Copyright 2008, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Kuala Lumpur, Malaysia, 35 December 2008.

This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society Committees
of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum 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
where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.


Abstract
Gas hydrates are recognized as a massive source of fossil fuel that could be far in excess of conventional hydrocarbon
resources. The evaluation of formations that contain gas hydrates is therefore receiving renewed emphasis through
contemporary petrophysical technology. A key factor is the use of logging-while-drilling (LWD) to sense hydrate-bearing
intervals before drilling-induced thermal invasion and thence hydrate dissociation take hold. Recent advances in LWD
technology have brought most of the potentially diagnostic tools onto the drill string, so there is little disadvantage in not
having a wireline database. Moreover, modern tools have a sharper spatial resolution and a greater capability for differential
depths of investigation. Petrophysical models have to be capable of distinguishing hydrates from ice in permafrost regions:
this complication does not exist subsea. In general, pristine hydrates are characterized by high resistivity, low sonic transit
time, and low density, possibly in conjunction with gas shows from mud logs. High neutron porosity can also be diagnostic
away from permafrost. Other tools with a role to play include dielectric tools, for distinguishing ice from methane hydrate;
electrical imagers, for identifying the mode of hydrate occurrence; and magnetic resonance tools, for contributing to estimates
of hydrate volume by difference, because of hydrate invisibility. These matters are important, because a hydrate-supported
structure will not produce as well as a framework-supported structure due to pore collapse with dissociation. The evaluation
process is guided by some laboratory measurements of recovered and synthetic hydrates and sand/hydrate mixtures. The
methodology is illustrated by field examples. Although formation evaluation for gas hydrates remains largely semi-
quantitative, current interests are driving towards interpretation protocols that target estimates of producibility. Indicators are
provided as to how this objective might be best approached.

Introduction
Gas hydrates or clathrates were discovered in 1810 by the British chemist Sir Humphry Davy. They were first sampled in
deep ocean sediments in the Black Sea (Yefremova and Zhizhchenko, 1974) and subsequently through the Deep Sea Drilling
Project (DSDP) (Kvenvolden and Barnard, 1983). Today, they are recognized as being present in all continents. Indeed, it has
been estimated that gas hydrates contain hydrocarbon resources, typically methane, considerably in excess of the global fossil-
fuel energy base. Their potential energy value has assumed a high profile, because it brings together issues of Earth, sea,
environment, climate and society.
Because it is technically difficult and expensive to sample preserved gas hydrates for laboratory measurement, well logs
are especially important in the evaluation of gas-hydrate-bearing intervals. The use of well logs to evaluate gas hydrate
formations has received attention over the past 25 years in a manner that was naturally conditioned by the logging technology
of the day. Mathews (1986) set out the log signature of methane hydrate using information from the DSDP and the North
Slope of Alaska. Subsequent analyses include Prensky (1995), Collett (1998a), Lovell et al. (2002), and Murray et al. (2005).
This paper provides a contemporary overview of the petrophysical evaluation of formations that contain gas hydrates with
a view to their exploitation. The description adopts the terminology of the Petroleum Resource Management System (PRMS)
published by the Society of Petroleum Engineers and others (March 2007). Thus, for example, the term hydrocarbon
volume quantifies a gas volume in place referred to standard conditions of 60 F and one atmosphere. On the other hand, the
term resources relates to that portion of an in-place volume (at standard conditions) that is considered to be recoverable and
that may be undiscovered or discovered. Yet again, the term reserves relates to resources that are discovered, remaining (in
the subsurface), recoverable, commercial, and for which there exists a development plan.


IPTC 12610
Petrophysical Evaluation of Gas Hydrate Formations
Paul F. Worthington, SPE, Gaffney, Cline & Associates
2 IPTC 12610
Nature of Gas Hydrates

Gas hydrates are naturally-occurring crystalline inclusion compounds. They comprise compressed molecules of gas (usually
methane) that are lodged within a solid lattice of water molecules. For this reason, the gas molecules are called guests and
the water molecules are called hosts. Gas hydrates form where there are sources of water and methane under favourable
thermodynamic conditions of relatively high pressure and low temperature. There are three molecular structural types.
Structure I (from biogenic gases) forms with a predominance of lighter hydrocarbons such as methane. Structure II
(thermogenic) forms with a predominance of heavier hydrocarbons such as propane and butane. A third structure, H, contains
still heavier hydrocarbons (Sloan, 1998). Because over 99% of naturally occurring gas hydrates are believed to be methane
hydrates, Structure I will be implicit here. Structure II hydrates have been encountered where hydrates overlie petroleum
reservoirs (e.g. the Caspian Sea). Structure H hydrates have been found occasionally (e.g. in the Gulf of Mexico).
The hydration reaction of methane (CH
4
) is given by:

CH
4
+ N
hyd
H
2
O = CH
4
N
hyd
H
2
O (1)

where N
hyd
is the hydration number, the molar ratio of water reacting with methane. For methane, the nominal hydration
number = 6.0 (Moridis and Sloan 2007). Between 80% and 100% of lodging sites are occupied by guest molecules
(Mathews, 1986).

Occurrence of Gas Hydrates
Gas hydrates occur in (permafrost) continental environments and in shallow marine sediments beneath deep water on outer
continental margins. Fig. 1 explains the physical differences between methane hydrates in a permafrost zone and those below
this zone or elsewhere. The former can be at relatively low pressure and very low temperature (e.g. 4.5 MPa at -6 C); the
latter are at higher pressure and low temperature (e.g. 12 MPa at 2 C up to 17.5 MPa at 15 C. Global estimates of methane
gas in place within marine sediments have been up to 100 times greater than those for continental occurrences. Methane
hydrates can occur as massive layers, as thin interlayers, as structural nodules or discrete grains within sediments, or
distributed within the pore spaces of silts and sands, especially where porosity is high. The mode and scale of occurrence do,
of course, have profound implications for the approach to petrophysical evaluation. One volume unit of (massive) methane
hydrate (under conditions of full hydration) equates to 164 volume units of methane at standard conditions of 60 F and one
atmosphere (0.1 MPa).

Classification of Gas Hydrate Deposits
Several schemes have been proposed for describing different occurrences of gas hydrate. Here, they are brought together by
referring to scale.

Microscale. This is the pore scale. It covers a range from pore imaging (0.1 100 m) to core-plug analysis (30 mm). It has
been prognosed that hydrates can grow within a pore network in four identifiable ways (e.g. Kleinberg and Dai 2005):

at grain-to-grain contacts (contact cementing model);
as a uniform coating over the grains (grain-coating model);
as a structural combination with mineral grains (matrix-supporting model);
within the free pore space away from mineral grains (pore-filling model).

Although the distribution of hydrates within sediments has been seen to be fairly uniform (Booth et al. 1996), the hydrate-
growth mechanism has profound implications for petrophysics. For example, the grain-coating model changes the perceptions
of a wetting phase. In addition to this scheme, hydrates can exist as millimetre-scale laminae, within laminar sediments, or as
fracture-fill.

Mesoscale. This is the scale of whole-core measurements (up to 300 mm). It is below the spatial resolution of many logs, i.e.
a log may not record the true parametric value of a 300-mm layer (after environmental corrections) because there is no
position in the wellbore at which the logging tool sees the target layer to the exclusion of adjacent layers. However, such a log
may detect the layer through an apparent parametric value. The thickness of identified gas-hydrate sedimentary sections varies
from a few centimetres to a few tens of metres (Collett 1993). It may not be possible to resolve through logs a hydrate-bearing
sedimentary section or a subsection (see below). Where such a sedimentary section is very thick, the mesoscale may still be
pertinent if the constituent layers are below tool spatial resolution. In these cases, hydrate petrophysics is impacted by the
same problems that beset the petrophysical evaluation of conventional sand-shale sequences. These same comments apply to a
massive hydrate layer such as that reported by Mathews (1986).

IPTC 12610 3
Macroscale. The lower limit of this scale is the sharpest spatial resolution of conventional logs, e.g. about 300 mm, although
many tools cannot achieve this level of sharpness. The following classification is placed at the macroscale to simplify the
discussion by addressing the problem of hydrate character without sub-resolution complexities. Four classes of gas-hydrate
occurrence have been proposed on the basis of their environmental setting (Moridis and Collett 2003, 2004; Moridis and Sloan
2007). This classification is summarized for a non-permafrost situation in Table 1, which also includes some petrophysical
implications.
Class 1. A two-zone hydraulically-bounded occurrence: upper zone is hydrate-bearing with either free gas (Class 1G) or
water (Class 1W); lower zone contains water and free gas. The zonal interface is the base of the hydrate stability zone. This
interface is associated with the Bottom Simulating Reflector noted in marine seismic surveys.
Class 2. A two-zone hydraulically-bounded occurrence: upper zone is hydrate-bearing with either water or free gas; lower
zone contains movable water. The zonal interface is not necessarily the base of the hydrate stability zone.
Class 3. A one-zone hydraulically-bounded occurrence: the zone is hydrate-bearing with either water or free gas. The
base of the zone is not the base of the hydrate stability zone.
Class 4. Hydraulically-unbounded occurrence: dispersed accumulations of low hydrate saturation within marine
sediments.

Physical Properties of Gas Hydrates
Gas hydrate can co-exist with free gas, water and/or ice (permafrost regions only) (Fig. 1). Some physical properties of
hydrates have been re-compiled by Prensky (1995).

Petrophysical Data
Conventional methods of evaluating gas accumulations are not applicable to gas hydrates. In permafrost zones, it is difficult to
distinguish between the log signatures of hydrate-bearing deposits and sediments containing ice. Standard formation tests are
not possible without decomposing (thawing) the hydrates so that flow can occur: thus, physicochemical equilibrium has to be
disturbed in order to assess recovery. Otherwise, formation permeability is very low and this inhibits the recovery of any free
gas that might co-exist in the pore system. Moreover, the relative volumes of hydrate gas and free gas have proved difficult to
estimate from core analysis. In the face of the above complexities, the basic tools available to the petrophysicist remain mud
logs, core analysis, well logs (wireline and logging-while-drilling (LWD)), and repeat formation tests.

Mud Logs. Mud logs indicate related free gas but will not show hydrate unless there has been dissociation. A comparison of
gas shows with well logs (preferably run on the drill string) should allow a qualified assessment to be made of the source of
the gas. Where the gas appears to originate from the hydrate stability zone, this does not necessarily imply dissociation,
because some free gas can exist here. Gas shows immediately below the base of the hydrate stability zone are especially
significant (see below).

Core Analysis. The overarching aim of hydrate core analysis has been to assess gas hydrate concentration through methane
mass-balance calculations based on the depressurization of cores recovered while maintaining in-situ pressure. This objective
is even more poignant given that the complexity of naturally occurring hydrates makes it very difficult to replicate them in the
laboratory (Schultheiss et al. 2008). Core recovery uses a pressure core barrel that can maintain, say, a one-metre length of
sample at reservoir pressure. Coring can be undertaken with a cooling additive that keeps the drilling mud below 2 C and a
non-invasive gel that reduces the effects of filtrate invasion. Once removed from the wellbore, the sample is maintained at
high pressure and low temperature until it reaches the laboratory autoclave, where it is removed from the core barrel and
transferred. The drawbacks of this approach are that it is expensive, there may be some dissociation during the transfer
process, and the (unlined) core is subject to hydrate dissociation and degassing in the autoclave.
Recent advances have introduced a core autoclave that can be decoupled from the main (percussion or rotary) coring tool
after recovery (Schultheiss et al. 2008). The core remains under pressure and cooled by ice as it is transferred to specialist
chambers for analysis, subsampling and storage. A core logging chamber allows physical properties to be measured through a
window in a temperature-controlled environment. Because the core is in a pressure-maintaining steel housing, only a
limited number of measurements can be made at present. For example, density (through gamma-ray attenuation) and P-wave
velocity can be measured but resistivity cannot. The core can also be allowed to dissociate allowing reliable estimates of the
quantity of methane produced, provided that porosity and pore-water chemistry are known. This is a key calibration for the
estimation of hydrates in place.

Well Logs. The primary thrust of core analysis has not yet resulted in the degree of laboratory groundtruthing that would be
expected in conventional hydrocarbon petrophysics. This is because it has not been possible to establish in the laboratory
those interpretative algorithms that will guide the quantitative interpretation of well logs. Moreover, in unconsolidated oceanic
sediments, there are hole instability problems and potentially deep invasion, so wireline log responses can be seriously
degraded. As a result, different logs have provided different insights into hydrate occurrence. For example, the one-metre
core of massive hydrate recovered at DSDP Site 570 off the Pacific coast of Guatemala was interpreted as having a thickness
of 4 m (sonic log), 2.7 m (density log) and 0.6 m (laterologs) using equipment of 1982 vintage (Mathews 1986; Laherrere
4 IPTC 12610
2000). The situation should be improved through logging-while-drilling (LWD), which was not available in 1982, and
through the advances made in sharpening the spatial resolutions of standard logs so that they are more consistent. Fig. 2
shows a suite of logs through a sequence of hydrate-bearing intervals with different levels of hydrate manifestation.

Formation Tests. Gas can be produced from hydrates through three methods of induced dissociation: depressurization,
thermal stimulation, and the use of hydration inhibitors such as salts and alcohols (Moridis et al. 2006). Formation testers
measure the flow of gas so, if this is to reflect the recoverability of hydrate as opposed to free gas, their deployment will be
post-dissociation.

Petrophysical Evaluation
There are three key issues. The first relates to the amount of hydrate that is present within a hydrate-bearing interval. The
second concerns the mode of occurrence of hydrate within that same interval. The third is concerned with the nature of the
immediately underlying beds, specifically whether these are impermeable or of reservoir quality and, if the latter, whether they
contain water and free gas. These three factors are set within a culture that seeks to progress in-place hydrocarbon volumes to
(recoverable) resources and thence (commercially recoverable) reserves.

Bulk Volume of Hydrate. The overriding issue is whether or not the downhole temperature-pressure regime is conducive to
hydrate formation and, if so, what solids and fluids might fill the pore space. This information is discernible from geothermal
and hydrostatic pressure gradients through Fig. 1. Another important issue is whether a hydrate-bearing layer can be resolved
by conventional logs. A layer is resolvable where a logging tool at the centre of the layer records a true parametric value for
that layer after environmental corrections. If hydrate-bearing layers are not resolvable, the problem becomes one of meso-
scale occurrence and therefore it is far more complex. In this case, a three-component induction approach affords the most
promising way forward. This aspect will not be discussed further here. A third important factor is whether there are any core
data available. In the following discussion, it will be assumed that there are not, but indications will be given of additional
options if there had been.
Where a hydrate-bearing layer is resolvable, petrophysical evaluation is directed at estimating the fractional volume of
hydrate within the reservoir rock. This quantity is the product of hydrate saturation and porosity. Hydrate tends to form where
porosity is high ( 0.40), where grain size has a silt and/or clay component (< 62.5 m) (Booth et al. 1998) and where
formation-water salinity is low (e.g. < 25000 ppm NaCl equivalent) (Ginsburg 1998). Under these conditions, hydrate
saturations can be high (0.5 0.9), and at these higher levels there is a corresponding increase in the rigidity of the host rock.
I ce. There are two possible scenarios: the hydrate is located within permafrost or it is not. Within permafrost, there is a
need to distinguish between hydrate and ice. The high-frequency dielectric log is the best way of doing this. However,
dielectric measurements are shallow and, therefore, there should be no thermally-induced dissociation of the hydrate near the
wellbore at the time that the logs are run. This suggests that an LWD approach be favoured, a principle that should be
extrapolated to all tools. On the drill string, resistivity is sensed through a propagation measurement, and this can be processed
to furnish an estimate of dielectric constant with specialist input.
Shale. The fractional volume of wetted shale V
sh
is required if it is intended to use the neutron or sonic logs for the
evaluation of porosity in shaly intervals. This quantity is calculated from a shale indicator such as the gamma ray log.
Inferred values of V
sh
are notoriously unreliable and they can overestimate the requisite parametric value by a factor of four
(Worthington, 2008). Corrections can only be made if core data are available. Given the complexity of the petrophysical
problem in the evaluation of gas hydrates, it is recommended that the total porosity model be used instead. This model does
not use V
sh
in the determination of porosity from well logs (see below).
Porosity. In the total porosity system of petrophysical interpretation the electrochemically bound water associated with
clay minerals is included within the porosity. The dry clay mineral fraction is assumed to have the same density as the rock
matrix. The density log is used to evaluate total porosity as though the formation were lithologically clean. As part of this
process, it is necessary to estimate the collective density of the pore-filling substances, i.e. some or all of ice, hydrate, water
and gas. In the absence of core data, this process is tenuous. It is therefore recommended that some core be cut in a key well,
even though core data can be expected to be less reliable than usual without special precautions (Schultheiss, 2008).
Logistically, this recommendation calls for a logging-while-coring approach, which might not be achievable given the need for
pressured cores. The total porosity method can break down where hole conditions are poor or where the rock matrix shows
pronounced depth variations in composition. Under either of those conditions, conventional petrophysics makes recourse to
the effective porosity system, which allows use of density, neutron and sonic logs. (In bad hole, the sonic log is the only
option, but this can fail as a predictor of porosity where the reservoir is unconsolidated.) These logs must first be corrected for
the effect of clay minerals using the quantity V
sh
. Even if this can be done meaningfully, there is now the problem of
representing up to four pore-filling substances in the clean-sand equations for porosity. These matters have been addressed by
Collett (1998b) for the cases of density and neutron logs and by Akihisa et al. (2002) for sonic logs.
An alternative approach, which sets out to simplify identification of the pore-filling substances, uses the fact that hydrate
(and ice) are invisible to magnetic resonance imaging (MRI) logs, which see only water and free gas. By comparing this
information with porosity from core or perhaps from a standard porosity log subject to certain simplifications, an estimate of
IPTC 12610 5
hydrate (plus ice) volumes can be made. For example, Murray et al. (2005) used a combination of density and MRI logs to
infer the proportions of hydrate and water in shaly marine sediments that were assumed to contain no free gas (see below).
More complex situations could require a geomathematical approach involving the simultaneous solution of log response
equations. However, the author considers that a sounder understanding of the petrophysical character of hydrate-bearing
intervals is needed before such methods can be applied with appropriate control. In particular, the density, MRI and sonic
tools are shallow sensing, so any near-wellbore hydrate dissociation will lead to unrepresentative data.

Saturations of Pore-filling Substances. The conventional approach to the evaluation of hydrocarbon and water saturations
involves the use of deep resistivity logs. These logs require that formation water resistivity be known. They allow an
interpretation to be made of water saturation, on the one hand, and the collective fractional pore volume of ice/hydrate/free
gas, on the other. Where it is present, it might be possible to distinguish ice from hydrate using a high-frequency dielectric log
(which performs in a manner that is independent of water salinity but has a shallow depth of investigation). The remaining
problem is therefore to quantify any free gas saturation. MRI logs are potentially useful here, although the MRI porosity will
be net of ice and hydrate, which are not seen. An alternative approach for marine sediments (no ice) is to use a carbon/oxygen
log to infer water saturation.
The problem becomes more complex where the formations are shaly, because shale effects are exacerbated by the
comparatively low formation-water salinity associated with hydrate occurrence and by high saturations of ice/hydrate, which
are seen as insulators. Under these conditions, the Archie equation is unlikely to be appropriate. Some have used shaly-sand
equations to try to overcome this problem, but it is highly likely that these equations, too, will break down where there are
large concentrations of (insulating) hydrates (Worthington, 2006). An appropriate response in such situations is to use a
pseudo-Archie approach to estimate water saturation (Worthington, 2007).
As an example of the pseudo-Archie approach, consider a hydrate occupying 80% of the pore space of a rock of 40%
porosity and of resistivity 40 ohm m (as per the depth of 910 m in the Mallik 5L-38 well (Anderson et al. 2008). It will be
assumed that there is no free gas and that the formation is lithologically clean. Using a regional temperature profile
(Majorowicz and Smith 1999), the temperature at this depth has been calculated to be 9.7 C. Assuming a formation-water
salinity of 25 000 ppm NaCl equivalent (Collett et al. 1999), water resistivity is estimated as 0.35 ohm m. On this basis, the
pseudo-values of the Archie porosity exponent m and saturation exponent n are estimated as 1.92 and 1.85, respectively
(Worthington 2007). These values depart from the conventional values of m n 2, which others have used to evaluate
hydrate volumes elsewhere, even though the formation appears to be clean. These departures are caused by the freshness of
the formation waters within the hydrate stability zone. Overestimates of m and n will lead to an underestimate of hydrate
saturation. Therefore, the use of the pseudo-Archie approach should lead to a better conjunctive use of resistivity, dielectric,
MRI and carbon/oxygen logs in the evaluation of hydrate volumes.

Mode of Hydrate Occurrence. The way in which hydrate forms within a pore network has profound implications for the
mechanical stability of the host rock and, most importantly, how that stability might change if the hydrates dissociate.
Provided that the hydrate-bearing beds can be resolved by well logs, the problem sits at the micro-scale (see above). The sonic
log is potentially the most diagnostic of hydrate growth habit. Model studies have indicated that compressional transit time t
is an indicator of the mode of hydrate occurrence for a given value of hydrate saturation (Kleinberg and Dai, 2005; Murray et
al., 2005). The model studies of Kleinberg and Dai (2005) were carried out for a two-phase system (hydrate and water). The
relationship between compressional wave transit time and hydrate saturation for the four modes of hydrate occurrence are
represented by the type curves of Fig. 4a (Kleinberg and Dai 2005). Fig. 4b shows data derived from well logs of intervals
that are (presumably) below the permafrost zone at the Mallik 5L-38 well: the data indicate a matrix-supporting model of
hydrate growth. The same authors have adopted a similar approach for shear waves, which reinforced the p-wave diagnosis.
This conclusion means that hydrate dissociation will take away some of the structural framework of the rock with implications
for formation stability.
The above method ought to be applied iteratively. Once there are indications of the mode of hydrate occurrence, it might
be considered appropriate to use an alternative method for the evaluation of hydrate saturation, e.g. where a conventional
resistivity algorithm assumed a pore-filling model but the mode of occurrence now appears to be grain-coating.

Nature of Adjacent Beds. This relates to beds above and below the hydrate stability zone. Well logs will indicate whether
hydrates belong to Class 1, 2, 3 or 4. Classes 1-3 are bounded by impermeable rock above and below the hydrate reservoir
zone. Class 4 is not. The difference between Classes 1 and 2 is that the layer beneath the hydrate stability zone contains water
and free gas in the case of Class 1 but only water in the case of Class 2. The separation of the neutron-density logs might be
diagnostic of the free gas, but this effect can be suppressed by invasion or shaliness. The presence of gas might be better
discerned through the use of carbon/oxygen logs, magnetic resonance imaging, or deep resistivity logs.

Implications for Reporting Reserves. Once the hydrate class has been identified, and in place volumes have been estimated,
the next task is to progress towards an estimate of gas resources. This requires a knowledge of formation volume factor and
recovery factor. Formation volume factor is already known for pure methane hydrate that does not contain quantities of CO
2

6 IPTC 12610
and H
2
S. Recovery factor is not known. Depressurization is more effective than thermal stimulation as a dissociation inducer.
There is a need for a development strategy. Because of the unique nature of hydrates, it may be necessary to define a
dissociation area around a vertical well. Indeed, deep resistivity logs have indicated different dissociation radii at different
depth levels from a thermal test at the Mallik 5L-38 research well (Fig. 3). Moreover, the dissociation area can be expected to
vary with the hydrate class. In the case of Class-3 hydrates, it may be necessary to discount offsets. In the fullness of time,
analogs will become available to guide the identification of a meaningful recovery factor.
Class 1. This is the most promising type of hydrate deposit from a production perspective (Moridis et al., 2007a). Because
it is located close to the base of the hydrate stability zone, small changes in temperature and pressure will induce dissociation.
Moreover, the existence of a free gas (as opposed to residual gas) zone beneath the hydrates will ensure some early production.
The Class 1W model represents a possibly water-wet situation in the hydrate zone (depending on the mode of hydrate
occurrence): if so, it ought to be approached using quasi-conventional petrophysical methodologies. The Class 1G model
represents a non-water-wet situation and will require pronounced departures from conventional petrophysical approaches.
Magnetic resonance logs should be able to distinguish between these two cases, and to recognize the gas-bearing interval
below. This is an important observation, because the upper and lower waters in Class 1W could have very different salinities,
and this would complicate an electrical approach. More generally, one might expect three phases (hydrate, water, free gas) in
hydrate zones outside permafrost intervals. Depressurization through depletion of the underlying gas/water layer is
accompanied by some wellbore heating to prevent flow blockage due to the formation of secondary hydrate and ice.
Class 2. Gas production from Class 2 hydrate accumulations have been reviewed by Moridis and Reagan (2007a) for the
particular case of ocean sediments. Magnetic resonance logs should be able to distinguish Class 2 from Class 1 on the basis of
the fluid responses in the layer beneath the hydrate. Again, depressurization with some wellbore heating is favoured. Initial
production will be lower than for Class 1.
Class 3. Strategies for gas production from Class 3 hydrates have been reviewed by Moridis and Reagan (2007b) for the
specific case of subsea sediments. The exercise is inhibited by the low effective permeability of the hydrate layer, which leads
to low rates of dissociation and gas production, and by the absence of an underlying water layer, which could be produced to
induce dissociation above. A possible strategy might be to induce dissociation at the base of the hydrate layer, thereby
creating a weak Class-2 situation. Note that the existence of an underlying seal should be evident from standard logs. Without
that seal, the effectiveness of depressurization is considerably diminished in Class 1, 2 and 3 situations.
Class 4. In these dispersed hydrate accumulations, hydrate saturation is low in the absence of hydraulically bounding
strata. The gas production potential of Class 4 hydrates in marine sediments has been prognosed as sub-economic with the
additional complication of high water production (Moridis and Sloan 2007).

Workflow for Petrophysical Evaluation
Putting these matters together, the key message is to approach the petrophysical evaluation of hydrate-bearing rocks in an
integrated manner. The following notes propose a workflow to this end. The workflow is simplified, because it does not take
account of such factors as wettability and electrical tortuosity, which will have to be accommodated as perceptions evolve.
The workflow does not consider the comparatively rare case of massive hydrates but rather deals with hydrates within
sediments. To simplify the discussion, it is assumed that the hydrate layers do not contain free gas. Where they do, some
additional fluid characterization may be required in the form of carbon/oxygen logs.

Prerequisites. The context is that of a key well. The prognosis is that a key well would have log data only. The logging tools
are on the drillstring. If core data are required, this should be achieved through a dedicated pressure-cored well. The
acquisition of core and log data in the same key well seems inadvisable, given the sensitivity of the target hydrocarbons to
changes in temperature and pressure.
Resolution. Layers should be resolvable. If they are not, a three-dimensional solution should be sought based on the
principles of meso-anisotropy. The decision should be partly based on electrical imaging logs.
Phase conditions. This analysis is based on entering the expected formation temperature and pressure into an appropriate
phase diagram. An important petrophysical issue is whether the evaluation interval is above the base of a permafrost zone.
Physical properties of reservoir components. An inventory of physical properties of all expected components of the
system will include matrix and possibly shale properties.
Hole conditions. Good hole conditions can be accepted where the wellbore is circular in cross-section and shows no
borehole enlargement greater than 2.5 cm on any face, i.e. less than 5 cm on the caliper log. Where hole conditions are good,
all porosity logs can be used. Where they are bad, the sonic log should be used, perhaps in the effective porosity system. The
following discussion assumes that hole conditions are good.

Ice-free hydrate occurrences. From a petrophysical standpoint, there are two types: clean formations and shaly (clay-
mineral-bearing) formations.
Clean formations. Obtain porosity from the density log where rock matrix is fairly uniform, otherwise use the neutron-
density combination. Use the magnetic resonance log to infer hydrate-free porosity
MR
. Hydrate saturation can be established
by difference. Check via a pseudo-Archie screening approach (Worthington 2006), which uses a deep resistivity log and for
IPTC 12610 7
which an estimate of formation-water salinity is needed in the hydrate stability zone. Once the hydrate saturations have been
affirmed, use the sonic log to investigate the mode of hydrate occurrence. Finally, extend the evaluation interval to include
shoulder beds and thence place the hydrate accumulation in one of the designated classes.
Shaly formations. If matrix and clay-mineral compositions are fairly uniform, obtain total porosity from the density log.
Use the magnetic resonance log to infer hydrate-free total porosity: this takes account of clay-mineral-bound water. Hydrate
saturation in the total porosity system can be established by difference. Otherwise, use the neutron-density log combination
and correct the log responses for shaliness using fit-for-purpose shale indicators (Worthington, 2008): the resulting porosity is
an effective porosity. Use the magnetic resonance log to infer hydrate-free effective porosity: this does not take account of
clay-mineral bound water. Hydrate saturation in the effective porosity system can be established by difference. In both cases,
the results should be affirmed using a pseudo-Archie screening approach, which does not require a shale indicator and is
therefore applicable to both clean and shaly formations. In investigating the mode of hydrate occurrence, it is recommended
that this be done twice: without a correction for shaliness and then after correcting for shaliness using a fit-for-purpose shale
indicator. The reason for this is that we are still learning how to use this approach in more complex lithologies, and it is
important to be aware of the prevailing uncertainties. Finally, classification is effected.

Ice-bearing hydrate occurrences. The procedure follows the above, with the exception that the porosity log magnetic
resonance log approach yields a combined ice plus hydrate saturation. These components can be separated using a dielectric
log but, as noted earlier, this is a specialized deliverable. Knowing the dielectric constants of the constituents, it is possible to
separate ice from hydrate and thence infer hydrate saturation from porosity and water saturation.

Discussion
Where the petrophysical evaluation of hydrates is targeted, logs should be run on the drillstring. This strategy minimizes the
risk of hydrate dissociation in the near wellbore region. Petrophysical evaluation must take account of the scale and mode of
hydrate occurrence. Where the thickness of the hydrate layers is below the resolution of drillstring tools, log analysis faces all
the problems presented by conventional laminated systems. Where the hydrate occurrence departs from a pore-filling mode,
traditional saturation models can break down. A pseudo-Archie approach has been applied elsewhere in the presence of a
range of departures from classical assumptions, and this same approach is advocated here. A prerequisite remains the
evaluation of porosity.
Methods for evaluating porosity and saturation are evolving. They are not limited to the hydrate layers themselves,
because of the significance of underlying layers for production scenarios. At this stage in the development of a subject, it is
imperative that different physical models be used to cross-check the resulting petrophysical estimates. This requirement has a
key role in managing uncertainty.
Petrophysics has a philosophy of indirectness: it is not possible to measure directly the properties we wish to quantify but
rather some other property that is related to that which we wish to know. The subject is founded on empiricism with the
support of physical concepts. The key element is that interpretation should be data-driven: the data will often indicate how an
interpretation should be effected. In the case of hydrates, we are not yet at the stage where the databases can contribute fully
to this philosophy. Progress will call for greater investment in pressure sampling of undissociated hydrate-bearing cores for a
wider range of physical measurements in the laboratory under simulated subsurface conditions. The optimum way forward in
log analysis will not be identified until this has been done.

Acknowledgements
The authors interest in gas hydrates originated through his involvement with the Ocean Drilling Program (ODP) from 1983-
1993, during which time he chaired the ODP Downhole Measurements Panel (1986-1992). Special mention is made of Robert
B. Kidd (1947-1996), who encouraged that initial engagement. More recently, gas hydrates have received considerable
attention as a potential unconventional hydrocarbon resource. The author recognizes Timothy S. Collett and George J.
Moridis, whose work has provided a reference for this overview.

References
Akihisa, K., Tezuka, K., Senoh, O. and Uchida, T. 2002. Well log evaluation of gas hydrate saturation in the Miti Nankai-
Trough well, offshore south east Japan. Transactions SPWLA 43
rd
Annual Logging Symposium, BB1-14. Society of
Petrophysicists and Well Log Analysts, Houston, Texas.
Anderson, B.I., Collett, T.S., Lewis, R.E. and Dubourg, I. 2008. Using open hole and cased hole resistivity logs to monitor
gas hydrate dissociation during a thermal test in the Mallik 5L-38 research well, Mackenzie Delta, Canada. Petrophysics
49, 285-294.
Booth, J., Rowe, M. and Fischer, K. 1996. Offshore Gas Hydrate Sample Database, U.S. Geological Survey Open File Report
96.
Booth, J.S., Winters, W.J., Dillon, W.P., Clennell, M.B. and Rowe, M.M. 1998. Major occurrences and reservoir concepts of
marine clathrate hydrates: implications of field evidence. In: Henriet, J-P & Mienert, J. (eds), Gas Hydrates: Relevance to
World Margin Stability and Climate Change. Geological Society, London, Special Publications, 137, 113-127.
Collett, T. 1993. Natural gas hydrates of the Prudhoe Bay and Kuparak River area, North Slope, Alaska. American
8 IPTC 12610
Association of Petroleum Geologists Bulletin 77, 793-812.
Collett, T. 1998a. Well log evaluation of gas hydrate saturations. Transactions SPWLA 39
th
Annual Logging Symposium,
MM1-14. Society of Petrophysicists and Well Log Analysts, Houston, Texas.
Collett, T. 1998b. Well log characterization of sediment porosities in gas-hydrate-bearing reservoirs. SPE Paper 49298,
Society of Petroleum Engineers, Richardson, Texas.
Collett, T.S., Lewis, R.E., Dallimore, S.R., Lee, M.W., Mroz, T.H. and Uchida, T. 1999. Detailed evaluation of gas hydrate
reservoir properties using JAPEX/JNOC/GSC Mallik 2L-38 gas hydrate research well downhole well log displays. In:
Dallimore, S.R., Uchida, T. & Collett, T.S. (eds), Scientific Results from JAPEX/JNOC/GSC Mallik 2L-38 Gas Hydrate
Research Well, Mackenzie Delta, Northwest Territories, Canada: Bulletin of the Geological Survey of Canada 544, 295-
312.
Collett, T., Lewis, R. and Uchida, T. 2000. Growing interest in gas hydrates. Oilfield Review 12(2), Schlumberger, Sugar
Land, Texas.
Ginsburg, G.D. 1998. Gas hydrate accumulation in deep-water marine sediments. In: Henriet, J-P & Mienert, J. (eds), Gas
Hydrates: Relevance to World Margin Stability and Climate Change. Geological Society, London, Special Publications,
137, 51-62.
Kleinberg, R.L. and Dai, J. 2005. Estimation of the mechanical properties of natural gas hydrate deposits from petrophysical
measurements. OTC Paper 17205, Offshore Technology Conference, Society of Petroleum Engineers, Richardson, Texas.
Kvenvolden, K.A. and Barnard, L.A. 1983. Gas hydrates of the Blake Ridge, Site 533. Initial Reports of the Deep Sea
Drilling Project, 76, 353-365.
Laherrere, J. 2000. Oceanic hydrates: more questions than answers. Energy Exploration & Exploitation, 18, 349-383.
Lovell, M., Jackson, P., Gunn, D., Rochelle, C., Bateman, K., Nelder, L., Culshaw, M., Rees, J., Long, D., Francis, T.,
Roberts, J. and Schultheiss, P. 2002. Petrophysical characterisation of gas hydrate sediments. Transactions SPWLA 43
rd

Annual Logging Symposium, CC1-13. Society of Petrophysicists and Well Log Analysts, Houston, Texas.
Majorowicz, J.A. and Smith, S.L. 1999. Review of ground temperatures in the Mallik Field area: a constraint to the methane
hydrate stability. In: Dallimore, S.R., Uchida, T. & Collett, T.S. (eds), Scientific Results from JAPEX/JNOC/GSC Mallik
2L-38 Gas Hydrate Research Well, Mackenzie Delta, Northwest Territories, Canada: Bulletin of the Geological Survey of
Canada 544, 45-56.
Mathews, M. 1986. Logging characteristics of methane hydrate. The Log Analyst 27(3), 26-63.
Moridis, G.J. and Collett, T.S. 2003. Strategies for gas production from hydrate accumulations under various geologic
conditions. Report LBNL-52568, Lawrence Berkeley National Laboratory, Berkeley, California.
Moridis, G.J. and Collett, T.S. 2004. Gas production from Class 1 hydrate accumulations. In: Taylor, C. & Qwan, J. (eds),
Recent Advances in the Study of Gas Hydrates. Kluwer Academic, New York.
Moridis, G.J., Kneafsey, T., Kowalsky, M. and Reagan, M. 2006. Numerical, laboratory and field studies of gas production
from natural hydrate accumulations in geologic media. Paper LBNL 61765, Lawrence Berkeley National Laboratory.
Moridis, G.J., Kowalsky, M.B. and Pruess, K. 2007. Depressurization-induced gas production from Class 1 hydrate deposits.
SPE Reservoir Evaluation & Engineering 10, 458-481.
Moridis, G.J. and Reagan, M.T. 2007a. Gas production from oceanic Class 2 hydrate accumulations. OTC Paper 18866,
Offshore Technology Conference, Society of Petroleum Engineers, Richardson, Texas.
Moridis, G.J. and Reagan, M.T. 2007b. Strategies for gas production from oceanic Class 3 hydrate accumulations. OTC
Paper 18865, Offshore Technology Conference, Society of Petroleum Engineers, Richardson, Texas.
Moridis, G.J. and Sloan, E.D. 2007. Gas production potential of disperse low-saturation hydrate accumulations in oceanic
sediments. Energy Conversion & Management 48, 1834-1849.
Murray, D., Kleinberg, R., Sinha, B., Fukuhara, M., Osawa, O., Endo, T. and Namikawa, T. 2005. Formation evaluation of
gas hydrate reservoirs. Transactions SPWLA 46
th
Annual Logging Symposium, SSS1-10. Society of Petrophysicists and
Well Log Analysts, Houston, Texas.
Prensky, S. 1995. A review of gas hydrates and formation evaluation of hydrate-bearing reservoirs. Transactions SPWLA
36
th
Annual Logging Symposium, GGG1-12. Society of Petrophysicists and Well Log Analysts, Houston, Texas.
Schultheiss, P.J., Holland, M.E. and Humphrey, G.D. 2008. Borehole pressure coring and laboratory pressure core analysis
for gas hydrate investigations. OTC Paper 19601, Offshore Technology Conference, Society of Petroleum Engineers,
Richardson, Texas.
Sloan, E.D., Jr. 1998. Physical/chemical properties of gas hydrates and application to world margin stability and climatic
change. In: Henriet, J-P & Mienert, J. (eds), Gas Hydrates: Relevance to World Margin Stability and Climate Change.
Geological Society, London, Special Publications, 137, 31-50.
Worthington, P.F. 2006. Quality assurance of the evaluation of hydrocarbon saturation from resistivity data. SPE Paper
103075, Society of Petroleum Engineers, Richardson, Texas.
Worthington, P.F. 2007. Petrophysical type curves for identifying the electrical character of petroleum reservoirs. SPE
Reservoir Evaluation & Engineering, 10, 711-729.
Worthington, P.F. 2008. Quality-assured evaluation of effective porosity using fit-for-purpose estimates of shale volume
fraction. Transactions SPWLA 49
th
Annual Logging Symposium, DDD1-12. Society of Petrophysicists and Well Log
Analysts, Houston, Texas.
IPTC 12610 9
Yefremova, A.G. and Zhizchenko, B.P. 1974. Occurrence of crystal hydrates of gases in the sediments of modern marine
basins. Doklady Akademii Nauk SSSR, 214(5), 1179-1181.


Table 1

Hydrate Classification and Petrophysical Implications in a Non-Permafrost Setting

Class Phases Zones Position in Hydrate
Column
Petrophysical Implications

1G (Solid) Hydrate;
(Liquid) Water;
Free Gas
A Hydrate and Free Gas
B Water and Free Gas
Base of hydrate stabilization
zone
Hydrate saturation must be
separated from free gas in A
1W (Solid) Hydrate;
(Liquid) Water;
Free Gas
A Hydrate and Water
B Water and Free Gas
Base of hydrate stabilization
zone
Hydrate saturation does not
have the complication of free
gas in A
2 (Solid) Hydrate;
(Liquid) Water
A Hydrate
B Water
Can be near base of
hydrate stabilization zone
High hydrate saturation with
other pore filling substances
unspecified
3 (Solid) Hydrate Hydrate Within (main part of) hydrate
stabilization zone
ditto
4* (Solid) Hydrate;
(Liquid) Water;
Free Gas
None Dispersed in and around
hydrate stabilisation zone in
the absence of an integral
cap rock
Low hydrate concentrations;
higher water saturations
A above; B below; * present only in marine sediments (Moridis and Sloan 2007)

































Fig. 1 Phase diagram for a system of pure water and pure
methane. (Re-drawn from Collett et al. 2000)
-10 0 10 20 30 40
100
50
10
5
1
0.5
0.1
Methane
Gas + Ice
Methane Hydrate
+ Ice + Gas
Methane Hydrate
+ Water + Gas
Methane
Gas + Water
I
c
e
-
w
a
t
e
r

p
h
a
s
e

b
o
u
n
d
a
r
y
H
y
d
r
a
t
e
-
g
a
s

p
h
a
s
e

b
o
u
n
d
a
r
y
P
r
e
s
s
u
r
e
,

M
P
a
Temperature, C
10 IPTC 12610



























Fig. 2 Open hole well logs from the permafrost zone in the
Northwest Eileen State-2 well. Units C, D and E are
interpreted to be hydrate-bearing. The low density zone
below Unit C suggests that this unit might be classified as
Class 1 or Class 2. (Re-drawn from Collett 1998)






















Fig. 3 Comparison of open hole (RAL5) and cased hole
(CHFR) resistivity logs from the Mallik 5L-38 research well
during a thermal dissociation test. Zones of dissociation are
indicated by the curve separation, most notably around 919 m
depth. (Re-drawn from Anderson et al. 2008)































































800
750
700
600
500
650
550
Gamma Ray
0 API 100
Bulk-Density
1.9 g/cm
3
2.4
Neutron-Porosity
15 (%) 45
Velocity
1.5 km/sec 4.5
Resistivity
1 ohm-m 10000
Unit
E
D
C
B
D
e
p
t
h
,

i
n

M
e
t
e
r
s

B
e
l
o
w

G
r
o
u
n
d

S
u
r
f
a
c
e
930
920
910
900
1 10 100 1000
Resistivity (ohm-m)
D
e
p
t
h

(
m
e
t
e
r
s
)
Attic Gas?
Collar
Collar
P
e
r
f
o
r
a
t
i
o
n
s
CHFR log
RLA5 log
CHFR log
RLA5 log
IPTC 12610 11























Fig. 4 (a) Hydrate growth habit modelled through
compressional wave velocity as a function of hydrate
saturation. (b) Superposition of log data from the Mallik 5L-
38 well, indicating a matrix supporting hydrate occurrence
with stability implications for dissociation. (Re-drawn from
Kleinberg and Dai 2005)

2
3
4
5
0 0.2 0.4 0.6 0.8 1
Hydrate Saturation
C
o
m
p
r
e
s
s
i
o
n
a
l
V
e
l
o
c
i
t
y

(
k
m
/
s
)
Grain Cementing
Grain Coating
Matrix Supporting
Pore Filling
2
3
4
5
0 0.2 0.4 0.6 0.8 1
Hydrate Saturation
C
o
m
p
r
e
s
s
i
o
n
a
l
V
e
l
o
c
i
t
y

(
k
m
/
s
)
Grain Cementing
Grain Coating
Matrix Supporting
Pore Filling
2
3
4
5
0 0.2 0.4 0.6 0.8 1
Hydrate Saturation
C
o
m
p
r
e
s
s
i
o
n
a
l
V
e
l
o
c
i
t
y

(
k
m
/
s
)
Grain Cementing
Grain Coating
Matrix Supporting
Pore Filling
2
3
4
5
0 0.2 0.4 0.6 0.8 1
Hydrate Saturation
C
o
m
p
r
e
s
s
i
o
n
a
l
V
e
l
o
c
i
t
y

(
k
m
/
s
)
Grain Cementing
Grain Coating
Matrix Supporting
Pore Filling
(a) (b)