Anda di halaman 1dari 37

UNIVERSIDAD AUTNOMA DEL

CARMEN
Dependencia Acadmica de Ciencias Qumicas y Petrolera
Ingeniera Petrolera

SPE-180361-MS
Models of Thermal EOR
in Fractured Reservoirs Materia: EOR NFY
V. M. Ziegler, California Resources Profesor: M.I.P Armando Rene Godina R.
Corporation Presented by:
Copyright 2016, Society of Petroleum Engineers
This paper was prepared for presentation at the SPE Western Regional Meeting held in Anchorage, Alaska, USA, 2326 May
2016.
Isaas Canche Canche
Marco Antonio Alonso Martnez

A 23 de Marzo del 2017


Cd. Del Carmen, Campeche
UNIVERSIDAD AUTNOMA DEL
CARMEN
Dependencia Acadmica de Ciencias Qumicas y Petrolera
Ingeniera Petrolera

SOUTH BELRIDGE FIELD


Between Los Angeles and San Francisco, in the San Joaquin Valley, the
Southern Portion of the Central Valley of California, Western United States.

&
ANTELOPE SHALE RESERVOIR
San Joaquin Valley of central California, United States

SPE-180361-MS
Models of Thermal EOR in Fractured Reservoirs
V. M. Ziegler, California Resources Corporation
Copyright 2016, Society of Petroleum Engineers
This paper was prepared for presentation at the SPE Western Regional Meeting held in Anchorage, Alaska, USA, 2326 May 2016.
INDEX
1-ABSTRACT

2-SOUTH BELRIDGE FIELD UBICATION

3-ANTELOPE SHALE FIELD UBICATION

4-GEOLOGICAL INFORMATION FOR BOTH OIL FIELDS (SOUTH BELRIDGE FIELD


, ANTELOPE SHALE FIELD)

5-MODEL DEVELOPMENT

6-MODEL VALIDATION

7-COMPARISON WITH THERMAL SIMULATION

8. APLICATION AND DISCUSION

9-CONCLUSIONS
ABSTRACT
Models for hot-water or steam injection into a fractured shale or diatomite reservoir are developed from existing
analytic models of energy transport and counter-current imbibition.

Radial convective heat flow through a horizontal fracture system is modeled with conductive heat flow into the low
permeability matrix. The flow geometry approximates hot fluid injection into a five-spot pattern. Recovery
mechanisms accounted for in the models include capillary imbibition and thermal expansion. Temperature dependence
of viscosity and interfacial tension are included in the imbibition estimate.

Laboratory data is needed to quantify the magnitude of the imbibition mechanism, which is usually the primary
contributor to oil recovery. Data representative of either the Belridge Diatomite or the Antelope Shale, two giant
fractured oil reservoirs, are used for the model forecasts.

The steamflood model has been partially validated against a large-scale project in the Belridge Diatomite. Using
public-domain information, a reasonable match was obtained between the model and field project over a four-year
injection period. Comparison to conventional thermal simulation was also performed and indicated reasonable
agreement with the steamflood analytical model.
The models have been used to determine the key factors determining the success of thermal recovery in fractured,
low-permeability reservoirs. Steam injection is shown to be superior to hot water injection in heating the matrix.
Key factors enhancing recovery include: reduced fracture spacing, increased matrix permeability and increased
injection temperature.

Model results indicate that steamflood recoveries in excess of 40% of the original oil-in-place (OOIP) may be
achieved by injection in both diatomite and shales containing light oils. Application to diatomites containing heavy
oil also shows good performance.

Successful application in diatomite reservoirs is forecast to be possible in the current low oil price
environment. Economic application in fractured shales, would require a higher oil price because of the
higher well costs and lower oil content relative to diatomite projects.

Because of the significant volumes of remaining oil in place in both the diatomite and shale reservoirs, application
of thermal enhanced oil recovery (EOR) to these resources represents the logical next step in steamflood
development.
SOUTH BELRIDGE FIELD UBICATION

The South Belridge field is located in an active tectonic


setting between Los Angeles and San Francisco, in the
San Joaquin Valley, the Southern portion of the Central
Valley of California, Western United States.

The field is 100 mi (161 km) North of Los Angeles and 40


mi (64 km) West of Bakers field. SouthBelridge field is 12
mi (19 km) northeast of the basin - bounding San
Andreas wrench fault zone, a prominent and important
geologic feature.
ANTELOPE SHALE RESERVOIR
The Antelope Shale is a geologic formation in the san Joaquin Valley of central California, United States
GEOLOGICAL INFORMATION
MODEL DEVELOPMENT

The flow geometry


approximates hot fluid
injection into a five-spot
pattern.

The models developed are


based on previous analytic
models of energy transport
and capillary imbibition. A
schematic of the model
geometry is shown in Figure
2.
Validation of the model concept is substantiated from outcrop studies of the fractured Monterey Shale and
results of steam injection into diatomite. The photo shown below indicates the extent and frequency of iron-
stained halos around conductive fractures. This implies a zone of increased permeability in and around
fracture zones.
Steam injection into the Belridge
Diatomite, similarly provides evidence of a
layered system with higher permeability
intervals.
As indicated, the model consists of an injection well with a sequence of equally-spaced horizontal
fracture or high-permeability zones, with low-permeability matrix between the fractures. In the areal plane, the reservoir is
subdivided into 10 equal-area cylinders. Assumptions made in the calculation sequence are the following:

1. Hot fluid injection is evenly distributed between each repetitive element.


2. The thickness of each repetitive element is equal to total reservoir thickness divided by the number of fracture zones.
3. Hot fluid flow is radial in the fracture zones.
4. Heat flow in the fractures is by convection only; whereas, heat flow in the matrix is by conduction normal to the fracture.
5. Thermal properties of the fracture zone and the matrix are the same.
6. There is no heat conduction in the radial direction in either the fracture or the matrix.
7. The pattern area is treated as an equivalent radial area.
8. Oil expelled from the matrix in each radial segment is assumed to flow through the fracture system to the pattern
producers.
9. Oil expulsion occurs by both thermal expansion of the matrix and its fluids and counter-current imbibition.
10. Thermal expansion results directly from the conductive heat transfer between the hot fluid in the fracture and the matrix
blocks.
11. Only oil is assumed to be expelled from the matrix upon heating.
12. The magnitude of oil recovery by counter-current imbibition needs to be estimated from laboratory measurements and
then scaled to time according to Equation 1
12. The magnitude of oil recovery by counter-current imbibition needs to be estimated from laboratory
measurements and then scaled to time according to Equation 1.
The definition presented in Equation 1 has been found to normalize counter-current imbibition oil
recovery over a broad range of oil viscosities and mobility ratios.13 This is illustrated in the nomograph
shown in figure 5, which is derived from counter-current displacements in strongly water-wet diatomite
cores at low temperature.
Laboratory measurements have shown that oil recovery by counter-current imbibition increases with
temperature.6,15 Data for a fractured diatomite reservoir containing medium gravity oil (24 API) has been
used to develop the following correlation between imbibition recovery and temperature. As indicated, the
incremental recovery, above that obtained by imbibition at reservoir temperature, follows a nearly linear
trend with temperature increase above reservoir temperature. Figures 5 and 6 provide the empirical
methodology for estimating incremental oil recovery by capillary imbibition as a function of temperature
and time.
Tang and Kovscek5 have shown that in systems with different wettabilities, fractional recovery versus
dimensionless time may not collapse into a single curve. This implies that there should be a cos term in the
numerator of Equation 1 to account for changes in wettability, per the Young-Laplace equation.5 We will
assume that the relation in Figure 5 is valid and rely on laboratory data, such as shown in Figure 6, to account
for changes in wettability.

What remains is to estimate the temperature distribution within both the fracture or high-permeability system
and the low-permeability matrix blocks. The energy transport assumptions described previously are identical to
those of Lauwerier9 for the case of hot-water injection and Marx-Langenheim12 for steam injection, modified
assuming radial flow.
The Lauwerier heat transport model calculates the temperature distributions in both the high permeability
interval and the low-permeability matrix, normal to the direction of flow. To calculate the temperature in the
matrix for the steam injection case, we use the following equation from heat conduction.
MODEL VALIDATION
Steamflooding has been piloted in the diatomite formation at the South
Belridge field since 1986. These tests were conducted in a portion of the
formation containing light oil. Beginning in 2011, the pilot was expanded into
a 600-well project. Production and steam injection are shown in Figure 7.

As indicated, project expansion began in July 2011. By August 2015,


injection reached 19 mbspd into 215 injectors. It is assumed that two
injectors are injecting per pattern. The average injection rate per well is
approximately 90 bspd. At this low rate, it is likely that very low quality steam
or hot water is being injected. Incremental oil production in response to this
injection is estimated to be about 3,000 bopd, in agreement with the reported
value. A model forecast was generated using the reservoir data shown in
Table 1 for the Belridge Diatomite, but using the average injection rate
estimated above and with four high-permeability zones over a 1,000
ft injection interval (see Figure 4). A comparison of the model forecast with
project performance is shown in Figure 8.
As indicated, the match between project
and model is reasonable given the
coarseness of the production data and
uncertainty in extrapolating beyond 2015.
A better match may be achievable if
steamflood producers are known and
production is estimated on a per well
basis.

One outcome of the model calculations is


the relative importance of the two
recovery mechanisms: thermal expansion
and counter-current imbibition. At the low
steam injection rate for the Section 29
project, model calculations indicate
heating of the matrix is very poor. At the
low heating rates, overall recovery is low
(~14%) and dominated by thermal
expansion. This is illustrated in the figure
below.
COMPARISON WITH THERMAL SIMULATION
Comparison to conventional thermal simulation was made to improve confidence in the analytical steamflood model
and to assess the validity of the assumptions made in its development. This comparison was made for the Belridge
Diatomite using the properties shown in Table 1. The simulation model represented one-eighth of a 5-spot pattern and
was subdivided into 75 layers; 5 of which were thin, evenly-spaced, high-permeability intervals representing horizontal
fractures. Steam was injected into the model at a rate of 1.5 B/D/AF-CWE. The temperature profiles at 5 and 10 years
of injection are shown
in Figure 10.
As indicated, steam is initially evenly distributed between the fractures. With time however, the majority of
the steam goes into the centrally located fractures. Despite this, the average reservoir temperature
predicted by the model and simulator are reasonably close, as shown in Figure 11.
Forecasts of oil recovery were made with both the numerical simulator and the analytical model. Fluid
properties were representative of a light oil (30 API oil gravity and 5 cp viscosity). Rock properties, such as
relative permeabilities and capillary pressures, were taken from previous simulation studies of diatomite.20 A
comparison of the oil recoveries predicted by the analytical and numerical models are shown in Figure 12.

In general, the simulation shows


delayed heat transport and oil recovery
in the first five years relative to the
analytical model.

The analytical model assumes that oil


displaced from the matrix near the
injector is immediately produced. The
simulator has a lag built in which
accounts for fluid flow in the fracture.
On average, the analytical and
numerical models show similar heat
transport and oil recovery over a 10-
year injection period.
APPLICATIONS AND DISCUSSION
After validation, the analytical models were used to evaluate production response from a set of fractured, low-
permeability reservoirs: the Belridge Diatomite and the Antelope Shale. Properties of these reservoirs are listed in
Table 1 and were obtained either from the literature or regulatory sources.2,6,14,18,21 Key parameters assumed for
the two test cases were the light oil nature of the hydrocarbons (30 API oil gravity and 5 cp viscosity) and the
relatively high porosity values.

The majority of the Belridge Diatomite contains light oil as does the Antelope Shale. There are diatomite intervals,
however, that contain more viscous, heavier oil and these are evaluated as a sensitivity case. Porosities of diatomite
(Opal A crystal structure) are typically in the 0.50 to 0.60 range. Porosities of Antelope Shale range from 0.20 to
0.35. We focused on the higher porosity range which would be characteristic of the shallower, fractured Antelope
Shale (Opal CT crystal structure).
Effect of Injection Fluid

Test cases were run to evaluate the effect of injecting


either hot water or steam at 500 F. Results for the
diatomite reservoir are shown in Figure 13.
As indicated, steam is much more effective than hot water in recovering oil from fractured diatomite. This
behavior was also observed for the fractured Antelope Shale reservoir. The principal reason for this is the
greater reservoir temperature achieved with steam than with hot water. This is illustrated in Figure 14, which
shows produced fluid temperature for either hot water or steam injection.
Wellbore heat losses during hot water
injection significantly reduces the
injection temperature from 500 F to 445
F for the base case.22 Wellbore heat
losses also occur during steam
injection,23 however, these are
manifested in a loss in steam quality, not
a loss in steam temperature. Because oil
recovery is primarily dependent upon
temperature level, as shown in Figure 6,
hot water injection is noticeably inferior
to steam injection in enhancing recovery
from fractured reservoirs. For this
reason, further
evaluation was restricted to steam
injection only.
Effect of Reservoir Properties
Reservoir Type A comparison of steam injection into Belridge
Diatomite and Antelope Shale served to evaluate the effect of
different fractured reservoir types. A plot of oil recovery versus
steam injection duration is shown in Figure 15.
Steamflood performance is good in both
reservoir types, with recoveries exceeding 40%
OOIP following 10-to-12 years of injection.
Recovery is better from the diatomite, however,
because of a greater oil content than the shale
(i.e., Soh 77 ft versus 61 ft for shale).
Recovery Mechanisms There are two recovery mechanisms assumed On a percentage basis, thermal
in the model: thermal expansion and counter-current imbibition. Figure expansion contributes
16 shows the relative importance of each mechanism in contributing to approximately 25% of total oil
total recovery. recovery, with the majority of
recovery coming from counter-
current imbibition. This is
similar to what has been
reported in laboratory tests
which include both thermal
expansion and imbibition.
Because of the time
dependence of imbibition, the
majority of the early production
is by thermal expansion with
imbibition beginning after about
two years. The behavior shown
for a diatomite steamflood
pattern in Figure 16 is
replicated for a
shale steamflood.
Fracture Spacing Because heating of the matrix occurs primarily by
conduction, it is expected that spacing of the horizontal fractures or high-
permeability zones will have a significant effect on oil recovery. This is
confirmed in the figure shown below for a diatomite steamflood. Significant
recovery is achieved when the fracture spacing is 70 ft or less. Similar
results shown in Figure 17 are also observed for a shale steamflood.
Matrix Permeability Theoretical analysis has shown that the
rate of imbibition is dependent upon both the individual phase Similar results shown in Figure 18
mobilities and the total system mobility.13 Since mobility is for Belridge Diatomite are also
equal to the effective permeability divided by viscosity, this observed for Antelope Shale. While
implies that increasing permeability should increase the rate of theory can be used to estimate the
imbibition. The rate of imbibition should be further increased effect of permeability and viscosity
with increasing temperature, since this reduces the phase on the rate of imbibition, no
viscosities. All of these parameters are incorporated into the satisfactory theory is currently
dimensionless imbibition time shown in Equation 1. Results available to determine the amount
shown in Figure 18, confirm the theory. of imbibition which may occur with
an increase in temperature. Similar
to steam flooding in conventional
reservoirs, ultimate recovery must
rely on physical measurement of
residual oil saturations. Therefore,
laboratory measurements such as
those shown in Figures 5 and 6 will
be required to estimate oil recovery
from fractured reservoirs.
Oil Saturation While variations in oil saturation do not have a material effect on
the recovery factor forecasts, they do impact the cumulative oil recovered over
the life of a project. This is illustrated in Figure 19 for a diatomite steamflood
pattern. Similar results are observed for a shale steamflood.
Oil Viscosity As mentioned previously, there are portions of the Belridge
Diatomite and other diatomite deposits that contain heavy oil. The impact of
steaming a diatomite containing more viscous oil should be to extend the
dimensionless time scale (Equation 1) based on the effects of a higher oil
viscosity on oil mobility and the mobility ratio. This is what is observed for
Belridge Diatomite containing viscous 15 API oil.
Injection Rate Increasing injection
rate for a fixed steam temperature
will accelerate the rate of matrix
heating. This is illustrated in the
figure below for a diatomite
steamflood pattern.
Figure 22

Steamflooding in conventional
reservoirs is usually conducted at
an injection rate of 1.5 bbl per day
per acre-ft of reservoir volume. This
rate appears to be reasonable
target for a fractured reservoir. Any
increase beyond this value has a
marginal effect.
Economic Selection To evaluate the
economic viability of steamflooding a
fractured reservoir, economics were run for
9-pattern pilots in both the Belridge
Diatomite and the Antelope Shale.
Reservoir parameters were assumed to be
those shown in Table 1. Five-spot pattern
geometry was assumed for each pilot. Once
steam was estimated to fill the fracture pore
space, a steam rate reduction schedule was
implemented which maintained bottom-hole
steam quality at 35% or higher. Costs were
based on current data from existing thermal
and shale projects. The ratio of capital
spent on the shale pilot was about 2.4 times
that spent on the diatomite pilot; primarily
because of more expensive well costs for
the deeper shale wells. A value of 15 was
assumed for the ratio of oil price to gas
price. With these parameters, the variation
of project internal rate of return (IRR) with
oil price was determined and is shown
below.
As indicated, the diatomite project is
forecast to be economic over the entire
oil price range. Again, this is probably a
result of the favorable oil content
assumed for the diatomite. In contrast,
the shale project is forecast to require an
oil price of about $80/bbl to meet most
economic hurdles (e.g., IRR 12%).
There are significant original oil-in-place
(OOIP) volumes for both diatomite and
shale in California. The diatomite is an
attractive target right now. Once oil
prices rebound, then the shallower
intervals of the Antelope Shale may
represent an additional attractive target
for steamflooding.
CONCLUSIONS
Analytical models of hot waterflooding and steamflooding in fractured reservoirs have been developed
and validated against field tests and thermal simulation. Application of these models leads to the
following
conclusions.
1. Steam injection is superior to hot water injection for enhancing recovery from either fractured diatomite
or shale reservoirs.
2. Oil recoveries in excess of 40% OOIP may be achievable by steam injection into either Belridge
Diatomite or Antelope Shale, with the diatomite exhibiting better performance because of higher oil
content.
3. At reasonable injection rates, thermal expansion is an early contributor to enhanced oil recovery during
a steamflood, with counter-current imbibition dominating subsequent recovery and yielding approximately
75% of the total recovery.
4. Reduced fracture spacing accelerates conductive heating of the matrix, leading to accelerated and
increased oil recovery from fractured diatomite and shale reservoirs.
5. Increased matrix permeability and the effects of increased temperature accelerate recovery by
capillary imbibition.
6. Laboratory measurements are needed to determine the amount of oil expelled by counter-current
imbibition as a function of temperature and time.
2. Oil recoveries in excess of 40% OOIP may be achievable by steam injection into either Belridge
Diatomite or Antelope Shale, with the diatomite exhibiting better performance because of higher
oil content.
3. At reasonable injection rates, thermal expansion is an early contributor to enhanced oil recovery
during a steamflood, with counter-current imbibition dominating subsequent recovery and yielding
approximately 75% of the total recovery.
4. Reduced fracture spacing accelerates conductive heating of the matrix, leading to accelerated and
increased oil recovery from fractured diatomite and shale reservoirs.
5. Increased matrix permeability and the effects of increased temperature accelerate recovery by
capillary imbibition.
6. Laboratory measurements are needed to determine the amount of oil expelled by counter-current
imbibition as a function of temperature and time.
7. Increased injection rate and/or steam temperature accelerates and increases steamflood oil recovery
from fractured diatomite or shale reservoirs.
8. An injection rate of 1.5 bbl/day/ac-ft, the standard for conventional heavy oil reservoirs, appears
to be a reasonable target for fractured reservoirs.
9. Economic review of potential 9-pattern pilots in both the diatomite and shale reservoirs indicates
that the diatomite project would have favorable profit indicators at current oil prices. A shale
project would require an oil price approximately 75% higher to meet most economic hurdles.
10. Based on the estimated large OOIP volumes, fractured diatomite and shale reservoirs may provide
the next phase of steamflooding in California.

Anda mungkin juga menyukai