School of Civil and Environmental Engineering

Ph.D. Thesis Defense Announcement

GEOPHYSICAL FLUID FLOW DURING HYDROTHERMAL VENTING AND CARBON SEQUESTRATION

By

Joshua E Smith

Committee members: Dr. J. David Frost (CEE), Dr. Haiying Huang (CEE), Dr. Sheng Dai (CEE),

Dr. Robert P. Lowell (EAS), and Dr. Lawrence C. Murdoch (Clemson)

Date and time: Monday July 17th, 2017 at 2:00 pm

Location: Mason 2119

Fluid flow influences mechanical processes in the earth's crust, but many aspects of these processes remain poorly understood; in large part, because of a scarcity of controlled field experiments or measurements at appropriate scales. For example, advective heat transfer data from hydrothermal sites are necessary for understanding the geochemical and nutrient fluxes to seafloor biological communities and for constraining subsurface models at mid-ocean ridges. Yet, such data are limited or lacking. This work provides the most comprehensive database currently available for fluid flow and heat output from seafloor hydrothermal systems. We describe 120 new measurements on the Juan de Fuca Ridge (North Pacific), Lau Basin (South Pacific), and East Pacific Rise (Equatorial Pacific) collected on the seafloor with submersibles Alvin and Jason.

These data constrain models of turbulent buoyant plumes that are difficult (if at all possible) to reproduce in the laboratory. Taylor's entrainment hypothesis, specifying the entrainment coefficient, has become a standard approach for closing models of turbulent plumes. The value of the coefficient is determined in laboratory experiments or field measurements and is commonly considered as being virtually a universal constant. We studied five black smoker plumes, for which we have detailed measurements. We reduce seawater density to a linear function of temperature. If the error in this assumption is small, the entrainment coefficient under black smoker conditions is half the expected value. This is significant as it would strongly affect the heat flux estimates for hydrothermal plumes in the near seafloor water column. We then characterized the flow regimes of 99 seafloor hydrothermal vents (mostly black smokers) based on our measurements, and our results suggest lazy plumes constitute 85% of hydrothermal discharge and the remaining 15% are forced plumes. The percentage of forced plumes is low, but they play a disproportionate role in the support and dispersal of biological communities. It appears that in the mid-ocean ridge environment, forced hydrothermal plumes have considerably larger entrainment and higher height than lazy plumes. In turn, forced plumes are likely to play a much more effective role in larvae transport mechanisms near hydrothermal vents and are more effective in supporting microbial communities within the plume.

The last topic of this work is concerned with monitoring fluid flow in geologic formations. Specifically, there is significant concern on how to monitor fluid migration during carbon storage and petroleum operations. During fluid injection, pressure redistribution and formation properties affect the deformation pattern, and this effect is possible to interpret from field measurements of the strain tensor. Modern borehole strainmeters are now capable of measuring multiple components of strain and tilt in the shallow subsurface, and these measurements can be used to interpret processes at much greater depths. The first field test of this technique will occur during a waterflooding operation at the North Avant oil field in Osage County, OK. This field is a representative example of geological formations proposed for carbon storage. To design the field test, we developed a model of the poroelastic response to fluid injection and determined zones of deformation optimal for measurements. Currently, two boreholes have been drilled for instrument installation based on this modeling. The model is based on our geologic analysis of the North Avant field site, but it can be applied elsewhere. The model shows it is indeed possible to use monitoring wells at significantly shallower depths than the reservoir for measuring strain signals generated by waterflooding or carbon sequestration operations. Additionally, permeability is likely to vary within reservoirs, but boundaries are challenging to identify. We show that it is feasible to identify channels of high permeability in deep formations using the strain tensor measured in shallow boreholes. This is significant as such channels are common in petroleum formations consisting of fluvial deposits and strongly affect the fluid flow pattern.