UC Santa Barbara

A tracer injection experiment was performed in 3.5 Myr old seafloor comprised of sediment-buried abyssal hills oriented N20°E, and located 100 km east of the Endeavor Segment of the Juan de Fuca spreading ridge in the northeastern Pacific. In the summer of 2010, a mixture of tracers (metal salts, dissolved sulfur hexafluoride [SF6], microspheres) was injected into the crust via borehole 1362B as part of a 24-hour injection experiment during IODP Expedition 327. Fluid samples were subsequently collected from 1362B, and from four additional holes (1026B, 1362A, 1301A, and 1301B) located 300 to 550 m away from the injection hole. The borehole array penetrates a hydrothermal fluid flow system thought to be flowing from SW to NE along a buried abyssal hill often referred to as Second Ridge (SR). Hydrothermal flow is thought to be controlled by a series of exposed volcanic outcrops located along the same buried hill. According to the hypothesis, recharge of bottom seawater occurs through Grizzly Bare, an exposed outcrop 52 km south of the borehole array, and discharge seeps through Baby Bare and possibly also Mama Bare outcrops, which are both located within 5 km to the existing boreholes. The goal of this study is to test hypothesized fluid flow direction, flow velocity and crustal permeability using the conservative gas tracer SF6.

Two small cylinders of SF6 were injected at a fluid pumping rate of 6.7 L/s for 20.2 hours, resulting in a mean concentration of 47.6 µM or 47,600 nM (total of ~23 mol of SF6 was injected). Borehole fluid was continuously sampled in 1.8 mm ID copper tubes using osmosamplers (OS) from each of the long-term, subseafloor observatories (known as CORKs) that had previously been installed in the boreholes. The OS spools were recovered using the ROV Jason in 2011 and 2013. Following recovery, fluid samples were transferred into evacuated vials for measuring SF6 via gas chromatography in a shore-based laboratory. Results of samplers recovered in 2011 indicate the first arrival of injected SF6 ~305 days after injection at Hole 1301A, located 550 m south of the injection Hole 1362B. This suggests that the mean lateral transport of tracer is ~1.8 m/day (660 m/yr), a rate at the upper end of previous upper crustal fluid flow velocity estimates.

The first detection of an SF6 patch south of the injection source apparently contradicts the prior hypothesis that crustal fluid is flowing northward from Grizzly Bare to Baby and Mama Bares, but there are extensive leaks in Hole 1301A, which may have inadvertently pulled SF6 southwards. CORKs 1301B and 1026B are also leaking hydrothermal fluid, and any data collected from these holes should be interpreted carefully. Both 1362A and injection Hole 1362B are operating as intended with no known leaks, although a discharge valve was opened at 1362B in 2011 and closed in 2013. Following this perturbation to the flow field, the 1362A valve was opened in 2013 and then closed in 2014. Tracer was detected in both 1362A and 1362B from the beginning of sampling in 2013, one year after injection. The mean tracer concentration for Holes 1362B and 1362A was 4.6 nM and 4.5 nM, respectively. The sustained signal at these CORKs suggests a hovering of an SF6 patch spanning at least 311 meters in the region. The detection of a considerable amount of tracer to the north at 1362A is consistent with the hypothesized SW-NE fluid flow direction. The hydrogeological fabric at our site appears to be heterogeneous, with fluid transport occurring through small, isolated permeable zones found in the upper volcanic portion of the ocean crust.