are not directly
for corrosion to develop new materials with enhanced corrosion resistance. Computational
material methods, therefore, make it possible to use data that might not be directly linked to
corrosion to understand the nature of corrosion in potential materials.
Obot et al. (2015) has reported that computational modeling can be used to provide useful
reactivity parameters that can be used to explain the corrosion inhibition effect of compounds
before they are synthesized and tested.
Sastri et al. (2005) used a number of quantum chemical parameters which include dipole
moment, ionization potential, electrophilicity, etc. to rationalize the corrosion inhibition
efficiency of several organic inhibitors.
Different environmental conditions need models that are unique to them and, being a field of
study that is just beginning to gain unprecedented attention due to big advances made in the
field of computing, it is imperative that a nation like Nigeria, whose major source of foreign
exchange and revenue is oil, explore this field and delve into the task of developing corrosion
models that are unique to its oil and gas operations environment. It is in consideration of this
that I am embarking on this research.

2.0 Aim and Objectives

This study aims to develop a model that predicts the corrosion behaviour of steel and its existing
and potential alloys in the Nigerian coastal region. The objectives are:
1. To harvest data from Open Quantum, Materials Database, MatNavi and Questec
relating to steel and its chemistry.
2. To use computational data analysis and established models and algorithms to clean and
test various possibilities.
3. Created different test models based on the environmental conditions of Nigerian coastal
regions and sea offshores.
4. To test different steel materials at different combinations of conditions in the coastal
and offshore environment.
5. To create a master model for the prediction of corrosion behaviour.
3.0 Materials and Methods
The materials needed for this research include:
1. A computer system, at least 7 core CPU and 16GB RAM.
2. Data harvesting software to include LabView and Matminer
3. Data analysis software including Excel, SPSS, and Stata
4. Steel material and its alloys
5. Corrosion testing equipment
6. Seawater from Nigerian offshore fields.
The methods that will be followed in the following order include:
1. Data harvest:
This will be done using data harvesting software and algorithms. Data will be harvested from
online materials databases and findings of other research works.
2. Data cleaning
3. Quantum chemical calculations
4. Model development:
Both linear and non-linear machine learning methods for the prediction of corrosion will be
evaluated. The linear method to be used in this study is the partial least squares (PLS). This is
a multivariate linear method normally utilized to study the relationships between two sets of
dependent and independent variables.
5. Model validation:
To validate the developed model, stratified and random splitting will be used. These are for the
external validation of the developed models. For the internal validation of the model, Leave
One Out Cross Validation (LOO-CV) will be used.
6. Laboratory Experimentation:
Practical laboratory experimentation will be done using randomly selected variables to
test the practical performance of the developed model.

References
Calderon, C. E., Plata, J. J., Toher, C., Oses, C., Levy, O., Fornari, M., . . . Nardelli, M. B.
(2015). The AFLOW Standard for High-throughput Materials Science Calculations.
Computational Materials Science, 108, 233-238.
Dong, C., Ji, Y., Wei, X., Xu, A., Chen, D., Li, N., . . . Li, X. (2021). Integrated Computation
of Corrosion: Modelling, Simulation and Applications. Corrosion Communications, 2,
8-23. doi:10.1016/j.corcom.2021.07.001
Gong, X., Dong, C., Xu, J., Wang, L., and Li, X. (2020). Machine Learning Assistance for
Electrochemical Curve Simulation of Corrosion and Its Application. Materials and
Corrosion, 71(3), 474-484.
Hou, B., Li, X., Ma, X., Du, C., Zhang, D., Zheng, M., . . . Ma, F. (2017). The Cost of Corrosion
in China. npj Materials Degradation, 1(1), 1-10.
Jain, A., Ong, S. P., Hautier, G., Chen, W., Richards, W. D., Dacek, S., . . . Ceder, G. (2013).
Commentary: The Materials Project: A Materials Genome Approach to Accelerating
Materials Innovation. APL materials, 1(1), 011002.
Koch, G. (2017). Cost of Corrosion. Trends in oil and gas corrosion research and technologies,
3-30.
Obot, I., Macdonald, D., and Gasem, Z. (2015). Density Functional Theory (DFT) As a
Powerful Tool for Designing New Organic Corrosion Inhibitors. Part 1: an Overview.
Corrosion Science, 99, 1-30.
Saal, J. E., Kirklin, S., Aykol, M., Meredig, B., and Wolverton, C. (2013). Materials Design
and Discovery with High-throughput Density Functional Theory: the Open Quantum
Materials Database (OQMD). Jom, 65(11), 1501-1509.
Sastri, V., Elboujdaini, M., and Perumareddi, J. (2005). Utility of Quantum Chemical
Parameters in the Rationalization of Corrosion Inhibition Efficiency of Some Organic
Inhibitors. Corrosion, 61(10), 933-942.
Ser, C. T., ouvela, P., and Wong, M. W. (2020). Prediction of Corrosion Inhibition Efficiency
of Pyridines and Quinolines on an Iron Surface Using Machine Learning-powered
Quantitative Structure-property Relationships. Applied Surface Science, 512, 145612.
Shi, Y., Fu, D., Zhou, X., Yang, T., Zhi, Y., Pei, Z., . . . Shao, L. (2018). Data Mining to Online
Galvanic Current of Zinc/copper Internet Atmospheric Corrosion Monitor. Corrosion
Science, 133, 443-450.
Taylor, C. D., Lu, P., Saal, J., Frankel, G. S., and Scully, J. R. (2018). Integrated Computational
Materials Engineering of Corrosion Resistant Alloys. npj Materials Degradation, 2(1).
doi:10.1038/s41529-018-0027-4
Zhi, Y., Yang, T., and Fu, D. (2020). An Improved Deep Forest Model for Forecast the Outdoor
Atmospheric Corrosion Rate of Low-alloy Steels. Journal of Materials Science &
Technology, 49, 202-210.