Gas hydrate stability and the assessment of heat flow through continental margins

Abstract

Summary A prominent feature across some continental margins is a bottom‐simulating reflector (BSR). This seismic reflection generally coincides with the depth predicted for the base of the gas hydrate stability field. Because the occurrence of gas hydrates is controlled by temperature and pressure conditions, it has been suggested that BSRs mark an isotherm and they have therefore been used to estimate the heat flow through continental margins; crucial parameters are the temperature at BSR depth and at the seafloor and the thermal conductivity structure between the BSR and the seabed. However, very often the required parameters are not available and therefore they have been derived from models for gas hydrate stability and empirical relationships to obtain thermal conductivities from seismic velocities. Here, we use downhole temperature, thermal conductivity, porosity and logging data from 10 Ocean Drilling Program (ODP) sites drilled into and through the gas hydrate field to investigate the quality of estimates. Our analyses and application of constraints to the Makran margin off Pakistan indicate the following. (i) The temperature at BSR depth could be approximated by a seawater–methane system, although capillary forces, chemical impurities or non‐equilibrium conditions can lower (or increase) the temperature. If calibration by heat probe measurements is possible, errors of geothermal gradients are less than 10 per cent, otherwise uncertainties of 20 per cent (or even higher) may arise. In addition, seasonal variations of bottom water temperature have to be considered, because they may affect thermal gradients by up to ∼10 per cent. (ii) The impact of typical quantities of low‐thermal‐conductivity gas hydrate on the bulk thermal conductivity is insignificant. (iii) The thermal conductivity profile between the BSR and the seabed can generally be approximated by a mean value. Thus, (iv) seabed measurements should be used instead of empirical relationships, which may produce errors of 5–30 per cent. Consequently, in addition to high‐quality seismic data, a prerequisite should be a large data set of thermal conductivities and oceanographic data. Heat probe measurements are recommended to constrain geothermal gradients. In this case the uncertainty of heat flow is 5–10 per cent of the estimated heat flow. If these data are not available errors/uncertainties can reach 50–60 per cent of the calculated value.

Observed correlation between the depth to base and top of gas hydrate occurrence from review of global drilling data

“…The program is available online at http://hydrates.mines.edu/CHR/Software.html. Additionally, sites reported in Table 1 with known pressure and temperature conditions from drilling at the base of gas hydrate stability and those reported by Grevemeyer and Villinger [2001] are included (Sites 889,892,(859)(860)(861)688,and 808). Error bars reflect uncertainties in the depth of the base of gas hydrate stability zone and geothermal gradients (Table 1).…”

“…Two scenarios of the gas hydrate phase boundary are included for seawater (35 ppt) and methane gas, and an arbitrary mix of 90% methane and 10% ethane (with seawater of 35 ppt). We also added previously published data [Grevemeyer and Villinger, 2001] to complete the data set, although some of these sites did not yield useable information for depth to the TGHOZ or the depth to the BGHSZ. Most plotted sites fall between the pure water and seawater gas (methane) hydrate stability boundaries.…”

Low heat flow from young oceanic lithosphere at the Middle America Trench off Mexico

“…Hyndman et al [15] concluded that in situ data favour a base of the hydrate stability field controlled by the pure methane/pure water phase boundary. However, other authors (e.g., [12,16]) have favoured the pure methane/seawater boundary, and this boundary better fits the data from recent Ocean Drilling Program (ODP) legs [12] (Fig. 4).…”

“…No in situ data are available for our survey area, and borehole data show a scatter of +/À 20% or more in thermal conductivity for unconsolidated sediments at a given depth in the upper 500 m of the sediment column in similar accretionary wedge settings [12,17]. We could have used a depth-dependent value that accounted for compaction by relating the conductivity to seismic velocity (following [18]), but have instead used a constant value as suggested by [12], since the predicted variation with depth is small.…”

“…Heat flow was estimated from BSR depths following an approach first developed by Yamano et al [5] and reviewed in detail in [12]. BSR picks were made only where the reflector depth could be readily picked; in some parts of the survey area a BSR was clearly present, but strong interference from sedimentary reflectors made picking difficult.…”