The enhancement of brine corrosivity due to the presence of acetate (or anions of other weak acids) in CO2 containing waters has been discussed frequently in recent literature. It has been confirmed that the observed increase in corrosivity is related to the equilibrium concentration of undissociated acetic acid (HAc) in the water and this in turn can be determined via solution speciation calculations - providing the composition of the water is known.
As with all weak acids (and regardless of the in-situ pH), organic acids will be present in both their undissociated form and dissociated form (anions) and since propionic and butyric acids essentially share the same pKa as acetic acid, their molar concentrations can be typically summed under the generic name of the dominant species, i.e., acetate ions (Ac-) and acetic acid (HAc).
This paper discusses the implications of significant levels of organic acids in produced waters (> 1mM) with regards fluid corrosivity. Specifically, how can we account for this in corrosion prediction modelling, as well as ensuring any testing programs conducted correctly reflect the water chemistry and speciation? Our approach to water speciation calculations and subsequent use of this will be discussed in relation to a specific example. The rate of corrosion is not determined solely by solution composition, however, we can't understand or estimate the corrosion rate without considering the solution speciation.
Soon after the realization that CO2 was a corrosive agent in gas condensate wells in the early 1940's, it was suspected that the presence of weak organic acids in the aqueous phase could also introduce a significant corrosive threat1. However, this was not confirmed until the early 1980's because of the difficulty or failure to correctly analyse the produced waters for organic acids. The potential impact of acetates on aqueous corrosivity was alluded to by Crolet and Bonis; however the relationship between brine composition and corrosivity was not understood in more detail until much later.
Some understanding of the role of HAc in CO2 corrosion can be gleaned from field experience as related to Top-of-Line-Corrosion. However, until recently, few studies have endeavoured to study the basic effect of HAc on the electrochemical corrosion process. Although HAc has for many years been reported to increase corrosivity in CO2 systems, only in recent years have the roles of in-situ pH, acetic acid, acetate and bicarbonate contents, ionic strength, partial pressure of CO2 and temperature been more clearly defined with regards their combined effect on aqueous corrosivity. A comprehensive review of CO2 corrosion in the presence of HAc has recently been published.
It appears to be generally accepted that the main cause of the enhanced corrosion in the presence of acetic acid/acetate is the equilibrium concentration of undissociated acetic acid.[5–10] Several authors have proposed undissociated acetic acid has a key role as a reactant at cathodic sites, whether this involves HAc acting as a source of hydrogen ions via dissociation at the surface or via direct reduction of HAc at the surface.[5,8] Irrespective of this, in practical terms the ultimate requirement is to be able to sensibly predict when acetate enhanced corrosion is likely and to estimate the magnitude of the effect.
A significant number of CO2 corrosion prediction models currently exist that are used in support of E&P system design and generally for quantifying the magnitude of the potential risk of corrosion. Often the challenge is to demonstrate the acceptable use of carbon and low steels either alone or together with inhibitor treatment and the inclusion of a corrosion allowance on wall thickness.
Traditionally, corrosion prediction models have been based on either empirical or semi-empirical correlations to laboratory or field data.[11–15] However, mechanistic models[16–18] are becoming more common as the different aspects that influence CO2 corrosion have become more deeply understood. With the need to understand additional and more complex corrosion mechanisms, mechanistic models have been developed and will likely continue to become more widely used.