The Effect of HAZ Microstructure on the Utilisation Potential of 13%Cr Flowlines

Abstract

Supermartensitic stainless steel pipelines are used in the oil and gas industry mainly to transport oil and gas products from wells to offshore installations or even to shore. Most of these pipes are installed by the reeling process, involving up to 3% plastic strain of the pipe wall during installation. Supermartensitic stainless steels are a relatively new group of alloys that contain 11-13% chromium, 2-6.5% nickel and 0- 2.5% molybdenum. They contain less than 0.015% carbon. In comparison, more common martensitic stainless steels contain 0.1-0.2% carbon. The decreased carbon content increases the weldability of the steel by reducing the hardness in the heat affected zone and reducing carbide precipitation.

This work has investigated the most typical failure mode of supermartensitic stainless steel pipelines in the oil and gas industry. Published results of failure investigations have revealed that this failure mode is hydrogen induced cracking in the weld coarse grained heat affected zone. The hydrogen source is either weld consumables or hydrogen developed on the steel surface because of the negative potential induced by the cathodic protection system. In the latter instance, the cracks have typically appeared in the toe of the fillet weld that attaches the sacrificial anode connection doubler plate (anode pad) to the pipe wall.

The objective of the present work was to investigate the influence of the microstructure on the hydrogen cracking susceptibility of two typical pipeline supermartensitic stainless steels. The work has concentrated on the coarse-grained heat affected zone because this is the typical crack initiation point. The hydrogen cracking susceptibility was tested by tensile testing notched steel samples under varying conditions.

Three different sets of tensile tests have been carried out. All specimens were subjected to weld heat affected zone simulation using a Smitweld simulator. All samples were first heated to a temperature close to the melting temperature in order to create a microstructure that is typical of the coarse-grained heat affected zone. Some samples were also subjected to a second heat cycle representing an adjacent second weld pass. The peak temperature of this second heat cycle was varied. The tensile test samples were machined after simulation with a 1 mm deep notch in the simulated zone. Two different materials were used throughout the testing, designated steel A and B. Both were rich grade supermartensitic stainless steels typical of the pipeline steels used in the North Sea. The distinct difference between them is that steel A is alloyed with titanium in order to avoid selective corrosion along previous austenite grain boundaries, while steel B is not alloyed with titanium.

The first set of samples were tested without any influence of hydrogen and monitored during testing in order to investigate any tendency of local cracking during straining. After the tensile tests these samples were subjected to light microscopy and hardness testing in order to characterise the microstructure of each sample. The second peak temperature was found to have an impact on hardness. A second heat cycle to 500-550oC resulted in an increase in hardness; while a second heat cycle to 625-650oC decreased the hardness.

The second set of samples were charged with hydrogen in 3.5%NaCl electrolyte prior to tensile testing in order to investigate the tendency for hydrogen embrittlement from hydrogen present in the material prior to straining. This is similar to the case in which hydrogen is introduced through the welding process, and the pipe is installed by reeling at a later stage. The tensile testing was performed at a slow rate (10-6 s-1) in order to allow for the necessary hydrogen diffusion to the crack tip to achieve stepwise hydrogen cracking. A significant reduction in ductility compared to samples without hydrogen was observed.

The third set of samples were tested under constant load immersed in 3.5% NaCl and subjected to a negative potential corresponding to that induced by the cathodic protection system. The load was increased by a small amount every second day in order to establish the threshold stress for hydrogen cracking during service in seawater with hydrogen being introduced on the steel surface by the cathodic protection system. A significant influence of the second heat cycle was observed. This effect was attributed to both a variation in yield strength and the influence of precipitates on hydrogen solubility and diffusion.

Supermartensitic stainless steels contain a certain amount of retained austenite, and this is regarded as beneficial because it increases toughness and ductility. The retained austenite is produced intentionally in pipe products by heat treatment at a temperature where the alloy is partially transformed to austenite, normally 600-700oC. The tendency to form retained austenite in the two materials was investigated by heat treatment of samples at different temperatures and time. Phase transformation temperatures were measured by dilatometer tests and X-Ray diffraction. The highest amount of retained austenite was achieved after a double tempering heat cycle, and the formation of retained austenite required longer heat treatment times in steel A than in steel B.

Retained austenite may play a role in the hydrogen embrittlement process because austenite has a much higher hydrogen solubility than martensite has. The effect of the retained austenite on hydrogen cracking was investigated by tensile testing of standard round bar specimens that had been heat treated in order to achieve different levels of retained austenite. A significant effect of the retained austenite was observed. Samples with high amounts of retained austenite experienced a much higher reduction in ductility after hydrogen charging than samples with low amounts of retained austenite. In order to explain this effect, the hydrogen solubility of samples containing different levels of austenite and precipitates was measured. This was achieved by charging the samples to saturation in an electrolyte and performing hydrogen analysis. Carbide precipitation was also found to influence hydrogen solubility, and carbides precipitated at different temperatures were identified by transmission electron microscopy.