Beruflich Dokumente
Kultur Dokumente
This paper was prepared for presentation at the 2000 SPE/CERI Gas Technology Symposium
held in Calgary, Alberta Canada, 3-5 April 2000.
regions. Under the correct pressure and temperature regime,
upward migrating thermogenic or near-surface biogenic
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as methane bonds with water to form hydrates. Stable methane
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to hydrates are found at the temperature and pressure conditions
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at that exist near and beneath the sea floor where water depths
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper exceed 1,000 to 1,600 feet [ft] (300 to 500 meters [m]).
for commercial purposes without the written consent of the Society of Petroleum Engineers is Hydrates are also stable in conjunction with permafrost at high
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 latitudes.
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. Methane hydrates are cages of water molecules that
surround and trap methane molecules. They are crystalline
solids that form under moderate pressure and at temperatures
Abstract
that are low but above the freezing point of water. According
Methane hydrates represent a potentially enormous supply of
to a recent paper by Makogon et al.,3 there are three
natural gas, assuming the technology can be developed for
morphological types of hydrate crystals: massive, whiskery,
commercial gas production from this reservoir type. An
and jelly. Fig. 1 shows a massive methane hydrate and Fig. 2
important part of this required technology is the development
shows a whiskery growth of hydrate crystals. Both of these
of mathematical models for predicting methane production
hydrate samples were formed under laboratory conditions.
rates for various field operational strategies. This paper
presents a review and analysis of ten types of predictive
models for gas hydrates. These models vary from simple
energy balance models, to single-phase analytical porous-flow
models, to robust three dimensional, three-phase numerical
simulators. The models can be broadly classified into
depressurization models, and thermal injection models. Each
of the ten models focused on specific characteristics of the
dissociation, migration, and production of methane from gas
hydrates. The models are compared based on what we believe
are necessary features for a full-featured hydrate simulator.
Introduction
Methane hydrates represent a potentially enormous supply of
natural gas. Utilization of a portion of this resource could help
eliminate concerns of domestic energy shortages in the 21st
century. Published estimates of the total energy reserves
trapped in methane hydrates are speculative and highly varied.
The United States Geological Survey (Kvenvolden 1, 1993)
estimates that the total amount of methane in gas hydrates
around the world is approximately 7 x 105 trillion cubic feet Fig. 1 – Example of massive methane hydrate formed in
[ft3] (2 x 1016 cubic meters [m3]). According to Sassen 2, the the laboratory (after Makogon3, 1999)
resource potential of methane in hydrate exceeds the
combined worldwide reserves of conventional oil and gas
reservoirs, coal, and oil shale by a wide margin.
Most hydrates form offshore in deep ocean sediments or
onshore at shallow to moderate depths in arctic permafrost
2 SAWYER, W.K., BOYER, C.M., II, FRANTZ, J.H., JR., AND YOST, A.B., II SPE 62513
Review of Models
We have reviewed what we believe to be a representative suite
of gas hydrate reservoir simulators that have appeared in the
literature over the past two decades. Obviously, other models
may exist. However, the models reviewed in this study
include the primary production mechanisms proposed for gas
hydrates. In addition, these models include the major physical
and chemical processes involved in the dissociation, transport,
and production of methane from gas hydrates, based on
current understanding of these processes.
In the remainder of this report, we present our detailed
review of ten gas hydrate reservoir simulators. The models
are discussed in a chronological order to provide a historical
perspective.
program which builds upon the existing knowledge base of IMPERMEABLE ROCK
gas hydrates.4
The Federal role provides for, among other things, the HYDRATE ZONE
INJECTION
3300 ft
DISSOCIATED AREA PRODUCTION
WELL
HYDRATE ZONE WELL
ORIGIN x
50 ft
Plan view of fracture-flow production system
z
Finite difference elements used in simulation. Multiple z elements z HYDRATE-FREE “FRACTURE” FLOW PATH
used only for temperature profile FOR TEMPERATURE
PROFILE
The results of this study were as follows: (1) gas hydrates, Representation of the fracture flow path by a series of constant-
width elements (10 elements were used in the model)
if found adjacent to a conventional gas reservoir, may
contribute significantly to total gas production, (2) the
percentage of produced gas coming from the hydrate zone Fig. 5 – Schematic of hydrate dissociation by linked
increases with time, and (3) the gas/hydrate interface gets vertical fracture (after McGuire6, 1982)
colder with time, resulting in a greater heat flux and hence an
increased dissociation rate. It should be noted that water Results from the thermal stimulation models indicate that
production from hydrate dissociation was not considered in hydrate gas production is only economically feasible if (a) the
this model. hydrate reservoir has high in-situ permeability or (b) a
geothermal aquifer exists below the hydrate reservoir to
Model 2 (McGuire6, 1982). Two thermal models and one provide hot brine to the injection wells. To be a candidate for
depressurization model were developed by McGuire6. The thermal stimulation, the hydrate reservoir should be at least
two analytical thermal models included a frontal-sweep model 4.6 meters (m) thick and have a porosity of at least 15%.
(Fig. 4) and a fracture-flow model (Fig. 5), which represented A one-dimensional porous flow model was used to
the upper and lower bounds on hydrate gas production, simulate hydrate gas production by depressurization from a
respectively. Parametric studies were made with these two hydraulically fractured well. A salt-fracture technique was
models to determine the importance of porosity, bed thickness, proposed to establish and maintain a high-permeability, ice-
injection temperature, and hydraulic fracture length. free flow path in the porous media itself (as opposed to
creating a propped-open crack in the rock). A viscosified
“super-saturated” brine using CaBr2 and/or CaCl2 was used to
attack the ice and hydrate on both sides of the hydraulic
fracture (Fig. 6).
PRODUCTION
WELLS
HYDRATED SEDIMENT
DISSOCIATED
DISSOCIATION FRONT
AREA
HYDRATE FREE SEDIMENT
INJECTION
WELLS
UPPER BOUND
NO SOAK
PRODUCED REGION T IN SITU
TS
permeability = k1 AFTER INJECTION
TEMPERATURE
DISSOCIATION FRONT
TEMPERATURE
FINAL
HYDRATED SEDIMENT
T ADIABATIC permeability = k2
TI
TO
DISTANCE FROM WELLBORE INITIAL HYDRATE NEW HYDRATE
FRONT FRONT
CAP ROCK Wellbore heat losses and heat losses to strata above and
below the hydrate zone were considered. It was assumed that
a fracture treatment would create a sufficient void volume to
permit initial steam injection. It was also assumed that the
permeability of the rock in the planned dissociated zone was
sufficient to permit constant steam injection and fluid
DISSOCIATED
HYDRATE
production rates.
ZONE
The results of this study indicated that steam cycling might
be feasible as a gas hydrate recovery scheme. Fig. 9 shows
results as a function of porosity, depth, and thickness of the
BASE ROCK
hydrate zone.
Schematic of the cyclic steam injection process
RESERVOIR DEPTH, FT
OVERBURDEN
ENEGY EFFICIENCY RATIO
9
3250
2800
HYDRATE
LOWER BOUND
2500
8
HYDRATE - GAS
INTERFACE
GAS
7 RESERVOIR
POROSITY: 0.3
THICKNESS: 100 FT
6 UNDERBURDEN
0 100 200 300 400 500 600
ELAPSED TIME (DAYS) Gas reservoir with hydrate cap
10
UPPER BOUND
INCREASE IN
ENERGY EFFICIENCY RATIO
SALT
CONCENTRATION
INCREASING PRESSURE
8
100
ZONE THICKNESS, FT
THERMAL STIMULATION
HYDRATE
50
6
FREE GAS
DEPRESSURIZATION AND WATER
Variance of energy efficiency with time and zone thickness General hydrate phase diagram illustrating dissociation mechanisms
Fig. 9 – Energy efficiency ratios for methane recovery Fig. 10 – Hydrate dissociation mechanisms and schematic
from hydrates by cyclic steam injection (after Bayles7, of gas reservoir with hydrate cap (after Burshears8, 1986)
1986)
A radial geometry was assumed with a single well at the
center, producing at a constant gas rate. The temperature at
Model 4 (Burshears8, 1986). In this study, a 3D, two-phase any point on the hydrate-gas interface was assumed to be the
gas/water simulator was used to address fundamental equilibrium dissociation temperature at the local pressure.
questions regarding the feasibility of producing gas from Heat flow to the gas-hydrate interface was modeled by
hydrates when a conventional gas reservoir is in contact with a conduction.
hydrate cap (Fig. 10). Dissociation of hydrates is affected by For a wide range of realistic parameters, the simulations
depressurization at the hydrate-gas interface. The hydrate indicated that hydrate can be dissociated without an extra
dissociates and cools the reservoir until the gas pressure energy source; i.e., the sensible heat of the reservoir provides
reaches the equilibrium value at the new (colder) temperature. the energy necessary for dissociation and the required heat
The resulting temperature gradient causes the flow of heat flux is sustained as production continues. It was found that
toward the gas-hydrate interface. the water evolved during hydrate dissociation did not result in
undesirably high produced water-gas ratios and did not impair
gas flow to a notable degree. It was concluded that
depressurization is a feasible method for dissociating hydrates
6 SAWYER, W.K., BOYER, C.M., II, FRANTZ, J.H., JR., AND YOST, A.B., II SPE 62513
The rate of advance of the dissociation front was found to TIME (min)
be proportional to t-1/2. Results indicated that the dissociation Model match of experimental results, cumulative gas produced
rate is dependent on the porosity of the porous medium, but
essentially independent of permeability. Calculated energy
efficiency ratios ranged from 6.2 to 11.4, which compare Fig. 11 (cont.) – Comparison of experimental and model
favorably to those of Bayles7 (Fig. 9). results for hydrate dissociation by depressurization (after
Yousif10, 1991)
Model 6 (Yousif10, 1990). A one-dimensional, three-phase
Model 7 (Moridis11, 1998). In September 1998, an equation-
(gas-water-hydrate) numerical finite-difference simulator was
of-state CH4-Hydrate module (EOSHYDR) was added to the
developed to simulate the process of gas production from
existing code of the TOUGH2 model. The TOUGH2 model is
hydrate in a Berea sandstone sample in the laboratory by
a state-of-the-art thermal, fluid-flow heat-transfer simulator
means of a depressurization process. The model was validated
which has a wide variety of applications for modeling gas
by comparison with laboratory experiments. Numerical model
hydrates. It combines many advanced features found in
and laboratory results of produced gas and the location of the
industry simulators with the special characteristics of gas
dissociation front were in good agreement (Fig. 11).
hydrates via the EOSHYDR module.
The TOUGH2 model can simulate hydrate dissociation by
16 either thermal stimulation or depressurization. Both
equilibrium and kinetic conditions of hydrate formation and
DISSOCIATION FRONT LOCATION (cm)
14
dissociation are considered. In the gas phase, both the
12
Klinkenberg effect and molecular diffusion are taken into
account. The heat flux terms include both conduction and
10 convection.
Test results are presented for three hypothetical problems
8 that include different mechanisms and strategies for
production from hydrate reservoirs. The results indicate that
6
recovery of methane from hydrates has significant potential,
4 using either depressurization or thermal stimulation.
EXPERIMENTAL DATA
2 MODEL Model 8 (Ahmadi12, 1999). In this study, a one-dimensional
analytical model was used to describe natural gas production
0 from the dissociation of methane hydrate in a confined
0 60 120 180 240
reservoir by depressurization. Pressure and temperature
TIME (min)
distributions are calculated during hydrate dissociation. The
Model match of experimental results, location of hydrate-dissociation interface fluid energy equation was used which includes convective
heat transfer in the dissociated zone. The release of water
Fig. 11 – Comparison of experimental and model results
during hydrate dissociation was not considered.
for hydrate dissociation by depressurization (after
Results indicated that, under favorable conditions, natural
Yousif10, 1991)
gas can be produced from a hydrate reservoir by
SPE 62513 COMPARATIVE ASSESSMENT OF NATURAL GAS HYDRATE PRODUCTION MODELS 7
depressurization. It was found that reservoir permeability in the reservoir fluid system is internally represented as an
affects the rates of convective heat transfer and consequently extra “enthalpy” component, which occurs in each of the three
the gas production rate. Gas production was found to be phases. In this manner, the heat required for phase transitions,
strongly dependent on reservoir pressure, temperature, such as dissociation of the hydrate, is properly accounted for
permeability, and well pressure. in the model. The calculation method also included the Joule-
Thompson effect, which may play an important role in
Model 9 (Masuda13, 1999). A two-phase, gas-water producing hydrate-associated gas.
numerical finite-difference simulator was developed based on The authors of this study concluded the following: (1)
the kinetic theory of hydrate dissociation. The permeability of current thermal simulation tools in standard use in the oil and
the hydrate was assumed to be a function of hydrate gas industry provide the functionality required to model gas
saturation. Heat transfer by conduction and convection was and water production from hydrate and hydrate-capped
considered within the porous media containing hydrate. In reservoirs, (2) modeling the depletion of a hydrate-capped gas
addition, heat transfer to the surroundings by conduction was reservoir through a horizontal well shows the importance of
considered. The Kim-Bishnoi equation 14 is used to determine the Joule-Thompson cooling effect around the wellbore, (3)
the dissociation rate. dissociation of the hydrate cap caused by pressure depletion in
a hydrate-capped reservoir will be slowed down considerably
TEMPERATURE SENSORS (THERMOCOUPLES)
T1 T2 T3
due to cooling at the gas hydrate interface, and (4) the
production of gas from solid hydrates requires large numbers
PRODUCTION
SURFACE of producing wells, handling of considerable water volumes,
and large amounts of energy input.
END PLUG RUBBER SLEEVE CORE CORE HOLDER END PLUG Discussion and Comparison of Evaluated Models
Structure of core holder and location of thermocouples Each of the models reviewed focused on either
depressurization or thermal stimulation to recover gas from
hydrates.
PRESSURE
INDICATOR
The models reviewed have made a significant contribution
DIFFERENTIAL PRESSURE
TRANSDUCER
BACK-
PRESSURE
toward the development of a practical, realistic, hydrate gas
REGULATOR
production simulation model. Each model has focused on
CORE
specific characteristics of the dissociation, migration, and
HOLDER
GAS METER
production of methane from gas hydrates. An exhaustive
AIR
BATH
literature review would undoubtedly reveal the existence of
other hydrate models. However, the models reviewed in this
NO. 1 NO. 2
GAS-WATER
METHANE
RESERVOIR RESERVOIR
SEPERATOR &
BALANCE
study encompass a wide range of capability and applicability
WATER
GAS
RESERVOIR RESERVOIR
FLUID TANK FOR
for predicting hydrate gas recovery.
OPERATING BPR
NO. 1 PUMP
Tables 1 and 2 compare the ten models based on their
primary production mechanism, i.e., depressurization or
thermal. Our study concluded that there are six primary model
NO. 2 PUMP
features that are required for a full-featured, field-scale
Experimental apparatus for hydrate dissociation in Berea sandstone hydrate simulator. These features are :
Fig. 12 – Laboratory configuration and core holder for § Porous flow of fluids
hydrate dissociation studies. § Heat transfer to surrounding rock
§ Dissociation kinetics
The numerical model results were in general agreement
with laboratory experiments using a one foot long Berea § Two-phase gas/water flow
sandstone core in which hydrate dissociation was initiated by § 3D Cartesian reservoir geometry
lowering and controlling the pressure at the outlet end (Fig. § Field-scale, multi-well capability
12).
We compared the six depressurization models with these
requirements and as shown in Table 1, none of the models
evaluated had all of the desired features. Table 2 compares
Model 10 (Swinkels15, 1999). In this study, a 3D, thermal the models reviewed that have a thermal injection capability.
finite-difference simulator developed by Shell Research 16 was Note that the Shell model (Swinkels15) and the Lawrence
used in which reservoir fluids are represented by three phases: Berkley National Laboratory (LBNL) TOUGH2 model
gaseous, hydrate, and aqueous. Each phase may consist of (Moridis11) are full-featured industry models. These models
two hydrocarbon components and a water component. Energy are the result of many man-years of research and development.
8 SAWYER, W.K., BOYER, C.M., II, FRANTZ, J.H., JR., AND YOST, A.B., II SPE 62513
Technology Laboratory, Department of Energy, 16. Regtien, J.J.M., Por, G.J.A., van Stiphout, M. T., and van
Morgantown, WV. der Vlugt, F.F.: “Interactive Reservoir Simulation,” Shell
Research Rijswijk, 1995, SPE 29146.
13. Masuda, Y., et al.: “Modeling and Experimental Studies
on Dissociation of Methane Gas Hydrates in Berea
Sandstone Cores,” Department of Geosystem
SI Metric Conversion Factors
Engineering, University of Tokyo, Japan.
ft3 x 2.831 685 E-02 = meter3 (m3)
14. Kim, H. C. et al.: “Kinetics of Hydrate Dissociation,”
Chem. Eng. Sci. (1987) 42, No. 7, 1645. ft x 3.048* E-01 = meter (m)
15. Swinkels, J.A.M., et al.: “Thermal Reservoir Simulation kcal x 4.186 8* E+03 = joule (j)
Model of Production from Naturally Occurring Gas
Hydrate Accumulations,” paper SPE 56550 presented at cm x E-02 = meter (m)
the 1999 Annual Technical Conference and Exhibition, scc x E-04 = meter 3 (m3)
Houston, October 3-6.
* conversion factor is exact