Gas Hydrate Characterization in the GoM using Marine EM Methods

• Main Author: Constable, Steven
• Other Authors: United States. Department of Energy.
• Format: Report
• Language: English
• Published: University of California, San Diego 2012
• Subjects: 03 Natural Gas; 58 Geosciences; Scripps; Methane hydrate
• Online Access: https://doi.org/10.2172/1049493; https://digital.library.unt.edu/ark:/67531/metadc839257/

In spite of the importance of gas hydrate as a low-carbon fuel, a possible contributor to rapid climate change, and a significant natural hazard, our current understanding about the amount and distribution of submarine gas hydrate is somewhat poor; estimates of total volume vary by at least an order of magnitude, and commercially useful concentrations of hydrate have remained an elusive target. This is largely because conventional geophysical tools have intrinsic limitations in their ability to quantitatively image hydrate. It has long been known from well logs that gas hydrate is resistive compared to the host sediments, and electrical and electromagnetic methods have been proposed and occasionally used to image hydrates. This project seeks to expand our capabilities to use electromagnetic methods to explore for gas hydrate in the marine environment. An important basic science aspect of our work was to quantify the resistivity of pure gas hydrate as a function of temperature at seafloor pressures. We designed, constructed, and tested a highpressure cell in which hydrate could be synthesized and then subjected to electrical conductivity measurements. Impedance spectroscopy at frequencies between 20 Hz and 2 MHz was used to separate the effect of the blocking electrodes from the intrinsic conductivity of the hydrate. We obtained very reproducible results that showed that pure methane hydrate was several times more resistive than the water ice that seeded the synthesis, 20,000 {Ohm}m at 0{degrees}#14;C, and that the activation energy is 30.6 kJ/mol over the temperature range of -15 to 15{degrees}#14;C. Adding silica sand to the hydrate, however, showed that the addition of the extra phase caused the conductivity of the assemblage to increase in a counterintuitive way. The fact that the increased conductivity collapsed after a percolation threshold was reached, and that the addition of glass beads does not produce a similar increase in conductivity, together suggest that while the surface of the gas hydrate grains are not intrinsically conductive, the presence of sand does increase their conductivity. In the field component of this project, we carried out an 18day cruise on the R.V. Roger Revelle in the Gulf of Mexico from 7th-“26th October 2008 to collect controlled-source electromagnetic (CSEM) data over four hydrate prospects; blocks AC 818, WR 313, GC 955, and MC 118. During these surveys we deployed 30 ocean bottom electromagnetic (OBEM) recorders a total of 94 times at four survey areas and towed the Scripps Undersea Electromagnetic Source Instrument (SUESI) a total of 103 hours. SUESI transmission was 200 A on a 50 m dipole antenna at heights of 70-100 m above the seafloor. We also towed a neutrally buoyant 3-axis electric field recorder behind the SUESI antenna at a constant offset of 300 m. The use of a towed receiver that is "flown" above the seafloor allowed us to operate in areas where seafloor infrastructure such as wellheads, pipelines, and installed scientific equipment existed. We reduced the data to apparent resistivity psuedosections. The most compelling results come from the hydrate observatory at MC 118, where a localized resistivity anomaly is clearly identified under the southeast crater in an otherwise uniform 1 {Ohm}m background. The data from MC 118 also show that authigenic carbonate does not necessarily express itself as a confounding resistor, as was feared at the start of this project. While the results from the other prospects are much more complicated, the data are well correlated with known geology, and line to line agreement is good. Although these data are not amenable to 1D inversion as was initially hoped, we expect to use a newly developed 2D CSEM inversion code to continue to get useful information from this rich data set.