Samples of 410 stainless steel were exposed to aerated and deaerated solutions of artificial sea water and distilled water, with and without packer fluid additions, with a CO2 atmosphere. Pitting rates were determined from autoclave exposures, and electrochemical parameters were obtained from potentiodynamic polarization curves.
The studies showed that in deaerated solutions, packer fluid had a mild inhibiting effect on liquid phase pitting in both seawater and distilled water. Packer fluid increased the rate of attack only in aerated solutions of distilled water. However, vapor phase pitting rates in uninhibited seawater, whether aerated or deaerated, are high enough to have caused the rates of attack observed in the field.
Disturbingly deep pitting in a number of 410 stainless steel wellhead valves on Dutch North Sea Platforms K-10-B and K-10-C were revoked. Initial investigation into the problem indicated that a specific packer fluid was responsible for the accelerated pitting. Later laboratory investigations were not conclusive. In order to determine the exact causes for the pitting experimental studies were conducted on pitting of 410 stainless steels in various environments which might have come in contact with the wellheads.
Conditions in these wells were as follows:
Water production - 0.8 bbl/mcf
Total depth-11000 feet
Pressure-220-240 bar (3390-3700 psia)
Flowing wellhead temperature-70 c (158 F)
Gas composition-1.7% CO2, bal. Methane
Brine composition-150,000 ppm Cl
After the wells were drilled to total depth, the drilling fluid was displaced with surface seawater. No attempt was made to control oxygen in this water. The tubing was run in, and seated into the packer. The sliding sleeve was opened to allow the seawater to be displaced with packer fluid inhibited packer fluid. Seawater was left in the tailpipe and production interval annulus.
At this point the christmas trees were installed. The production interval was perforated with through-tubing guns, and the wells were flowed to produce the packer fluid in the tubing, the seawater in the tailpipe, and then the formation brine.
The wells were flowed until no further liquid was produced, then shut in. In no case was any well produced longer than one month. The wells had been shut in for approximately 5 to 6 months when the pitting was discovered on disassembly.
Maximum depth pit on the wellheads from this field indicated a pitting rate of 47 mils per year. The actual solution in contact with the wellhead internals during shut-in is not known. Seawater, Packer fluid and formation brine all contacted the valves at one time or another. The saturation water content of methane at 70 C and 220 atmospheres, is approximately 0.4 barrels/mmcf1, which is well below the levels encountered in gas in these wells. Consequently water would condense out of the gas in these fields onto wellhead internals during shut-in. Such condensate could be relatively pure of minerals but would of course, be saturated with CO2 at the shut-in temperature and pressure.
Two experimental methods were used: evaluation of coupons exposed for four weeks in autoclaves under simulated shut-in conditions, and electrochemical potentiodynamic polarization studies on electrodes exposed under the same conditions.
The CO2 partial pressure in this field is 3.75 bar (58 psia). Therefore, the autoclave experiments and potentiodynamic polarization tests were conducted at 58 psia in 100 percent CO2. Solutions were maintained at 150 F during the tests.
In the autoclave experiments, 0.1-inch by 0.5-inch by 0.056-inch 410 stainless steel coupons were cleaned, weighed and then exposed in one liter 316 stainless steel pressure vessels. Approximately 500 ml of solution was used. Duplicate samples were exposed in both the vapor space and the liquid in each vessel. Teflon washers serrated as described in ASTM specification G78 were bolted onto each coupon to create an artificial crevice.
Deaerated solutions were sparged with nitrogen overnight before being poured into the autoclaves. After the autoclaves were sealed, the deaerated solutions were evacuated to 29 inches of mercury with a mechanical vacuum pump before the CO2 was introduced into the vessel. Aerated solutions were not sparged or evacuated prior to testing.
Inhibited solutions had the packer-fluid inhibitor added in the proportion 3 ml to 500 ml of solutions. This is equivalent to the recommended field dosage level of 25 gal/100 bbl H2O.
Seawater solutions were mixed by adding 41.96 gm/l artificial sea salt (as described in ASTM specification D1141) to distilled water at room temperature.
After four weeks' test time, the coupons were removed, cleaned and reweighed. Maximum pit depth was determined on a metallograph with a micrometer stage.
The potentiodynamic polarization experiments were conducted in a 250 ml Hastelloy C autoclave. A test electrode of ?? inch diameter 410 stainless steel rod was inserted into the autoclave through packing glands at the top and bottom. The exposed area of each test electrode was 7.9 sq. cm, or 1.22 sq. in. The test electrodes were insulated from electrical contact with the vessel using teflon bushings and Ryton (polyphenoline sulfide) spacers.
A Hastelloy C indicator electrode was introduced into the autoclave through a ?? inch NPT fitting. This indicator electrode was electrically insulated from the body of the vessel with teflon and polyethylene fittings. The stability of the Hastelloy C indicator electrode was checked against a commercial Ag/AgCl reference electrode. The body of the autoclave itself was used as the counter, or auxiliary, electrode.
Each test was conducted in 200 ml of solution. Solution mixing and aeration of deaeration practice were as described above for the autoclave tests. Packer-fluid was added to the inhibited solutions in the proportion of 1.19 ml to 200 ml of solution. Two tests in aerated fresh water were conducted in Houston tap water. After the autoclave was sealed and charged to 58 psia with CO2, the vessel was brought to 150 F with electric resistance heaters controlled through a Type J thermocouple. When the temperature had stabilized, the open circuit potential of the test electrode was measured with a digital voltmeter.
Starting with the test electrode at open circuit, applied current was measured as the potential of the test electrode was dynamically polarized in the cathodic direction at the rate of 5 V/hr. When either limiting current or the onset of concentration polarization effects was reached, the scan direction was reversed and continued in the anodic direction. The test electrode was polarized back through the open circuit potential and through the anodic passive current zone.
When the test electrode was well into the anodic transpassive zone, or when limiting current was reached, the scan direction was reversed once more and continued in the cathodic direction until the appearance of the current minimum signaling that the repassivation potential was reached.
Weight loss measurements made on the coupons from the autoclave tests were used to calculate general corrosion rates using the formula:
Maximum pitting rates were calculated by dividing the maximum pit depth observed on each coupon by the exposure time, and converting this number to mils per year.
Tables 1 and 2 present general and pitting corrosion rates calculated for vapor-space and liquid exposure in each autoclave. The values given are the average of the duplicate coupons from each condition. Photographs of the coupons after exposure are presented in Figures 1 through 8. The results of the electrochemical experiments are presented in Figures 9 through 18. These figures show applied current as a function of potential for the ten different environmental conditions tested. Direction of the scan at any given instant (cathodic of anodic) is given by the arrows on the potential current traces.