Abstract
The predictions of the so called CO2 or sweet corrosion and the so called H2S or sour corrosion are questions raised and still pending for some 60 years. A reliable prediction actually requires to understand when high or low corrosion rates are encountered in E&P operations (field data), how they can occur (comprehension of the various mechanisms ruling the corrosion layer protectiveness), and finally why they do occur (link between data and mechanisms). So far, however, and despite long lasting researches, these three aspects had always been studied separately.
In Part 1, the specific chronology of the CO2 and H2S themes was reminded, and the concept of corrosion layer protectiveness refined, especially in the presence of two acid gases with iron solubility values as different as for FeCO3 and the various iron sulfides. The why of the CO2 corrosion prediction was also made fully explicit. This field based prediction tool was indeed well known for some 25 years, but it remained partly unexplained. Therefore, the additional explanations of Part 1 not only support it, but also enable a much more simple formulation.
The present paper is aimed at doing the same for H2S corrosion, explaining the why of H2S corrosion, from the recently gathered field experience. It also refines the previous protectiveness mechanisms in sour media, especially through an in-depth analysis of 50 years experience in the production of the Lacq field in France (considered in the early 50's as a giant field of deep sour gas). This analysis extends from the very beginning of early production until the final exhaustion, and it emphasizes the never yet considered role of the isothermal decompression of reservoir water, as well as the inevitable trapping of water slugs at the bottom of HP gas wells. It also describes and explains a new type of profuse and non protective corrosion layer. It was named "calcic" (Ca) because its structure is made of both CaCO3 and FeS, and in H2S corrosion, it may form in any CaCl2 brine saturated in CaCO3. Despite huge PH2S in wells, the perfect solubility of dissolved iron within such layers is due to the immediate removal of any cathodic alkalization and any sulfide ion by CaCO3 precipitation.
This new Ca layer is to be added to the former non protective insoluble anionic layers (IA), in which the internal iron solubility is due to a local shortage in either H2S alone (IA1) or both H2S and CO2 (IA2).
The field data base presented in 2009 is also revisited in the light of this new information. Especially, it is shown that the well data base has still to be complemented, but in a mechanism minded way, i.e. by keeping in mind the background of these three non protective layers specific to H2S corrosion. Similarly, the data base on sour wells shall start from the very first minute traces of H2S in the acid gas mix, since regarding mechanisms, the so called "CO2 fields with H2S traces" are no longer relevant to a prediction by the previous algorithm of Part 1, but by the present Part 2, as "sour fields with minute H2S mole fractions".
Altogether, it already appears that sour oil wells are never corrosive at least above 3 % H2S in the acid gas mix. The same is true for gas wells with an active aquifer, i.e. producing their reservoir water in the same way as oil wells, and as long as depletion will not completely alter the raw petroleum data given in the present field data base. Conversely, the bottomhole of HP gas wells above 450 bar with no active aquifer will soon or later pass through a corrosive period, due to the presence of trapped water slugs, whereas the condensing zone at the top will never be corrosive.
Naturally, more field data are still necessary for refining the thresholds, but the present algorithm can already be used for predicting sour weight loss corrosion in a great many wells.
It is also worth noting that the present Part 1 and Part 2 can be illustrated by a common logigram of the downhole "weak acid corrosion".
