OTC 20801

Gulf of Mexico Gas Hydrate Joint Industry Project Leg II: Initial Results
from the Green Canyon 955 Site
D.R. McConnell, AOA Geophysics Inc.; T.S. Collett, U.S. Geological Survey; R. Boswell, National Energy
Technology Laboratory; M.C. Frye, W.W. Shedd, R. Dufrene, and P. Godfriaux, Minerals Management Service;
S. Mrozewski, G. Guerin, and A. Cook, Lamont Doherty Earth Observatory; and E. Jones, Chevron Energy
Technology Co.

Copyright 2010, Offshore Technology Conference

This paper was prepared for presentation at the 2010 Offshore Technology Conference held in Houston, Texas, USA, 3–6 May 2010.

This paper was selected for presentation by an OTC program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material does not necessarily reflect any position of the Offshore Technology Conference, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Offshore Technology Conference is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of OTC copyright.

Abstract
In April and May of 2009 the Gulf of Mexico Gas Hydrate Joint Industry Project realized its second field program (Leg II)
with the semi-submersible Helix Q4000 drillship. The three week expedition drilled seven logging-while-drilling (LWD)
holes at three sites that tested a variety of geologic/geophysical models for the occurrence of gas hydrate in sand reservoirs in
the deepwater Gulf of Mexico. At the GC 955 site, high saturation gas hydrate deposits in sands were found, where predicted,
at two of the three holes. The full research-level LWD assembly deployed for Leg II collected data on formation lithology
and porosity, and included quadrapole acoustic and high-resolution 3-D resistivity logs. No samples, only LWD data were
collected in Leg II. The three holes in GC955 were drilled where a wide and thick late Pleistocene channel complex has been
raised and fractured by salt uplift. A four-way closure with numerous amplitude anomalies at the base of gas hydrate
stability is near-to but west of the channel axis. The first well (GC955-I) was drilled very close to the channel axis in a
location with muted geophysical indications of gas hydrate. More than 300 ft of porous sands were encountered as predicted;
however the sands contained primarily water – with only modest indications of gas hydrate. The next hole (GC955-H)
targeted sands higher in the four-way closure. Fracture filling gas hydrate was detected above the deeper sand target, and, at
the target, 98 ft of sand fully saturated with gas hydrate with little to no gas beneath. A third well (GC955-Q) also
encountered at least 35 ft of fully saturated gas hydrate sand at the target depth, but drilling was aborted because of a gas
hydrate dissociation event or penetration of free gas and subsequent gas flow. The JIP's discovery of thick gas hydrate-
bearing sands at the GC 955 site validates the integrated geological and geophysical approach used in the pre-drill site
selection and provides increased confidence in assessment of gas hydrate volumes in the Gulf of Mexico.

Introduction
The Gulf of Mexico Gas Hydrate Joint Industry Program (JIP) is a cooperative research program run by an international
consortium of energy companies managed by Chevron in partnership with the U.S. Department of Energy. The primary
objectives of the JIP Leg II drilling program was to determine the occurrence of gas hydrate within sand reservoirs in the
Gulf of Mexico, to assess current approaches for interpreting gas hydrate occurrence from geologic and geophysical data, and
to determine the most suitable sites for additional drilling and coring in future phases of the JIP program. Gas hydrate is a
unique type of chemical substance in which a lattice of water molecules encase, without chemical bonding, a gas molecule.
Gas hydrates, if found in high concentration, are a dense energy resource. A cubic ft of methane hydrate contains 184,000
Btu of energy. Gas hydrate can be present in deep marine sediments where pore fluids are highly saturated with gas and the
pressure temperature conditions are favorable. Hutchinson et al. (2009)1 provide a review of the JIP’s site selection process
and of the analyses conducted to select the targets to be permitted for possible drilling. The final drilling program included
three sites, Alaminos Canyon block 21, Walker Ridge block 313, and Green Canyon block 955 (GC 955) in the northern Gulf
of Mexico (Figure 1). Initial summaries of JIP Leg II operations, scientific results, and logging-while-drilling data collection
methods and operations are provided by Collett et al. (2009)2, Boswell et al. (2009)3, Mrozewski et al. (2009)4, McConnell et
al. (2009a) 5, McConnell et al. (2009b) 6and Guerin et al. (2009) 7 respectively. This paper describes the scientific rationale
and initial results for the LWD program conducted at the GC 955 Site.
2 OTC 20801

The JIPs selection of sites for drilling in Leg II was based on an analysis of the various elements needed to support a gas
hydrate petroleum system (appropriate temperatures and pressures, gas and water sources and migration pathways) with the
likely presence of sand reservoirs (needed to enable gas hydrate concentrations to high levels). In addition, the site selection
process also incorporated attempts to directly detect and quantify gas hydrates from seismic data (see Hutchinson, et al.,
2009) 1. All of these elements were seen in the GC 955 area, providing a multitude of potential, high value sites for the JIP
program.

Figure 1. Northwestern Gulf of Mexico showing tabular salt and mini-basin province, Sigsbee Escarpment and AC 21, WR 313, and
GC 955 sites.
The prospectivity of the GC 955 site was based initially on the work of McConnell (2000)8 and Heggland (2004)9.
Although neither author mentioned the potential for gas hydrate deposits in the area, both described geophysical indications
for ample gas sourcing, clear gas migration pathways into the shallow sediments afforded by extensive faulting, and presence
of thick sand reservoirs associated with a large and persistent Pleistocene channel-levee complex. In addition, data from
seismic in conjunction will log data from an existing industry well (the GC 955 #001) indicated that a thick, sand-prone
section likely spans the base of gas hydrate stability (Figure 2). As a result, the GC 955 site was proposed to the JIP site
selection effort by the AOA Geophysics team in early 2006. Subsequent analyses conducted by the AOA Geophysics team
to determine the potential depth of the gas hydrate stability field from pressure, temperature and gas composition data (Figure
3) indicated that the observed distribution of anomalous high-amplitude reflectors are consistent with the inferred base of gas
hydrate stability (BGHS). To estimate whether the target reservoir was within the gas hydrate stability zone, temperature,
salinity, and pressure conditions were modeled by assuming a seabottom temperature of 4°C, normal seawater in the pore
space, and that the trend of shallowest gas at GC 955 #1 represented the base of gas hydrate stability at hydrostatic pressure.
The resulting geothermal gradient of 32°C/km is considered to be a reasonable value for sediments above shallow salt in the
GoM (Nagihara and Smith, 2008) 10. The geothermal gradient and pressure estimates place the target reservoir facies depths
within or near the BGHS.
In 2008, seismic data under license to Chevron were forwarded to AOA Geophysics to develop potential gas hydrate
targets for JIP Leg II. Numerous prospective targets were identified that tested a variety of geologic settings within the
block. Targets were developed within and proximal to the most prominent channel features, as well as within a salt-cored,
domal structure in the southwestern corner of the block. This structure features the most prospective geophysical anomalies
and the strongest indicators of gas occurrence. The highly-faulted structure displays little lateral continuity of the reflections
(Figure 4) and a complex array of responses that suggested the close association of both free gas (strong leading troughs) and
OTC 20801 3

gas hydrate (strong leading peaks). Despite numerous clear migration pathways further up in the section, afforded not only
by faults but by the inferred presence of permeable reservoirs, these amplitudes were not present in the shallower section,
suggesting that gas hydrate may be serving as a seal, restricting the vertical movement of gas.

Figure 2. Perspective view from 3-D seismic of top of channel facies showing JIP and industry well penetrations, and four-way
closure. Interpretation by AOA Geophysics.
Targets identified in the Chevron licensed 3-D seismic amplitude data were then further refined using data over GC 955
and one km south in GC 999 provided by JIP member Western Geco. Impedance and gas hydrate saturation models were
produced from the Western Geco 3-D volumes using techniques as described in Dai et al. (2009)11. Other seismic-derived 3-
D volumes produced in the gas hydrate modeling process include interval velocity, Poisson’s ratio, Vp/Vs, compressional
impedance, shear impedance, and gas hydrate saturations derived from both compressional impedance and the shear
impedance volumes. In addition to providing an opportunity to test the direct indicators of gas hydrate as determined through
the seismic analyses, GC 955 provided opportunities to test for the occurrence of gas hydrate that may not be well imaged in
seismic data, including highly heterogeneous, tabular, and steeply inclined occurrences related to fault zones, or gas hydrate
of low or transitional concentrations that resulted in lack of strong seismic reflectors.

Figures 3 and 4. Phase stability diagram for GC 955 (left) and, Perspective view from 3-D seismic of keystone faults on SE-NW
traverse and high amplitudes suggesting mixed gas hydrate and gas along the projected freeze line and relationship to channel
facies. Traverses are coincident with the Q and I wells (right).
4 OTC 20801

The GC 955 Gas Hydrate Petroleum System

The nine-square mile GC 955 lease block is approximately five miles basinward from Green Canyon proper, a reentrant
along the Sigsbee Escarpment, approximately 146 miles south of Fourchon, Louisiana (Figure 1). The Sigsbee Escarpment
is the seafloor expression of a large Jurassic allochthonous salt nappe that mobilized and extruded into and carries shallower
sediments basinward by gravitational processes (Amery, 1969 12; Rowan, 1999 13). The Sigsbee Escarpment, in general,
represents the seaward limit of salt extrusion, but two salt bodies are seaward of the Sigsbee in the vicinity of GC 955. One
of them, Green Knoll, is a large salt diapir, once thought to be isolated (Swiercz, 1991)14, but now understood to be
connected to the mother salt by a salt stock (Kendall, personal communication). Green Knoll is approximately 8 to 10 miles
in diameter with 1,850 ft of relief relative to the surrounding seafloor, approximately six miles to the east of GC 955 (Figure
5). The other salt body is an allocthonous, hourglass-shaped (in plan view), shallow salt body trending southwest to northeast
from WR 30 to GC 955 that is connected by a salt stock to the deeper mother salt beneath the Sigsbee front to the northeast
(Kendall and Pilcher, personal communications) (Figure 5). The shallow salt body has uplifted sediments, forming keystone
faults that extend within a few hundred feet of the seafloor in southern GC 955 (Figure 6). The salt bodies in the local area all
strongly contribute to faulting and impact the fluid flow dynamics to the target gas hydrate sands.

Figures 5 and 6. Green Canyon re-entrant on the Sigsbee Escarpment showing Green Knoll, hourglass-shaped salt pillow beneath
GC955 and WR30, and extent of 3-D seismic data (left), and Perspective view from 3-D seismic of seafloor amplitude showing JIP
wells and evidence of fluid flow to seabed (right).

Conventional oil and gas exploratory efforts have identified petroleum systems in the GC 955 block that resulted in three
wells, GC 955 #001, GC 955 #002, and a sidetrack delineation well from the GC 955 #002 location. GC955#001, drilled by
Statoil USA in 1998, was an exploratory well targeting potential hydrocarbons in the Plio-Pleistocene sediments forming
traps against salt. Hydrocarbons, if found, were not in quantities deemed to be commercial and the well was plugged and
abandoned. Kerr-McGee, later to become part of Anadarko, drilled the GC955 #2 well in 2006 that discovered oil in sub-salt
mid-Miocene sediments (Anadarko, 2006)15. Further delineation of the reservoir from the GC955 #2 location was completed
by Anadarko in 2008.

Figures 7 and 8. Overlay of Weimer, 1990, Mississippi Fan channel valleys over bathymetry data showing Channel 12h and 12g (in
red) (left) and Weimer, 1990, showing cross-section of Mississippi Fan showing Channel 12 (right).

Gas and fluids sources and migration pathways. Extensional faults extending from the shallow salt body in GC 955
into the gas hydrate stability zone are interpreted to be conduits for fluids, including gas, to move through the shallow
sediments. On southwestern and southern flanks up the uplift, there are three “horseshoe-shaped” escarpments showing
OTC 20801 5

seafloor failure (Figure 6). Orange et al.

(2003)16, showed similar features in the
“Mad Dog” development area 16 miles to
the northeast, and interpreted that they were
caused by internally driven failure by fluid
flow and gravity processes at overpressured
sand cropping out at the base of the Sigsbee
Escarpment face. The amphitheatres form
by seepage along the shortened flow paths
at the aquitard (a layer that impedes but
does not prevent fluid flow) (Orange,
1992)17. A similar mechanism for fluid
flow, albeit on a smaller scale, is
recognized in the “horseshoe-shaped”
seabed morphology and shallow subsurface
sediments in GC 955, that suggest eminent
and recent seafloor failures caused by
internally driven failure rather than by
oversteepening alone. In GC 955, most of
the faults extend to Horizon A, about 315
feet below the seafloor (fbsf) at the crest of
the uplift. Horizon A, acting in a similar
way to the failure models at the Sigsbee
face, also marks the base of the seafloor
failure in the SE quadrant of the lease
block. Heggland (2004)9 noted fluid flow
into the shallow sediments through gas
chimneys identified in the seismic data and
the seafloor failure. A mud volcano at the
seafloor above one of the buried faults is
further evidence that fluids move through
the sediments to the seafloor (McConnell,
2000 8; Heggland, 2004)9. This evidence
for fluid migration in addition to the high-
negative impedance gas and high-positive
Figure 9. Top of channel facies (transparent) showing gas hydrate targets below and
JIP well locations.
impedance gas hydrate anomalies centered above the crest of the shallow salt in the faulted sediments (discussed below)
strongly suggests fluid flow through the shallow sediments. Fluid migration along the faults is also strongly suggested by the
prominent seismic anomalies indicative of gas in the faulted four-way closure that prompted the search for gas hydrates in the
block.

Reservoirs. The shaded relief bathymetry map shows the complex seafloor topography in the tabular salt and mini-
basin province (Figure 1). At lowstand, numerous Late Pleistocene sediment fairways extended from the shelf edge to the
continental slope. Nibbelink (1999)18 shows several channel sands and perched sands on the upper part of the escarpment.
Similar sediment fairways extend from the shelf edge, across the mini-basin province, some through Green Canyon Proper,
past the Sigsbee Escarpment and onto the continental rise. Detailed mapping of the Pliocene and Pleistocene Mississippi Fan
seaward of the Sigsbee escarpment shows sand-prone channels and channel-levee systems interbedded within the thick (>4
km) mud-dominated Mississippi Fan (Weimer, 1990 19; see Figures 7 and 8). The large channel and channel levee system
sands targeted for drilling by the JIP corresponds to Weimer’s channel-levee system 12 of the 17 channel-levee systems (1 is
oldest, and 17, youngest) comprising the Mississippi Fan (Figure 8).
As it traverses GC 955, the top of the channel-levee complex 12 is approximately 1,000 fbsf and the channeled interval,
including channel fill sediments is in excess of 700 ft thick. The prominent late-stage channel is approximately 1.3 miles
across from levee crest to levee crest. Biostratigraphic data collected during drilling of the GC 955 #001 well is not yet
public, but is scheduled to be released in 2010. However, Weimer (1990 19; citing Walters, 1985), estimates the age at the top
of channel-levee system 12 at 0.5 Ma.
Because the most clearly imaged channel axis within this channel-levee system (Figures 2 and 9) occurred off the flank of
the prospective structure (and therefore away from the focus for gas and fluid flow above the shallow salt), the drilling targets
on this structure were assessed to be at increased risk of penetrating lower quality and thinner sands than observed in the GC
955 #001 well (see Hutchinson, et al., 2009)1.
6 OTC 20801

Figure 10. Synthetic gathers showing response of various concentrations of gas hydrate (green) over gas (red) as might be found at
the base of gas hydrate stability. This distinction between gas hydrate overlaying large amount of gas, A, and small amount, B, or
20
no gas, C can be difficult to determine from seismic. From Nur and Dvorkin, 2008 .

Geophysical Indications of Gas Hydrate. In addition to exhibiting indications of gas supply, migration pathways, and
reservoir into and within the GHSZ (all components conducive to hydrate formation), several features are observed at GC
955 which suggest the presence of gas hydrate. First among these is the occurrence of strong amplitudes with clear “leading
peaks” (strong positive-polarity reflectors). Such reflectors, where they exist within the potential GHSZ and within intervals
expected to be sand prone, are highly prospective for high concentrations of gas hydrate (McConnell and Zhang, 2005)21.
Analysis of the seismic data from the block also shows that such reflectors, while highly patchy and discontinuous, are most
common in the area of the closed structure (Figures 2, 4, and 9). A second indication of gas hydrate is the anomalous trend
of apparent gas trapping within the closure. Despite the presence of numerous fault-related pathways and the expected
porous and permeable strata in directly overlying unit, gas (as indicated by strong negative impedance amplitudes) is
restricted below a horizon of relatively consistent sub-seafloor depth with no clear geologic explanation. The occurrence of
this “pseudo-bottom simulating reflector (BSR)” within the sand-prone section suggests that the migrating fluids are forming
gas hydrate at the stability boundary, and restricted further upward gas migration. This phenomena is recognized as being
complex, with numerous high amplitude reflectors of both positive and negative polarity occurring in close association,
suggesting a mixed gas and gas hydrate system within the sand in the four-way closure.

GC 955 Gas Hydrate Prospect Models. The primary gas hydrate prospects identified by the site selection team all
trend near the base of gas hydrate stability in a local area of high fluid flux within or near the four-way closure. In the
prospect model for these locations, fluids oversaturated with gas move vertically along faults and fractures (with a potential
lateral fetch component from fluids moving along the salt face at the Sigsbee escarpment 6 miles to the N and along the salt
stock connecting the shallow salt body to the mother salt to the west) into thick sands spanning the gas hydrate stability field.
This gas then converts to gas hydrate within the sand along a relatively thin layer at the base of the stability zone. Other
potential mechanisms for gas hydrate emplacement, in situ gas methane hydrate production, upward diffusive flux and
exsolution, and concentration by recycling of low concentration gas hydrate in response to sedimentation, were not
discounted but were not considered in the pre-drill determination of the primary drilling targets. Where this zone of basal
hydrate accumulates to sufficient thickness to be seismically resolvable (either due to continued sourcing or due to
subsequent lowering of the BGHS), a strong peak reflector is produced at the top of the highly-concentrated gas hydrate
occurrence (Nur and Dvorkin, 2008)20. This model was considered the most likely case for those prospects within the
structural closure. However, given the interpretation of contiguous sand reservoir extending much further up in the section,
the GC 955 Site afforded the opportunity to test two lower-probability but intriguing hypotheses. In one alternative model,
gas hydrate extends above the peak reflectors but in low or in steadily decreasing concentrations (perhaps tied to reduced and
decreasing upwards reservoir quality, a reasonable possibility for channelized deposits) such that no interfaces of sufficient
OTC 20801 7

impedance contrast are produced. Second, gas hydrate extends above the peak reflectors, but is limited to areas very close to
the primary fault migration plains, hence creating gas hydrate units with high-angle orientations that are not readily resolved
in the seismic data.

Figure 11. Arbitrary line through amplitude volume showing JIP and industry wells with gamma and resistivity logs

Figure 12. Arbitrary line through gas hydrate saturation prediction volume showing JIP and industry wells with gamma and
resistivity logs

Shallow Hazards Analysis. The risks of shallow drilling hazards to personnel during riserless deepwater drilling
operations are negligible. However, geohazards, such as gas and water flows, can require significant time to properly control.
For a project such at the JIP, which is constrained to a set budget and time period, significant lost time translates directly into
undrilled holes and uncollected data. In addition, JIP Leg II deployed one of the most advanced and expensive LWD tools
strings ever assembled. Issues such a loss of borehole stability that can result in damaged, even lost tools, must be very
8 OTC 20801

carefully evaluated and managed. Because all of the potential holes considered by the JIP were to be drilled open hole
(without casing, risers, or blow-out preventers), the options available for management of well problems were limited to the
use of weighted mud and cement. Therefore, to minimize the chances for costly downtime, the JIP ordered a thorough
geohazards and wellbore stability analysis for all potential sites, even though regulations did not require that level of analysis
in every case.

Figure 13. Caliper, gamma, resistivity, density, and hydrate saturation logs at GC 955 I.

In order to provide optimal data throughout the section of interest, each planned well depth extended to 500 ft below the
base of gas hydrate stability. Therefore, one of the principal hazards in GC 955 was the potential for gas flows from free gas
zones in the bottom of the holes. Another key issue was the potential for slight overpressures that can cause water flows
and/or wellbore collapse in the unconsolidated marine sediment. The pre-drill hazards analysis evaluated all potential holes
in GC 955 for these risks and moved or eliminated potential locations as required. This work was particularly critical for
those location on the structural high were free gas indicators (strong “trough” amplitudes) were commonly associated with
(directly underlying) the most promising gas hydrate indicators. However, due to the weak sensitivity of seismic amplitude
to free gas saturation, it is not possible to confidently determine if such paired peak-over-trough signatures represent gas
hydrate over water or gas hydrate over free gas (see Figure 10; after Nur and Dvorkin, 2008) 20. The hazards work therefore
assumed such events represented potential gas flow hazards. The primary hazards mitigation strategy was to locate wells
where the “trough” amplitudes could be avoided or penetrated in locations with minimal downdip reservoir extent. Overall,
eleven possible locations in GC 955 were ranked for shallow hazards. Generally, those with the less prospective hydrate
potential, such as those with targets well above the inferred BGHS or those located off the structure, away from the strongest
amplitudes, were assessed with the lowest shallow hazard risks.

GC 955 Drilling Results

Three holes, locations I, H, and Q, were drilled at the GC 955 Site (Table 1). The rationale for the selection of these sites
from among the 11 permitted sites is described in Boswell et al. (2009)3. An arbitrary seismic traverse showing the drilled
locations and the industry wells with gamma ray and resistivity log overlays is shown as Figure 11. The predicted gas
OTC 20801 9

hydrate saturations along the same traverse with gamma ray and resistivity log overlays are shown as Figure 12. The same
traverse is shown the results obtained from the three wells drilled during the program are described below:

Table 1: Information on the 3 wells drilled in GC 955

Latitude Longitude Water Depth Total Depth
NAD1927: Clarke NAD1927: Clarke (ft) Feet below rig floor (fbrf)
GC 955-I 27 00 59.5305 N 90 25 16.8928 W 6770 9027

GC 955-H 27 00 02.0707 N 90 25 35.1142 W 6670 8654

GC 955-Q 27 00 07.3484 N 90 26 11.7156 W 6516 8078

Well GC 955-I. The GC 955-I location was drilled off the structural closure in a location proximal to both the pre-existing
GC 955 #001 well and the primary imaged channel-levee axis. The primary target at the I-location was a muted peak
amplitude anomaly 320 ft above the interpreted base of gas hydrate stability. The amplitude was stratigraphically correlative
to a 4 ohm-m resistivity anomaly logged in the #001 well interpreted to represent up to 15 ft of gas hydrate within the middle
of a 520 foot thick sand section (Figures 11 and 12). Given its proximity to both the #001 well and the channel-levee axis,
the I-location was assessed with a high probability of encountering clean sand within the GHSZ. In addition, the location of
the primary geophysical indicators relative to the BGHS and lack of strong amplitudes resulted in a low risk for gas flows.
The pre-drill gas hydrate saturation prediction at the GC955 I hole from the seismic inversion analyses conducted by
WesternGeco ranged from 28% to 69% over a small area around the well location, with most values above 45% (Hutchinson,
et al., 2009) 1.
In general, the quality of the downhole log data acquired from the GC 955-I well was high to good (Guerin et al., 2009)7.
For the most part, caliper data show that the hole was in gauge when logged except for within a sand-rich interval from
~1,240 fbsf to ~1,620 fbsf, which exhibited “wash-outs” on the caliper log measuring about 1-2 inches larger than the drill bit
(8.5 inch hole).
The sediments logged in the GC 955-I well can be generally divided into three major stratigraphic sections, with the
sedimentary section from the seafloor to a depth of ~1,240 fbsf characterized by uniform gamma-ray values of 75 API,
indicative of a mud-dominated sediment. From ~1,240 fbsf to 1,620 fbsf, gamma-ray log values are highly variable with
values ranging as low as 20 API indicative of an interbedded shale and sand section with individual sand beds ranging in
thickness from about 1-2 ft to as much as 10 ft and greater. This sand-prone section consists of three primary sub-units: a
basal unit denoted by generally decreasing gamma-ray, a 35-ft thick shale-rich middle unit containing only a few thin sands
(from 1,535 fbsf to 1,570 fbsf), and an upper sand section with sharp basal and a generally shaling-upward structure. The
section below ~1,620 fbsf appeared to be dominated by fine-grained sediments to the bottom of the hole. The resistivity log
from the GC 955-I well is very uniform over the entire well with values ranging from ~0.7 ohm-m in the sand units to 1.5 to
~2.0 ohm-m in the shale sections (Figure 13). For the most part these low resistivity log values indicate the presence of
water-bearing sediments (i.e., no gas hydrate or free-gas). The only evidence of elevated resistivity (3 to 5 ohm-m) occurs
within a few feet of sand within the middle shaley portion of the sand unit. These elevated resistivities directly correlate with
high velocities of 2290 m/s and 2190 m/s. Upon removal of the tool string from the hole, a small flow of water was
observed at the seafloor by the Q4000’s ROV which required nearly a day of effort to control (Collett et al., 2009) 2.

Well GC 955-H. The GC 955-H well targeted an anomalous seismic peak over strong trough in the interpreted sand facies.
The location is on the flank of the structural closure, in a location downthrown and to the east of a large normal fault (Figure
14). The location was assessed with a high probability of encountering sand in the GHSZ and a moderate risk for free gas
just below the gas hydrate target. Seismic inversion analysis indicated high potential for gas hydrate of at least 50%
saturation in an area at least 1,500 ft by 2,800 ft (0.12 sq miles) in size. The H location targeted the strongest amplitudes,
corresponding to a pre-drill gas hydrate saturation prediction from seismic inversion analysis of at least 95% (Hutchinson et
al., 2009) 1.
Drilling at GC 955 H revealed a shale-rich section extending from the seafloor to a depth of ~1,275 fbsf. Within the
upper shale section, resistivities gradually increase to about 1.5 ohm-m at 625 fbsf. From that depth to ~ 965 fbsf,
resistivities are highly variable, ranging up to 4 to 10 ohm-m (Figures 15 and 16). Elevated compressional velocities of 1905
m/s correspond with the 3.6 ohm-m resistivity anomaly between 703 and 730 fbsf within and above a thin sand interbed.
Between 887 and 957 fbsf, the 6 to 10 ohm-m shaley section does not correspond with anomalous compressional velocities,
although the velocities appear to be elevated above trend. Azimuthal resistivity images indicate the occurrence within highly-
inclined resistivity fractures indicated grain-displacing gas hydrate occurrence (Guerin et al., 2009) 7. No clear evidence was
found to support the hypothesis that gradually decreasing gas hydrate might be present in sands above the target peak
reflector but the highly inclined resistive sections correspond to intervals where several seismically-imaged faults are
interpreted to intersect the borehole.
10 OTC 20801

Figure 14. Arbitrary line through amplitude volume showing faults and GC955 H with gamma and resistivity logs.

From 975 fbsf (the base of the fractured-filling gas hydrate zone) to the top of the sand section at 1,275 fbsf, resistivity
varies across a range from 1.5 to 3 ohm-m. Where there are thin sands, between 997 and 1010 fbsf, elevated compressional
velocities of 1917 m/s correspond with the 3 ohm-m resistivity. From 1,275 fbsf to 1,600 fbsf, gamma-ray values are
generally lower, indicating a 325 foot thick sand section. The sand is highly-gradational at both top and base. Resistivity in
the unit is highly variable. For the upper 80 ft of the unit, resistivities steadily decline to ~0.8 ohm-m. There is then an
extremely sharp contact, despite lack of evidence for any major lithologic variation, and resistivities climb to ~20 ohm-m or
greater and compressional velocities sharply increase from 1,675 m/s to 2,458 m/s. For the next 86 feet (from 1,358 to 1,444
fbsf), resisitivities range from ~20 to ~200 ohm-m corresponding with high compressional velocities between 2,540 and
3,110 m/s (Figure 15). Gamma values indicate that the unit consists of numerous thin (1 to 3 ft) interbeds in this section
(Figure 15). At 1,444 fbsf, resistivities again abruptly drop to as low as 0.5 ohm-m indicating a sharp change in pore fill,
again with no clear lithologic control. This apparently water-bearing interval extend to 1,457 fbsf – at that level resistivities
again abruptly increase and vary 6 to 30 ohm-m with corresponding high compressional velocities of 2,650 m/s. This second
resistive interval is ~9 feet thick, and directly underlain by another low-resistivity unit ~4 feet thick. At 1,475 fbdf, a third
resistive unit (~10 to 15 ohm-m), ~3 feet thick, corresponding with high compressional velocities of 2,400 m/s, was logged.
From the base of this unit to the base of the sand at ~1,600 fbsf, resistivities are low, ranging from 0.7 to 1.5 ohm-m.
Formation velocities measured with both LWD acoustic tools showed very high velocities, up to 3,000 feet per second, in the
resistive intervals (Figure 15). The section is predominately fine-grained from 1,600 fbsf to the bottom of the hole.
This combination of high resistivity, high compressional velocity indicates likely high concentrations of gas hydrate pore
fill (Figure 16). Significant additional work will be required to understand the nature and controls on this occurrence,
particularly the alternating occurrence of gas hydrate and water pore fill within an apparently single sand reservoir. In
addition, there was no indication of any appreciable volume of free gas in the ~100 feet of sand reservoir underlying the
deepest gas hydrate occurrence in this well. Consequently, the strong trough event noted on the seismic data was driven
largely by the juxtaposition of hydrate over water.

Well GC955-Q. Based on the drilling results at the GC955-H well, the science team elected to drill a third well at the GC
955 site. The selected location is very near the structural crest of the closure, and marks the highest stratigraphic and
structural position with strong leading-peak seismic anomalies at this site. Expected depth to base of the stability zone was
not well constrained in this location, however, the site was thought to have the potential for thick and possibly multiple gas
hydrate bearing zones. As at the GC955-H well, this leading peak is accompanied by strong trough reflections as well, and
OTC 20801 11

Figure 15. Gamma, resistivity, and sonic logs at GC 955 H.

given the structural location, was assessed as an even greater gas flow risk in the pre-drill hazard analyses (Figures 2, 4, and
11). In addition, the section shows numerous strong amplitudes, both peaks and troughs, below the primary target horizon.
However, the findings at the H-well, as well as the Q4000’s ability to pump heavy muds, lead to a decision among the project
team to drill the location.
From the seafloor to a depth of about 1,320 fbsf, gamma-ray values were variable but generally low, indicating a mud-
rich section with minor thin silts and sands. Resistivity throughout this interval steadily increased from ~1 ohm-m in the
shallow section to ~2 ohm-m near the base of this unit. Despite the well crossing a major fault zone, and the general
proximity to the thick fractured-filling hydrate occurrence in the H well to east, the Q well shows little evidence of gas
hydrate in the upper mud-rich sediments.
.
12 OTC 20801

Figure 16. Caliper, gamma, resistivity, density, and hydrate saturation logs at GC 955 H.

From ~1,320 fbsf to ~1,440 fbsf (the deepest obtained gamma-ray reading), gamma-ray values steadily decreased with
depth, suggesting the top of a fining-upward sand section. The interval from ~1,360 to ~1,410 fbsf consists of thinly-bedded
sands with significant clay content that exhibit resistivity of ~1.5 ohm-m suggesting modest, if any, gas hydrate fill. A tight
interval (as noted on the density log 10 feet in thickness occurs from 1,410 to 1,420 fbsf , and below this, porosities and
resistivities (up to 10 ohm-m) abruptly increase. From 1,420 fbsf to the deepest resistivity reading at 1,454 fbsf, resistivities
are highly-variable (from 3 to 10 ohm-m, with a few thin spikes to 20 ohm-m) indicating thinly-laminated sand intervals
similar to those observed at the H location (Figure 17). The lowest tool on the tool string, the MP-3 acoustic tool, shows
clearly elevated velocities in the sand section in direct correspondence to the profile of the resistivity curve (Figure 17),
indicating the presence of gas hydrate, but below this, the tool shows a complex velocity structure that may indicate that free
gas was penetrated below the gas hydrate. Elevated velocities (greater than 2,150 m/s) are recorded to a depth of 1,466 fbsf,
the deepest LWD reading acquired in the hole, approximately 35 feet above the drill bit.
While drilling at 1,498, the ROV recorded a large expulsion of gas and sediment from the borehole. Although this event
was very short-lived, the team immediately began to implement the program’s gas flow control protocols, which included
cementing and abandoning the well (see Collett, et al., 2009 2; for more detail). At this time, it is not well understood what
role gas hydrate, free gas, and potential well-bore disturbances may have played in the gas release and flow issues at the GC
955 Q well. Further log studies, such as shear velocities, can determine if free gas was penetrated at the bottom of the hole.

Summary
Drilling at GC 955 was designed to test a range of geologic models for the occurrence and geophysical expression of gas
hydrate in sand reservoirs. The initial well drilled (GC 955-I) tested a muted seismic anomaly well above the base of gas
hydrate stability that appeared to correspond to the stratigraphic level of a potential 20 ft-thick gas hydrate occurrence in a
nearby, pre-existing well. By being close to the axis of the major sand dispersal fairway in the region, and being away from
the strong indicators of focused flow and gas charging on the closed structure in the southwestern corner of the block, the I-
location maximized the opportunity to encounter sand in the GHSZ while minimizing potential drilling hazards. The well did
OTC 20801 13

encounter thick, high-quality sands, but these sands were fully water-saturated. Only minor indications of gas hydrate were
seen.

Figure 17. Caliper, gamma, resistivity, density, and hydrate saturation logs at GC 955 Q.

The second well (GC 955-H) tested compelling direct geophysical indicators for gas hydrate (high-amplitude reflection
packages of appropriate polarity) on the flanks of the closed structure and removed from the channel axis. Therefore, the site
offered highly prospective targets with only slightly elevated geologic risk (potential lack of reservoir), but significantly
greater risk of drilling hazards. The well encountered ~100 feet of highly-saturated gas hydrate in sand reservoirs, in close
agreement with pre-drill predictions, and with no indications of free gas. Unexpectedly, the location also featured a thick
fracture-filling gas hydrate occurrence in the overlying muds. In detail, the gas hydrate occurrence at GC 955-H poses many
intriguing science questions related to the nature and existence of multiple apparent gas hydrate and water contacts within a
single sand unit. The third well (GC 955-Q) tested geophysical targets at the crest of the structure, maximizing the potential
for gas hydrate fill. Geologic risk at the well was again elevated, as the location is the most distal from the sand fairway, and
drilling risk, while still significant, had been mitigated by the findings at the H well. The well encountered sand as expected,
again saturated with gas hydrate. However, only the top of the apparently fining-upwards sand unit was logged, and the true
nature and extent of the gas hydrate occurrence is not known. Gas flow issues at the well, presently of unclear origin,
resulted in drilling being halted before the zone could be more fully drilled and logged.
The drilling program at GC 955 confirms the integrated geological-geophysical approach used within the JIP program to
assess gas hydrate occurrence from remote sensing data. The presence of fluid flow, a petroleum system, and potential
reservoir, were primary elements in the exploration model adopted by the JIP for prospecting for gas hydrate accumulations
in sands along the gas hydrate stability boundary. Several secondary hypothesis related to the potential occurrence of poorly-
imaged gas hydrate were tested; however, no gas hydrate was found in sand reservoirs that was not anticipated by the
geophysical analysis. The GC 955 site should provides a wealth of opportunities to advance our understanding of gas hydrate
systems, both the continuing evaluation of the LWD data collected in Leg II, and in future additional data collection
programs. The multi-component acoustic and high resolution electric logs, used here to assess the gas hydrate targets, will
also provide an extraordinary resource for Pleistocene stratigraphic and acoustic studies in the Gulf of Mexico.
14 OTC 20801

Acknowledgements
The authors, with the exception of Emrys Jones, made up only the on-board science party. This project would not have been
possible or successful without the guidance and considerable work of the GoM Hydrate Joint Industry Program executive
board, the site hazards and operational planning team, and the site selection team. The authors specifically thank Rana Roy
of Chevron; Debbie Hutchinson, Carolyn Ruppel, and Myung Lee of the U.S. Geological Survey; Diana Shelander and
Jianchun Dai of Schlumberger; and Zijian Zhang, Brenda Monsalve, Hunter Danque, Brenda Monsalve, Ana Garcia-Garcia,
Jim Gharib, Marianne Mulrey, Brent Dillard, and Adrian Digby of AOA Geophysics. We thank Western Geco for
permission to show the 3-D seismic data.

References
1 Hutchinson, D., Boswell, R., Collett, T.S., Dai, J., Dugan, B., Frye, M., Jones, E., McConnell, D., Rose, K., Ruppel, C., Shedd,
W., Shelander, D., Wood, W., 2009. Gulf of Mexico Gas Hydrate Joint Industry Project Leg II— Green Canyon 955 Site
Selection: Proceedings of the Drilling and Scientific Results of the 2009 Gulf of Mexico Gas Hydrate Joint Industry Project Leg
II. http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
2 Collett, T.S., Boswell, R., Mrozewski, S., Guerin, G., Cook,A., Frye, M., Shedd, W., and McConnell, D., 2009. Gulfof Mexico
Gas Hydrate Joint Industry Project Leg II— Operational Summary: Proceedings of the Drilling and Scientific Results of the 2009
Gulf of Mexico Gas Hydrate Joint Industry Project Leg II. http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
3 Boswell, R., Collett, T.S., Frye, M., McConnell, D., Shedd,W., Mrozewski, S., Guerin, G., and Cook, A., 2009.Gulf of Mexico
Gas Hydrate Joint Industry Project LegII — Technical Summary: Proceedings of the Drillingand Scientific Results of the 2009
Gulf of Mexico GasHydrate Joint Industry Project Leg II. http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
4 Mrozewski, S., Guerin, G., Cook, A., Collett, T.S., Boswell, R., 2009. Gulf of Mexico Gas Hydrate Joint Industry Project Leg II
— LWD Methods: Proceedings of the Drilling and Scientific Results of the 2009 Gulf of Mexico Gas Hydrate Joint Industry
Project Leg II. http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
5 McConnell, D., Boswell, R., Collett, T.S., Frye, M., Shedd, W., Guerin, G., Cook,A., Mrozewski, S., Dufrene, R., and Godfriaux,
P, 2009. Gulf of Mexico Gas Hydrate Joint Industry Project Leg II— Walker Ridge 313 Site Summary: Proceedings of the
Drilling and Scientific Results of the 2009 Gulf of Mexico Gas Hydrate Joint Industry Project Leg II.
http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
6 McConnell, D., Boswell, R., Collett, T.S., Frye, M., Shedd, W., Guerin, G., Cook,A., Mrozewski, S., Dufrene, R., and Godfriaux,
P, 2009. Gulf of Mexico Gas Hydrate Joint Industry Project Leg II— Green Canyon 955 Site Summary: Proceedings of the
Drilling and Scientific Results of the 2009 Gulf of Mexico Gas Hydrate Joint Industry Project Leg II.
http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
7 Guerin, G., Cook, A., Mrozewski, S., Collett, T.S., Boswell, R., 2009. Gulf of Mexico Gas Hydrate Joint Industry Project Leg II
— Green Canyon 955 LWD Operations and Results: Proceedings of the Drilling and Scientific Results of the 2009 Gulf of
Mexico Gas Hydrate Joint Industry Project Leg II. http://www.netl.doe.gov/MethaneHydrates/JIPLegII-IR/
8 McConnell, D. R., 2000. Optimizing deepwater well locations to reduce the risk of shallow-water-flow using high resolution 2-D
and 3-D seismic data: Proceedings, Offshore Technology Conference, Houston, Texas 1 4 May 2000, OTC 11973.
9 Heggland, R., 2004. Definition of geohazards in exploration 3-D seismic data using attributes and neural-network analysis,
AAPG Bull. V. 88, No. 6 (June 2004), p. 857-868
10 Nagihara, S. and M. A. Smith, 2008. Regional overview of deep sedimentary thermal gradients of the geopressured zone of the
Texas-Louisiana continental shelf, AAPG Bulletin, V. 92, No. 1 (January 2008), P. 1-14
11 Dai, J., Snyder, F., Gillespie, D., Koesoemadinata, A., Dutta, N., 2008. Exploration for gas hydrates in the deepwater, northern
Gulf of Mexico: Part I. A seismic approach based on geologic model, inversion, and rock physics principles. Journal Marine and
Petroleum Geology, (25), pp. 830-844.
12 Amery, G. B., 1969. Structure of Sigsbee Scarp, Gulf of Mexico, AAPG Bulletin, V. 58, No. 12 (December 1969),P. 2480-2482.
13 Rowan, M. G., M. P. A. Jackson, and B. D. Trudgill, 1999. Salt-Related fault families and fault welds in the northern Gulf of
Mexico, AAPG Bulletin, V. 83, No. 9 (September 1999), P. 1454-1484
14 Swiercz, A. M., 1991. Seismic stratigraphy and salt tectonics along the Sigsbee Escarpment, southeastern Green Canyon region,
in Geyer, R. A. and M. Ashwell eds., CRC handbook of geophysical exploration at sea: hydrocarbons: CRC Press, Boca Raton,
Fl., 496 p.
15 Anadarko, 2006. Press release, Anadarko strikes 250 ft payzone in US Gulf: Mission Deep,
http://www.oilandgasinternational.com/departments/exploration_discoveries/dec06_anadarko.aspx
16 Orange, D. L., D. Saffer, P. Jeanjean, Z. Al-Khafaji, G. Riley and G. Humphrey, 2003. Measurements and modeling of the
shallow pore pressure regime at the Sigsbee Escarpment: successful prediction of overpressure and ground-truthing with
borehole measurements, Proceedings, Offshore Technology Conference, Houston, Texas 5 8 May 2003, OTC 15201 Rowan, M.
G., M. P. A. Jackson, and B. D. Trudgill, 1999. Salt-Related fault families and fault welds in the northern Gulf of Mexico, AAPG
Bulletin, V. 83, No. 9 (September1999), P. 1454-1484
17 Orange, D. L. and N. A. Breen, 1992. The effects of fluid escape on accretionary wedges – seepage force, slope failure, headless
submarine canyons, and vents, Journal of Geophysical Research, vol, 97. no. B6, p. 9277-9295
18 Nibbelink, K., 1999. Modeling deepwater reservoir analogs through analysis of recent sediments using coherence seismic
amplitude, and bathymetry data, Sigsbee Escarpment, Green Canyon, Gulf of Mexico, The Leading Edge, May, 1999, p. 550
19 Weimer, P., 1990. Sequence stratigraphy, facies geometries, and depositional history of the Mississippi Fan, Gulf of Mexico,
AAPG Bull. V. 74. No. 4, p. 425-453
20 Nur, A. and J. Dvorkin, 2008. Seismic-scale rock physics of methane hydrate – Final report, prepared by Stanford University for
the National Energy TechnologyLaboratory
21 McConnell, D., and Zhang, Z., 2005. Using acoustic inversion to image buried gas hydrate distribution: FITI newsletter, Fall,
2005,p.3-5,http://www.netl.doe.gov/technologies/oilgas/publications/Hydrates/Newsletter/HMNewsFall05_HighRez.pdf#page=3