Inhibited calcium bromide brines can cause severe localized corrosion of steels as a result of CO2 acidification. In contrast, formate brines cause only slight general corrosion of steels after acidification.
An operator wishing to drill, complete or work over a well with a clear fluid has to decide between halide or formate brines. The corrosivity of the brine under expected wellbore conditions is a key criterion used in the fluid selection process.
With proper pH control, both calcium halide brines (CaBr2 and CaCl2) and formate brines (NaCOOH, KCOOH and CsCOOH) are non-corrosive over a wide temperature range in an oxygen-free environment. In addition, formate brines are known to be non-corrosive in the presence of oxygen and when contaminated with halides.
Little attention has been paid to the corrosivity of these brines in circumstances where pH control is lost as a result of a significant influx of acid gas into the well bore from leakage or gas lifting operations.
To fill this gap in our knowledge, an extensive laboratory test program was conducted using long-duration corrosion tests on brines acidified by CO2. The study was restricted to an examination of general and localized corrosion effects. The separate issue of stress corrosion cracking (SCC) of corrosion resistant alloys (CRAs) in halide brines was not addressed.
Testing has revealed important differences between the corrosivity of halide and formate brines when challenged by an influx of acid gas and corresponding drop in brine pH.
Weight loss and real-time linear polarization resistance (LPR) tests were conducted using specimens of C-steel, 13% Cr and "Super 13% Cr" (S13% Cr, Hypo II) exposed to two bromide (43% CaBr2) solutions and one formate (71% potassium formate) solution for up to 63 days in test autoclaves. All solutions had a specific gravity of 1.51. One of the bromide solutions was inhibited with a corrosion inhibitor commonly used in halide brines, and a potassium carbonate/bicarbonate buffer (20 + 40 g/dm3) was added to the potassium formate.
Temperature was maintained at 246?F (120?C) until the corrosion rate had stabilized. At this point, temperature was increased in intervals to 354?F (180?C) and then decreased accordingly to 246?F. The solutions were bubbled with CO2 at a constant partial pressure of around 10 bars. The autoclaves were thoroughly flushed with CO2 prior to heating to remove traces of oxygen. After the tests, the LPR corrosion rates were harmonized with weight loss data.
Severe localized corrosion that perforated some of the 2-mm thick C-steel specimens, developed on both the carbon steel and 13% Cr steel in the bromide solutions, whereas only uniform corrosion took place in the formate solution. Examples of weight loss specimens from the formate solution after exposure and cleaning are shown in Figure 1. Before cleaning, the C-steel specimens in all solutions and the 13% Cr specimens in the potassium formate were covered with a dark gray - apparently uniform - iron carbonate layer, whereas the 13% Cr specimens in the bromide solutions were only partly covered. The S13% Cr specimens in all solutions appeared blank and uncorroded, which was confirmed by very low weight losses. After the tests were completed, pH measurements in the solutions verified that acidification from CO2 influx had taken place.
The LPR corrosion rates throughout the test period for C-steel and 13% Cr in the two bromide solutions are presented in Figures 2 and 3. The corresponding data from the formate solution are given in Figure 4.
The following can be extracted from the LPR data:
Carbon steel corrodes with a high initial rate in the formate solutions, decreasing rapidly to values below 0.1 mm/year at 246?F. This decrease is attributed to the build-up of a thin (5-20 µm), dense and protective iron carbonate layer. The LPR corrosion rate increases linearly with temperature up to about 0.15 mm/year at 350?F. The corrosion rates decrease with time at all temperature levels.The smooth shape of the curve indicates that only uniform corrosion takes place.
A corresponding initial peak was not observed in the bromide solutions. Here, the time period with high corrosion rate lasts for several hundred hours and the scatter in the curves indicates that localized corrosion takes place during this period at 246?F. In the bromide brines, corrosion also levels off to about 0.1 mm/year at 246?F with a linear increase with temperature to about 0.20 mm/year at 350?F. The long-term effect is again attributed to the build-up of an iron carbonate layer. No effect of the corrosion inhibitor can be seen.
13% Cr corrodes with a low rate (0.05 mm/year) at 246?F in the bromide solutions. Rates increase with a rise in temperature. The scatter in the data above 282?F (140?C) indicates that localized corrosion of 13% Cr takes place at higher temperatures.
The corrosion of 13% Cr in the formate solution levels off at about 0.5 mm/year at 246?F after a higher initial rate. The corrosion rate increases linearly with temperature up to about 1.5 mm/year at 350?F. The shape of the curve indicates a uniform corrosion mechanism.
The main reason for the differences in corrosion behavior of C-steels in the acidified halide and formate brines is the presence of a small amount of formic acid in the formate brine (as a result of CO2 influx) that initially accelerates general corrosion of C-steel. With the low solution volume-to-surface ratio present in a well annulus, a fast build-up of ferrous ions (Fe2+) takes place in the solution and iron carbonate supersaturates and forms a stable layer on the corroding surface. In the bromide brines, with no formic acid and high amounts of halide ions, general corrosion is lower and extensive localized corrosion takes place before a protective iron carbonate film can be formed.
The presence of the small amount of formic acid can also explain the higher general corrosion of 13% Cr steels in formate brines compared to bromides. Again, the low general corrosion and the high halide concentration in the CO2-acidified bromides result in localized corrosion at elevated temperatures. The limited resistance of 13% Cr, in comparison to C-steel, after some hundred hours in the formate solution is attributed to the high chromium content. This prevents formation of an iron carbonate layer with as good protection ability as the layer formed on C-steel.
S13% Cr is practically resistant to both acidified formate and halide brines, if the cracking aspect of halide brines is ignored. The high resistance of S13% Cr, compared to the leaner 13% Cr steel, is attributed to the high molybdenum (Mo) content (2.1%). Mo has a significant effect on both the resistance to localized corrosion in moderately acidified halide-containing solutions without an oxidizing agent, such as oxygen, and the resistance to formic acid.
The corrosion inhibitor's failure to protect the metal in the bromide solution is most likely explained by the fact that the inhibitor is designed for higher pH environments. To our knowledge, this type of corrosion inhibitor is commonly used in oilfield brines, and we are not aware of any corrosion inhibitors that have been developed for acidic conditions.
Consequences and recommendations
The practical consequence of these new findings is that special care should be taken when selecting heavy brines and materials for well completion where influx of acid gas is a concern.
For short-term use, (e.g. for use as a completion fluid for up to a year), there should be no concern using formate brines as the CO2 corrosion is of a general form. For longer-term use, (e.g. as a packer fluid), there will be some temperature limitations with use of standard 13% Cr steel due to the relatively high general corrosion rates at elevated temperatures in the presence of CO2. For C-steel, overall corrosion will be low due to the formation of the high quality protective iron carbonate layer. CRAs of grade S13% Cr and higher will be practically resistant.
Halide brines, on the other hand, show significant localzsed corrosion on C-steel when acidified with CO2. Addition of a commonly used inhibitor does not improve this situation. Halide brines are therefore not recommended for use with C-steel if influx of acid gas is expected. Due to localized corrosion, there are clear temperature limits for use of halide brines with standard 13% Cr, whereas S13% Cr and higher alloyed CRAs are practically resistant if the SCC aspect is ignored.
Although not part of this study, SCC is a major problem in halide brines, and needs to be considered in the process of selecting well chemicals and materials.