Corrosion of Thermally Sprayed Aluminum in Flowing Seawater
MetadataShow full item record
Raw sea water can be pumped into subsea wells where pressure drops have occurred to increase the pressure again and thereby the production rate. Thermally sprayed aluminum (TSA) is considered as internal corrosion protection in pipes conducting this raw sea water. These pipes will have an internal diameter of 6 , be made of carbon steel (typically API 5L X 52) and the flow rates will be in the range 3 6 m/s.TSA is known to provide good and long-lasting corrosion protection both alone and in combination with sacrificial anodes. Little testing has however been done to evaluate its performance when exposed to higher velocities. Aluminum is known to be subject to localized pitting corrosion at negligible flow rates. As the flow velocity increases the corrosion regime changes from local to uniform as the protective oxide film dissolves chemically due to the increased mass transport to and from the surface. Cathodic protection is not expected to hinder this dissolution as it is purely chemical in nature and the protective effect of sacrificial anodes on TSA coated structures under these conditions are therefore doubtful. The dissolution rates were calculated from mass transfer correlations for the flow rate range of interest and found to be 0.15 0.30 mm/year for a smooth surface and 0.20 0.40 mm/year for a rough surface, depending somewhat on the amount of surface roughness.Last fall a rotating cylinder electrode (RCE) was used to simulate the flow conditions and TSA samples were tested at different velocities. The obtained corrosion rates were very high, up to 1,6 mm/year at the highest rotation velocity. Cavitation was however suspected to occur at the surface and there were therefore uncertainties related to the accuracy of the results. Another test design was therefore prepared, and cylindrical samples of TSA coated steel (sprayed with AlMg5 wire) were exposed to sea water in a constructed flow loop. Three different flow rates, corresponding to velocities of approximately 3.6, 4.2 and 4.8 m/s in a 6 internally coated pipe, were tested. The samples were exposed at open circuit potential (OCP) for a total of eight days, and polarized to -1040 mV SCE for 4 days. Ideally, the exposure times should have been much longer but were restricted by material deliveries. After exposure the sample surfaces were analyzed in the scanning electron microscope (SEM), then treated with chromic phosphoric acid for removal of corrosion products and weighed to determine their weight loss and corresponding corrosion rate, before finally being cut in order to examine their cross section, also in the SEM. All tests were also effectuated on solid AlMg5 samples in order to compare the effect of different surface structure and roughness.For the solid AlMg5 samples, the corrosion rates were found to match fairly well with the theoretical calculated values even though the surface analysis revealed a roughened profile and thus the possibility that the samples had been subject to wear. The presence of the exposed cylindrical samples in the channel where sea water flowed at high velocities caused irregularities in the flow regime. And a regular hydraulic regime in the fluid is known to be very important to avoid cavitation-corrosion. The observed wear might therefore have been caused by cavitation. For TSA the corrosion rates (about 1mm/year) were much higher than the theoretical calculated values indicating that other factors may have contributed to the degradation of the coating. Besides possible cavitation, this could have been the combined effect of surface roughness, sample geometry and flow turbulence causing the protective ability of the oxide film to be weakened. The surface roughness seemed therefore to cause increased corrosion rates, and not only because of a larger true available area. The measured open circuit potentials were also very low (down to -1100 mV SCE for TSA) suggesting that the surface was activated in some way. As no especial activating element was detected from the EDS analysis, the activation was thought to possibly be due to the high flow rates causing a less protective film and thus a more activated surface. Polarization to -1040 mV SCE was therefore in the anodic direction for TSA and small positive currents were recorded. For solid AlMg5, which had an open circuit potential of about -980 mV SCE, the polarization was cathodic but unusual high cathodic current demands were recorded (about -300 mA/m2). The current demand had however probably not stabilized and a much longer exposure would have been required to be able to conclude.Too short exposure times were a clear weakness of the performed experimental work. Questions also arose regarding the design of the test method since the varying conditions around the exposed cylinders were likely to have caused uneven corrosion rates and the velocity and pressure variations that occurred due to the obstacle these cylinders presented to the flow may have caused cavitation. In the case of internally coated pipes, these geometry effects would not be the same and the corrosion rates consequently not increased in the same way. The problem of cavitation was already suspected during previous tests with the RCE and the objective to avoid this here was not reached.The obtained corrosion rates for TSA indicate that it is not an ideal material to provide long term protection for the sea water pipes in question, but are probably falsely high. However, even the theoretical corrosion rates based on mass transfer considerations imply that the entire coating thickness would be dissolved within a year. If TSA is to be used for the desired purposes, an inhibiting factor hindering or strongly slowing down the chemical film dissolution rate must be found.