close

Вход

Забыли?

вход по аккаунту

?

j.jngse.2018.06.038

код для вставкиСкачать
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
Contents lists available at ScienceDirect
Journal of Natural Gas Science and Engineering
journal homepage: www.elsevier.com/locate/jngse
Uncertainty analysis of hydrocarbon in place calculation using 3D seismic
and well data during appraisal stage – Case study of Goldie Field, offshore
Sarawak
T
Mohd Azudin Yusofa,c, Augustine Agia, Afeez Gbadamosia, Radzuan Junina,b,∗, Azza Abbasa
a
Department of Petroleum Engineering, Faculty of Chemical and Energy Engineering, Universiti Teknologi Malaysia, UTM Skudai, 81310, Johor Bahru, Malaysia
Institute for Oil and Gas, Universiti Teknologi Malaysia, 81310, Johor Bahru, Malaysia
c
PETRONAS Carigali Sdn. Bhd. (PCSB), PETRONAS, Kuala Lumpur, Malaysia
b
A R T I C LE I N FO
A B S T R A C T
Keywords:
Uncertainty
Reserve estimation
Hydrocarbon
3D seismic and well data
Cuddy and Harrison-Skelt
Excel spreadsheet and chart
Goldie Field located offshore Sarawak was previously estimated to be less than 200 Bscf of gas initially in place
(GIIP) and well-4 of a major reservoir in the field was classified as a dry hole. In this study, the uncertainty of
hydrocarbon in place of a major reservoir in Goldie Field is estimated with limited seismic and well data to
determine if an oil column is present in the reservoir. 3D Stacking velocity incorporated into well velocity from
check shots were analysed and the best method or equation for time-to-depth conversion was determined. Gross
bulk volume (GBV) was developed using area depth method. Both deterministic and probabilistic calculations
were conducted and analysed with the uncertainties. The plot of true vertical depth (TVD) versus two-way time
(TWT) velocity function derived from seismic interpretation time at well location and velocity function derived
from pre-stack time migration (PSTM) were used to determine the accuracy of the time-to-depth conversion.
Pickett and Hingle graphical solution equation was used to derive the water resistivity (Rw). Log base saturation
height function was calculated using Cuddy's method and Harrison-Skelt's equation. The study has produced new
depth structure maps that shows a highly correlated values up to a regression coefficient of 0.99 and seismic
attributes that assisted in modelling of the reservoir in Goldie Field. Check shot analysis shows that a linear
function Vavg versus TWT and a second order polynomial function for TVD vs TWT are accurate for seismic time
conversion. Revaluation of well logs and other well data has produced new porosity, water saturation, net to
gross ratio and formation volume factor. Well-4 was classified as a dry hole in PETRONAS in-house report, but
the cuddy's plot for R3 sand in well-4 indicates existence of low saturation gas. This study has produced an
estimated GIIP of 360 Bscf for R3 sand and well-4 has penetrated deeper sands previously interpreted to be wet.
The seismic interpretation and petrophysical evaluation showed that Excel graphics and mathematical analysis
improved the quality of the analysis, and the results show good comparison to other specialized software provided by vendors.
1. Introduction
The determination of hydrocarbon-in-place or resource is very important in petroleum oil and gas exploration and production sector. The
success or failure of the project is to some extent dependent on the
initial estimate of reserve that will be developed in the future. This is
because reserve estimation always involves uncertainty. Uncertainty in
reserve estimation is due to the quantity and quality of data available,
and whether the probabilistic or deterministic method is used in calculating the reserves. During exploration and appraisal phase, only
minimal well and seismic data are available. Therefore, the integration
of all subsurface data is crucial in this process. Geologist, geophysicist,
petrophysicist, and reservoir engineers must work together to produce a
reliable geologic or reservoir model.
∗
Corresponding author. Department of Petroleum Engineering, Faculty of Chemical and Energy Engineering, Universiti Teknologi Malaysia, UTM Skudai, 81310,
Johor Bahru, Malaysia.
E-mail address: r-razuan@utm.my (R. Junin).
https://doi.org/10.1016/j.jngse.2018.06.038
Received 14 February 2018; Received in revised form 19 June 2018; Accepted 24 June 2018
Available online 04 July 2018
1875-5100/ © 2018 Elsevier B.V. All rights reserved.
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 1. Geologic province of sarawak basin (Petronas, 1999).
reservoir characterization involves the integration of 3D seismic and
well log data.
Seismic and well log data have been used by Lu and McMechan
(2002) and Wang et al. (2011) to access gas-hydrate reservoirs. They
used the relationship between P-impedance and porosity to interpret
gas-hydrate deposit. Meanwhile, 3D pre-stack inversion and rock
property models were used by Dai et al. (2008a, b) to estimate gashydrate occurrence. The gas saturation predicted from pre-stack inversion was compatible to that calculated from well log data. Yi et al.
(2017) used the deterministic method to estimate the gas-hydrate and
gas-in-place of a small area (12 km × 21 km) from pre-stack 3D
seismic data, log and core data. They concluded that the uncertainties
arising from simplified relationship between rock properties, elastic
attribute, nature of seismic data, non-uniqueness of seismic inversion
and variability in well log data cannot be fully overcome. But it can be
reduced provided rock physics modelling with pre-stack 3D seismic and
well log data are utilized.
Modelling of a reservoir is challenging due to the complexity of the
physics involved in the flow and lack of available data for modelling the
reservoir (Charles et al., 2001; Agi et al., 2017; Gbadamosi et al., 2018).
In most cases, the conceptual model is essential at the scale of the reservoir unit, but their accuracy is not sufficient to predict the distribution of the internal heterogeneity (Haldorsen and Damsleth, 1990;
Oluwadare et al., 2017). The internal reservoir heterogeneity and the
distribution of small scale sedimentary bodies are simulated using the
stochastic approach and geostatistical approach (Haldorsen and
Damsleth, 1990; Massonnat et al., 1993). These options are used to
evaluate the impact of different geological scenarios, which contribute
to the optimization of the field development plan. With advances in
computer technology, reservoir models can now accommodate detailed
3D data that illustrates the spatial distribution of the reservoir properties. The geologic model of the reservoir is established by augmenting
well data with seismic data to characterize the subsurface reservoir
(Patrick et al., 2002; Oluwadare et al., 2017). Therefore, a successful
239
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 2. Map showing 3D seismic outline and well location.
through Brunei into the southern part of Sabah (Scherer, 1980; James,
1984). It was formed during the late Eocene. The basement rock has
been interpreted as either highly deformed, metamorphosed deep
marine shale or turbidites. In some cases, the basement consists of
radiolarian chert, spilite and dolerite. In offshore Sarawak, the growth
fault is mainly SW-NE orientation in the south, becoming more progressively E-W trending in the north. Superimposed late Miocene to
Pliocene regional compressed deformation also took place. The intensity of this deformation is more toward the SE, which resulted to the
formation of a series of NE-SW trending anticlines. These anticlines
obliquely intersect the growth faults and it is at this intersection points
that the major hydrocarbon accumulation is located (Johnson and
Baldwin, 1986). The hydrocarbon accumulation is generally found at
the downthrown side of the growth faults, which is related to the
rollover structures, faults seals, and southerly-directed hydrocarbon
migration routes from more deeply buried downdip kitchen areas
(Johnson and Baldwin, 1986).
Sheriff and Geldart (1995) reported that the major source of uncertainty in volumetric computation is related to uncertainty in timedepth conversion. In the oil and gas industry, specialised and licensed
software are used for this purpose, but they are mostly limited to office
use only, and require high specification computers to run on. Also, the
use of these softwares will almost invariably produce different result
and the differences are compounded by structural and stratigraphic
complexes (Demirmen, 2007). Developing Excel spreadsheet is advantageous for researchers because it is user friendly and can be shared
easily for enhancement. Therefore, the objectives of this study are: (i)
To estimate the hydrocarbon in place of a major reservoir in Goldie
Field and to determine if an oil column is present in the reservoir, (ii)
To analyse 3D stacking velocity incorporated into well velocity from
check shot and to determine the best method or equation for time-todepth conversion, and (iii) To develop user friendly and readily available Excel spreadsheet for gross bulk volume (GBV), formation evaluation and probabilistic calculation with the aim of reducing uncertainties.
3. Methodology
2. Geology and location of the study area
3.1. Data acquisition
The study area nicknamed Goldie Field is in the Baram Deltaic
geologic province of the Sarawak Basin (Fig. 1). The Baram Delta
Province is in the northern part of Sarawak and extends north-eastward
The data used in this study consists of a 3D seismic data (Fig. 2)
(full, near, middle and far angle stack); well log data from four
240
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 3. Depositional environment model (Fui, 1978).
3.2. Data processing
exploration and appraisal wells and technical reports (drilling, petrophysical, geological, PVT, field development plan, seismic acquisition
and processing). The seismic, well log data and technical reports were
obtained from PETRONAS.
The key data used for correlation are the gamma ray and deep resistivity logs, since both data have been logged extensively to total
depth (TD) and show good response in a sand-shale sequence. Other
logs such as sonic, density and neutron were also used to fine-tune the
correlation and their quality depend on the borehole condition.
The geologic or facies model were derived from available well logs
and report. A depositional environment model has been established by
Shell for Sarawak Basin (Fig. 3) and it is still in use. Detailed formation
evaluation on the reservoir produced various input parameters such as
GBV, net-to-gross ratio, porosity and water saturation. Formation volume factor and core data were obtained from available reports. Both
the deterministic and Monte Carlo (probabilistic) analysis were
241
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
3.3. Data interpretation
PETREL software was used for the analysis and interpretation of the
well logs and seismic data. The check shot data loaded in PETREL was
checked against field reports and edited to remove bad data. This was
done for Well-1 (W-1) and Well-4 (W-4) (Fig. 4a and b). This is because
the only available check shot data was for well-1 and well-4. The best fit
check shot generated from W-1 and W-4 was applied to W-2 and W-3
for initial interpretation and synthetic seismogram generation. The
major faults were interpreted manually both inline and crosslines at a
spacing of 400m whereas, smaller spacing was used for the interpretation of minor faults. The fault interpretation was facilitated by
using variance and coherent data generated from 3D seismic (Fig. 5).
After satisfactory horizons and fault interpretation was achieved, fault
polygons for the horizon was constructed and gridded (50 ft × 50 ft) to
produce time structure map (Fig. 6).
3D pre-stack time migration (PSTM) velocity was incorporated into
the well velocity analysis to determine the best method or equation for
the time depth conversion. Three methods were selected to determine
the accuracy of the time-to-depth conversion namely; (a) single function derived from the check shots (Rider, 2011). The plot of true vertical depth (TVD) versus two-way time (TWT) will produce a best fit
second order polynomial; Z = at + bt2 whereas, Z is the TVD and t is
the TWT, (b) velocity function Vavg = Vo+ kt derived from seismic
interpretation time at well location, and (c) velocity function Vavg =
Vo+ kt derived from PSTM velocities. Whereas, Vo is the initial average
velocity at the time zero and k is a constant. It is like Z/t = a + bt
which is derived from a single function Z = at + bt2. The single
function was constructed using Excel (Fig. 4a and b) and the velocity
function derived from seismic interpretation time at well location. The
parameters were derived from each well location and gridded to produce Vo and k grids (Figs. 7 and 8) prior to the development of the raw
depth map. The process can be simplified if an average Vo value be used
instead of a variable grid value (Fig. 9). But it will increase the error at
the flanks of the field in the raw depth map. The velocity function Vavg
= Vo+ kt derived PSTM is the most complicated since the amount of
data is huge and it need to be edited and reduced to facilitate computation (Fig. 10). The original seismic velocity which includes the interval and the average velocity calculated from Dix's equation (Grechka
et al., 1997) is shown in Fig. 11. The original data at a depth of 2200 ms
are ambiguous since the interval velocity is decreasing with depth. It is
possible that overburden pressure is present at deeper levels, but geologically sedimentary rock will be highly compacted or metamorphosed
at high temperature and pressure. It is unlikely that the interval velocity
will show reversal in a large scale. Velocity error above 2200 ms is
likely to be attributed to the difficulty in picking the right stacking
velocity in areas where seismic multiples are dominant. To derive Vo
and k from stacking velocity that can be used for time-to-depth conversion, a suitable time range need to be selected so that a linear
function can be derived with high certainty. When stacking velocity
model is used as an explicit intermediate step in the time-depth conversion, the goal is to derive a robust model that can accurately predict
true vertical velocity at and between wells (Yilmaz, 2001; Etris et al.,
Fig. 4. a Check shot analysis of Well-1.
b Check shot analysis of Well-4.
conducted. Both methods were used because in practice as we move
towards more complex and multiple scenario, or fully defined probabilistic approach, both element of deterministic and probabilistic
methods are incorporated. In a probabilistic framework such as this,
determinism was introduced in form such as correlations between
variables which are not defined statistically by the data. Alternative
structure models requiring different model grids cannot be expressed as
a continuous variable and are rarely accommodated in a probabilistic
approach.
GBV was determined using area-depth method (Passey et al., 2010).
Excel spreadsheet and chart were developed to derive accurately the
GBV for minimum, most likely and maximum cases. The input data
required are the area, depth, gross reservoir thickness, hydrocarbon
column height and spill point (or fluid contact). The geometrical correction was used in the slab method calculation for undrilled prospect
(James et al., 2013), which can be interpreted using 2D seismic data.
Both the deterministic and probabilistic (Monte Carlo) calculations
(Krinitzsky, 2003) were conducted and analysed with the uncertainties.
242
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 5. Variance Cube Time Slice at 2300 ms.
equations (1) and (2) were used to derive Rw; while tortuosity (a) and
cementation exponent (m) were fixed at 0.65 and 2.15 respectively
which are common parameters for sandstone. Porosities were calculated from sonic and density logs and calibrated against neutron log for
each well. The average porosity was then calculated and used for water
saturation calculation (Ellis and Singer, 2007). In the absence of core,
log base saturation height function was calculated using Cuddy's
method and Harrison-Skelt's equations (3) and (4) (Cuddy et al., 1993;
Skelt and Harrison, 1995; Skelt, 1996).
2002). After reviewing the selected data, a time range of 500–2500 ms
is found to be sufficient (Fig. 12). The data were then gridded to produce Vo and k grids (Figs. 13 and 14) prior to generation of raw depth
maps. The resultant Vo and k grids looks erratic but the raw depth maps
are geologically sound with application of minor grid smoothing.
Petrophysical evaluation was conducted on R1, R2 and R3 reservoirs in each well. Input parameters such as gross thickness, net-togross ratio, porosity, and water saturation were derived to be used in
the hydrocarbon in-place calculation (Ellis and Singer, 2007; Krygowski
and Cluff, 2015; Mohamed and Kashlaf, 2016).
log(Rt ) = −m ∗ log(Ø) − n∗ log(Sw ) + log(a∗Rw )
1
(1)
1
1 m
Sw n ⎞ m
⎛ ⎞
=⎛
∗Ø
⎝ Rt ⎠
⎝ a∗Rw ⎠
BVW = 1 − A∗H ˆB
(3)
B ⎞C ⎞
Sw = 1 − A. exp ⎜⎛−⎛
⎟
h
+
D⎠ ⎠
⎝ ⎝
(4)
BVW = Sw * Ø
(5)
(2)
Whereas; Rt is the true resistivity, m is the porosity exponent, n is
the saturation exponent, Sw is the water saturation, Rw is the water
resistivity, a is the tortuosity factor and Ø is the total porosity.
Pickett and Hingle (Hingle, 1959; Pickett, 1973) graphical solution
Whereas, H is the height above free water level (FWL), A, B, C, and D,
are constants found by regression to core and log data.
The parameters used in the petrophysical calculation are shown in
243
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 6. Near base R3 time structure map.
observed, therefore, the depositional environment can be classified as
coastal to fluvial marine inner neritic (Wafa et al., 2013; Siddiqui et al.,
2013). Fig. 15 shows a structural log correlation using the mean sea
level as datum. It shows a correlation for an interval starting from Intra
Cycle-V to Base R3. Unfortunately, well W-2 penetrated 2 faults in this
interval, the missing section contributed to the reduction of the total
thickness which is not real for a geologic model. There is a thickening of
this interval from well W-4 (in the northeast) towards well W-3 (in the
southwest) as seen in the isopach generated from seismic data (Figs. 16
and 17).
Table 1. Although the calculated Rw for each well differs slightly, it
does not jeopardise the quality of the overall evaluation since the
average Sw will be decided based on the saturation height function
(Luthi, 2001).
4. Results and discussion
4.1. Geological and reservoir model
Within the zone of interest, no coal or carbonate layers were
244
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 7. Vo map generated from 4 well points.
Fig. 18 shows a stratigraphic log correlation flattened at the top of
R1, and it show more detailed correlation for an interval starting from
top R1 to Base R3. R1 and R2 sands show thickening in wells W-3 and
W-4, both located downdip of wells W-1 and W-2. However, R3 shows a
different trend of thickening. Thickness increases from well W-4 (in the
northeast) toward W-1 and W-2 (in the southwest). The net to gross
differ slightly in these sands and is generally more than 80%. The
blocky or coarsening upward nature seen in the Gamma ray log suggests
that these sands are coastal sands or offshore bars in a wave dominated
delta system (Siddiqui et al., 2013; Ezeh et al., 2016). Thin mudstone or
tight streaks can be seen from the neutron or density log, but their
occurrence is not extensive and will not become a flow barrier for reservoir fluids.
R2 values and shows that both wells are deviated. The composite check
shot plot (Fig. 19) shows only a small deviation. Therefore, the final
check shot is reliable. The check shot analysis shows that a linear
function for Vavg versus TWT and a second order polynomial function
for TVD versus TWT are accurate enough for seismic time-depth conversion. The 3D display of interpreted fault is shown in Fig. 20. The
field is dominated by growth faults (Rijks, 1981), but the appearance of
reverse faults at the northeast part of the field complicates the subsequent faults and horizon interpretation.
4.3. Time-to-depth conversion
(a) Single Function from Check shots: The near base R3 raw depth
structure map generated using single function is shown in Fig. 21.
The computation of regional correction function was done to tie the
raw depth at each well location against the actual well top depth
(Fig. 22). The raw depth was corrected to the nearest well top to
4.2. Seismic analysis
Results obtained from the check shot editing was good as seen in the
245
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 8. k Map Generated from 4 Well Points.
areas away from well control. The poor performance of the check
shot method could be attributed to the lack of check shot data. Only
2 out of the 4 wells were available, hence, the method using single
function from check shot appear to be less accurate.
(c) Vavg from Stacking Velocity: The near Base R3 raw depth structural
map generated using PSTM velocity is shown in Fig. 27. Fig. 28
shows the corresponding regional correction function using
stacking velocity and the post regional correction depth structure
map is shown in Fig. 29. The regional velocity using stacking velocity is higher than that of check shots or seismic interpretation
time. The error in the post correction maps are also higher but still
within ± 1% (Table 4). This is because the stacking velocity data
are not very precise. The desired information is often buried in
noise and the measurements are highly leveraged (Khan and
reduce error. Top R1 was penetrated in all four wells but Top R2
and R3 were faulted out in well W-2. Therefore, only 3 points are
available for calibration. The corrected Top R3 depth structure map
is shown in Fig. 23. Table 2 shows that the error has reduced tremendously after the application of the first correction.
(b) Vavg Function from Seismic Interpretation: The generated near base
R3 raw depth is shown in Fig. 24. The computation of regional
correction function was done to tie the raw depth at each well location against the actual well top depth (Fig. 25). The post regional
correction of the top depth structure map is shown in Fig. 26. The
error analysis of the post regional correction depth against actual
well top depth is shown in Table 3. The Vo+ kt method performed
better than the check shot method. This is because of less error in
the Vo+ kt method. However, the Vo+ kt method perform poorly in
246
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 9. k Map using Vo = 5964 ft/s, generated from 4 Well Points.
the downthrown side of the major fault. Whereas on the upthrown side,
the quality deteriorates, due to thinning and heterogeneity of the reservoirs. High seismic amplitude is seen in fault block tested by W-1
especially at the R3 interval. However, only weak amplitude is seen at
the vicinity of W-4 well. No structural conformable amplitude was seen
in R1 and R2 sands. Conversely, strong structural conformable amplitude can be seen in R3 far angle amplitude map in the fault block tested
by W-1 well. Qualitative analysis of near, mid and far amplitude maps
shows better contrast than the full stack amplitude map (Fig. 30a, b and
c), which is unusual due to the reduction in fold in the angle stack data.
This might be because of the improvement in signal to noise in low fold
data due to the latest 3D processing technique. The signal to noise ratio
after migration is higher than the ratio before migration. This is because
migration reduces the input random noise. Also, the accuracy of dipping event might have affected the full stack because lateral velocity
changes are assumed to be relatively small and flat. But this is not always true as anomalies might contain dipping components (Wu, 2001).
Another possible explanation is the detailed stacking velocity modelling
which might have improved the imaging of the data through PSTM.
PSTM supress input random noise more efficiently. The migrating noise
Akhter, 2011). Therefore, a small error in source data measurement
can lead to a large error in the output. Hence the difficulty is not in
the conversion but rather in gathering a statistically valid data set
which has adequate details for the purpose (Lindseth, 1982). The
check shot is the most reliable source of velocity information but
are not without errors. The source of error is in failure to recognise
the first arrival time due to excessive attenuation of the signal
transmission. This can result to velocity which is slower than the
actual value (Lindseth, 1982). It is recommended that the stacking
velocity should be calibrated with the check shot data.
4.4. Seismic attributes analysis
Displaying attributes maps with time or depth contours are beneficial for geologist. Geologic facies or hydrocarbon is modelled from
these maps. Good, consistent seismic and fault interpretation are prerequisite for good quality attribute maps. Seismic attribute is usually
reported as the maximum (positive or negative) amplitude value at
each sample along a horizon pick from a 3D volume.
Attributes generated for R1, R2 and R3 sands are better defined in
247
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 10. Location of original and selected stacking velocities.
Fig. 11. Original PSTM velocities.
Fig. 12. Edited PSTM Velocities used in the Study.
background comes from incomplete cancellation in migration, nonetheless, the cancellation is more complete with pre-stack than the full
stack (Wu, 2001; Luo et al., 2018). Usually, only amplitude near the
reflector point will contribute to the signal, others will cancel out with
the contribution from other traces. Such cancellations always leave
migration noise background. Obviously, the distribution of the pre-
stack migrated input is more even than that of the full stack, thereby,
leading to a complete cancellation in the pre-stack. Stacking the converted wave data always introduces extra errors, but with pre-stack
migration it is easy to handle the converted wave data accurately (Wu,
2001; Luo et al., 2018).
248
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 13. Vo map generated from PSTM velocities.
Fig. 14. k Map Generated from PSTM Velocities.
249
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Table 1
Caption: Petrophysical parameters.
Parameters
Values
Clean sand GR
ρ fluid
ρ matrix
Δt fluid
Δt matrix
Tortuosity (a)
Cementation exponent (m)
Saturation exponent (n)
< {GRmin+0.2*(GRmax-GRmin)}
1 gm/cc
2.64 gm/cc
181 μs/ft
55.5 μs/ft
0.65
2.15
2
The porosity distribution for R3 in well W-1 is shown in Figs. 32
and 33. After applying correction (Fig. 34), the three porosities
data points were better aligned in the cross plot. The difference is
not pronounced because the raw sonic and density porosities
4.5. Petrophysical evaluation
(a) Petrophysical Evaluation of R3 sand in Well 1: The raw sonic and
density porosities were plotted against neutron porosities (Fig. 31).
Fig. 15. Geologic correlation of intra Cycle-V to base R3.
250
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 16. Isopach from top R1 to base R3.
in well W-1 is close to that obtained from the Hingle plot (0.35),
thus, validating the results obtained from the Hingle plot for this
well.
(ii) Pickett Plot: Pickett plot devised by Pickett (1966, 1973) is a
simple and very effective method. It plots both the resistivity and
porosity on a logarithm scales (Fig. 35b) as compared to the uncalibrated porosity linear scale (Fig. 31). With resistivity on the xaxis and porosity on the y-axis, it is based on the observation that
Rt is a function of Ø, Sw and m. A straight line (100%Sw ) represents
the wet resistivity (Ro) and the use of Archie's equation is bypassed.
(iii) Cuddy Plot: The plot of BVW against height after FWL (Fig. 36)
plotted against neutron porosities were 90 and 92% respectively of
the neutron porosity values and a correction of 10 and 8% respectively was required.
(i) Hingle Plot: Hingle (1959) proposed a graphical solution to Archie's equation as shown in Fig. 35a. The x-axis shows porosity
increasing to the right while, the y-axis is a non-linear scale of
calculated Hingle value. From the location of the points, a water
bearing line was drawn to represent wet resistivity (Ro). From
Fig. 35a, the intercept of the Ro line with the horizontal axis is
0.23, which represent a*Rw. The value of 0.65 for a gives Rw of
about 0.35 Ω m. The estimated raw value of 0.3 Ω for R3 reservoir
251
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 17. Random seismic lines across W-1 and W-2.
that the seismic amplitudes are related to the reservoir properties
rather than water saturation. The depth maps generated using
check shot and stacking velocity are similar in appearance,
whereas, the map generated using Vo+ kt method derived from
well tops showed huge error away from the wells location. This
error could be attributed to the skills of the seismic interpreter or
the lack of control points. Therefore, only the depth maps using
check shots and stacking velocity were used in calculating GBV
(Fig. 39a and b). Although, oil has been seen in R1 sand in well-2,
the area tested by Well-2 is insignificant to affect the economics of
Goldie Field. Therefore, no further evaluation was conducted for
R1 sand. The Sw was not listed in Table 5 because the average Sw
for the field cannot be accurately estimated on well basis. The
average Sw can be estimated if the hydrocarbon column, permeability, and irreducible water saturation in the formation are
available. The initial water saturation calculated using Archie's
equation (Archie, 1942; Kumar et al., 2002) was corrected using
Cuddy's method (Cuddy et al., 1993) to cater for the height above
free water effect. The Harrison-Skelt function was used to model
the most likely saturation height function (Skelt and Harrison,
1995; Skelt, 1996; Harrison and Jing, 2001) as shown in Fig. 40.
Fig. 41, Tables 6 and 7 show the parameters and results for both
deterministic and probabilistic gas initially in place (GIIP). The
shows a rise and fall trend. The regression line indicates that the
BVW can be related to the height after FWL. Cuddy's plot for R3
reservoir indicates a high reduction in water saturation as height
increases, indicating that the reservoir is a gas bearing reservoir.
This is consistent with the well report and high resistivity values
that was observed in the section. The result also coincides with the
high amplitude in Fig. 30c. The summary of reservoir information
from all the wells is shown in Table 5.
(b) Gas in Place Calculation: The seismic interpretation and petrophysical evaluation indicates that only R1 and R3 sands have been
proven to have hydrocarbons in Goldie Field. The field have been
separated into different fault blocks as indicated by hydrocarbon in
R1 (Well-2), which was not the case in Wells-1 and 4. It is also
possible that R3 sand in Well-1 and Well-4 are in pressure communication, since the possible gas-water contact (GWC) is estimated at −10620 to −10660 ft s depending whether Well-1 or
Well-4 water point is used (Fig. 37). In the calculating GBV for R3
sand in Goldie Field, the top R3 depth maps generated by the
different methods were corrected to fit the well top at Wells-1 and
4 (Fig. 38a and b). The depth contour was overlain by full stack
seismic amplitude maps to determine if high amplitude is related
to the gas column. It is estimated that the GWC is estimated too far
down the dip of the high amplitude. Therefore, it is safer to assume
252
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 18. Stratigraphic correlation from top R1 to base R3.
Fig. 19. Composite Check shot Plot.
Fig. 20. 3D view of interpreted faults.
253
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 21. Near Base R3 Raw Depth Map using Check Shot.
a second order polynomial function for TVD versus TWT are accurate
enough for seismic time depth conversion. Whereas, method using velocity function V0 + kt derived from seismic interpretation time at well
location was the least accurate method. Time-to-depth conversion using
check shot data can be used if no strong lateral changes in velocity
occurs and it is usually used in field or prospect scale interpretation.
Seismic attributes from near, middle and far angle data provide
additional information that can be used to model the reservoir.
However, the seismic attribute map generated in this study does not
show any relationship to direct hydrocarbon indicator. The seismic
attributes map from this study has shown low to moderate noise level
which is above expectation. The concept of incorporating 3D stacking
velocity into limited well velocity from check shot has generated depth
maps that shows a highly correlated values for both wells up to a
GIIP difference between the two maps is about 7%, which is better
compared to the expected difference of 10–20% (De Hemptinne
et al., 1997). The GIIP for R3 reservoir in Goldie Field is estimated
at about 360 Bscf, which is sizeable for a single reservoir.
5. Conclusions
This study has produced new depth structure map and seismic attributes maps that assisted in modelling of the major reservoirs in
Goldie Field. The maps derived using different methods of time-todepth conversion have minimal effect on the calculated GBV. The use of
Vavg = V0 + kt is more practical than Vavg = V0 + kZ since the latter will
V
produce a function of = 1 0 ; which is not a polynomial function. The
t − k
check shot analysis shows that a linear function for Vavg versus TWT and
254
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 22. Regional Correction using Check shot.
Fig. 23. Top R3 regionally corrected depth map (check shot).
255
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Table 2
Caption: Depth conversion error analysis using check shot.
Well Name
Reservoir Top
Well Top
Raw Depth
Error (%)
Regional Correction
Error (%)
W-1
W-2
W-3
W-4
W-1
W-3
W-4
W-1
W-3
W-4
TopR1
TopR1
TopR1
TopR1
TopR2
TopR2
TopR2
TopR3
TopR3
TopR3
−9175.42
−8923.99
−10374.3
−9264.49
−9899.29
−11191.3
−9978.37
−10353.9
−11709.7
−10485.3
−8985
−8740
−10545
9240
−9960
−11625
−10205
−10630
−12500
−11050
−2.08
−2.06
1.65
−0.26
0.61
3.88
2.27
2.67
6.75
5.39
−9145
−8945
−10395
−9355
−9840
−11190
−10030
−10270
−11685
−10545
−0.33
−0.24
0.20
0.98
−0.60
−0.01
0.52
−0.81
−0.21
−0.57
Fig. 24. Near Base R3 Raw Depth Map using Vo+ kt.
256
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 25. Regional Correction using Vo+ kt.
Fig. 26. Top R3 Regional Corrected Depth Map using Vo+ kt.
Table 3
Caption: Depth Conversion Error Analysis using Vo+ kt.
Well Name
Reservoir Top
Well Top
Raw Depth
Error (%)
Regional Correction
Error (%)
W-1
W-2
W-3
W-4
W-1
W-3
W-4
W-1
W-3
W-4
TopR1
TopR1
TopR1
TopR1
TopR2
TopR2
TopR2
TopR3
TopR3
TopR3
−9175.42
−8923.99
−10374.3
−9264.49
−9899.29
−11191.3
−9978.37
−10353.9
−11709.7
−10485.3
−9117
−8885
−10323
−9221
−10030
−11270
−10050
−10682
−12017
−10772
−0.64
−0.44
−0.49
−0.47
1.32
0.70
0.72
3.17
2.62
2.73
−9165
−8930
−10375
−9274
−9930
−11195
−9948
−10362
−11710
−10428
−0.11
0.07
0.01
0.10
0.31
0.03
−0.30
0.08
0.00
−0.55
257
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 27. Near Base R3 Raw Depth Map using PSTM Velocity.
Fig. 28. Regional Correction using PSTM Velocity.
258
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 29. Top R3 Regional Corrected Depth Map using PSTM Velocity.
Table 4
Caption: Depth Conversion Error Analysis using PSTM Velocity.
Well Name
Reservoir Top
Well Top
Raw Depth
Error (%)
Regional Correction
Error (%)
W-1
W-2
W-3
W-4
W-1
W-3
W-4
W-1
W-3
W-4
TopR1
TopR1
TopR1
TopR1
TopR2
TopR2
TopR2
TopR3
TopR3
TopR3
−9175.42
−8923.99
−10374.3
−9264.49
−9899.29
−11191.3
−9978.37
−10353.9
−11709.7
−10485.3
−8796
−8612
−10384
−8955
−9715
−11441
−9868
−10348
−12285
−10658
−4.14
−3.50
−0.09
−3.34
−1.86
2.23
−1.11
−0.06
4.91
1.65
−9127
−8979
−10382
−9271
−9880
−11192
−10008
−10312
−11703
−10523
−0.53
0.62
0.07
0.07
−0.19
0.01
0.30
−0.40
−0.06
−0.36
259
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 30. a R3 maximum amplitude map from near angle stack.
b R3 maximum amplitude map from middle angle stack.
c R3 maximum amplitude map from far angle stack.
Fig. 32. Frequency histogram for R3 porosity in Well-1.
Fig. 31. Uncalibrated porosities for R3 reservoir W-1.
260
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 33. Cumulative distribution function for R3 in Well-1.
Fig. 35. a Hingle plot for R3 reservoir W-1.
b Caption: Pickett plot for R3 reservoir W-1.
Fig. 34. Calibrated porosities for R3 reservoir W-1.
regression coefficient of 0.99. This method has bypassed the source of
error at the initial stage and can be considered for enhancement at the
development stage.
Revaluation of well logs and other well data has produced new
porosity, water saturation, net to gross ratio, and formation volume
factor. The study has shown that porosity derived from neutron, density
and sonic logs in clean sand still have big deviations and need to be
calibrated to get a consistent porosity. In the absence of core data, it is
possible to get consistent porosity values. The initial step is to determine the best matrix and fluid parameter in tight reservoirs where
the porosity is near 0%. Revised Archie's exponent for tortuosity and
cementation exponent, and new saturation height functions calculated
Fig. 36. Cuddy's plot for R3 reservoir W-1.
261
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Table 5
Caption: Summary of reservoir parameters for R3 sand.
Well-1
Top R3
Base R3
Gross Thickness
Net to Gross
Porosity P10
Porosity P50
Porosity P90
Bg
Bo
Est. FWL or GWC
ft ss
ft ss
Ft
Decimal
Decimal
Decimal
Decimal
Cu ft/SCF
RB/STB
ft ss
Well-2
−10354
−10593
239
0.916
0.135
0.158
0.182
0.004
Well-3
Well-4
−11710
−11906
196
0.931
0.097
0.110
0.124
−10485
−10686
201
0.880
0.110
0.128
0.143
1.440
−10620
−10660
Fig. 37. Formation pressure data in R3 sand and adjacent reservoir.
Fig. 38. a R3 depth map (using check shot) overlain on maximum amplitude from full stack seismic.
b R3 depth map (using stacking velocity) overlain on maximum amplitude from full stack seismic.
262
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Fig. 39. a Area-Depth Plot for R3 Sand using Depth Map Derived from Check Shot.
b Area-Depth Plot for R3 Sand using Depth Map Derived from Stacking Velocity.
Fig. 40. Modelled saturation versus height function for R3 sand in goldie field.
have improved the accuracy of the calculated water saturation. Well-4
was classified as a dry hole in PETRONAS in-house report, but the
cuddy's plot for R3 sand in Well-4 indicates existence of low saturation
gas.
This information has produced an alternative gas in place figure but
the existence of substantial oil rim in these reservoirs cannot be proven
from these data. Initially, in-house report estimated the GIIP for Goldie
field to be less than 200 Bscf and more than 400 Bscf. But these figures
were calculated for reservoirs down to R3 sand only. This study has
produced an estimated GIIP of 360 Bscf for R3 sand and well-4 has
penetrated deeper sands interpreted as wet in previous report. The
porosity of these deeper sands might be lower, but the gas expansion
factor was higher. Therefore, another few hundreds Bscf of GIIP is
possible for Goldie Field. The method of using both Petrel and Excel has
improved the quality of the analysis. Besides being more user friendly,
Excel spreadsheet can easily be shared and improved by other users. A
good software is a bonus, the technical skills of the user is more important, especially when dealing with limited or bad data.
Fig. 41. a Probability versus GIIP for R3 sand in goldie field (depth map derived from stacking velocity).
b Probability versus GIIP for R3 sand in goldie field (depth map derived from
check shot).
263
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Table 6
Caption: Input Parameters for R3 Sand used GIIP Calculation (Depth Map Derived from Stacking Velocities).
R3 Sand
GBV
NG (%)
POR (%)
Shc (%)
FVF
X10ˆ6 (cu ft)
Min
ML
Max
Alpha
Beta
18806
21806
24962
1
1
88
91
93
1
1
11
14
18
1
1
45
50
55
1
1
240
245
250
1
1
Min
ML
Max
GIIP (bscf)
GIIP (bscf)
Deterministic
Probabilistic
196.6
340.3
574.6
P95
P90
P50
P10
P5
255.9
271.9
346.2
433.2
457.2
Table 7
Caption: Input Parameters for R3 Sand used for GIIP Calculation (Depth Map Derived from Check Shot).
R3 Sand
GBV
NG (%)
POR (%)
Shc (%)
FVF
X10ˆ6 (cu ft)
Min
ML
Max
Alpha
Beta
20201
23468
26946
1
1
88
91
93
1
1
11
14
18
1
1
45
50
55
1
1
240
245
250
1
1
Acknowledgement
Min
ML
Max
GIIP (bscf)
GIIP (bscf)
Deterministic
Probabilistic
211.2
366.3
620.2
P95
P90
P50
P10
P5
277.3
292.5
373.8
468.5
493.5
Hingle, A.T., 1959. Use of logs in exploration problems. In: Paper Presented at the 29th
International Annual Meeting SEG, Los Angeles, CA, USA, Nov.
James, D.M., 1984. In: The Geology and Hydrocarbon of Negara Brunei Darussalam,
Muzium Brunei, 169.
James, B., Grundy, A.T., Sykes, M.A., 2013. The depth-area-thickness (DAT) method for
calculating gross rock volume: a better way to model hydrocarbon contact uncertainty. In: AAPG International Conference and Exhibition, Cartagena, Colombia,
8–11 September.
Johnson, H.D., Baldwin, C.T., 1986. Shallow siliciclastic seas. In: Reading, H.D. (Ed.),
Sedimentary Environment and Facies, second ed. pp. 229–282.
Khan, K.A., Akhter, G., 2011. Work flow shown to develop useful seismic velocity models.
Oil Gas J. 109 (16), 52–61.
Krinitzsky, E.L., 2003. How to combine deterministic and probabilistic method for assessing earthquake hazards. Eng. Geol. 70, 1–2 157-163.
Krygowski, D.A., Cluff, R.M., 2015. Pattern recognition in a digital age: a gameboard
approach to determine petrophysical parameters. In: Presented at the SPWLA 56th
Annual Logging Symposium Held in Long Beach, California, USA, July 18–22.
Kumar, R., Cherukupalli, P.K., Lohar, B.L., Chandra, D., 2002. Saturation modelling in a
multi-layered carbonate reservoir using log-derived saturation-height function. In:
Paper SPE-75213–MS, Presented at SPE/DOE Improved Oil Recovery Symposium,
Tulsa, Oklahoma, USA, 13-17 April.
Lindseth, R.O., 1982. Digital Processing of Geophysical Data: a Review. Teknica Resource
Development LTD, Calgary, Alberta, Canada.
Lu, S., McMechan, G.A., 2002. Estimation of gas hydrate and free gas saturation, concentration, and distribution from seismic data. Geophysics 67, 582–593.
Luo, D., Wu, Z., Zeng, T., Zhang, X., Zhao, W., Liu, H., Xiao, G., 2018. Velocity anisotropy
characteristics and pre-stack time migration imaging of the marine sedimentary
strata in the south yellow basin. Studia Geophys. Geod. 62, 1–19.
Luthi, S.M., 2001. Geological Well Logs: Their Use in Reservoir Modelling. Springer, New
York.
Massonnat, G., Alabert, F., Guidicelli, C., 1993. Anguille marine, a deep sea-fan reservoir
offshore Gabon: from geologic to stochastic modelling. In: SEG Technical Programme
Expanded Abstract, 345.
Mohamed, A.K., Kashlaf, A., 2016. Comparative study between well logging and core
analysis of hawaz reservoir in murzuq basin, Libya. Egyptian Journal of Basic and
Applied Science 3, 301–313.
Oluwadare, O.A., Osunrinde, O.T., Abe, S.J., Ojo, B.T., 2017. 3-D geostatistical model and
volumetric estimation of ‘del’ field, Niger delta. J. Geol. Geophys. 6 (3), 2–10.
Passey, Q.R., Bohacs, K., Esch, W., Klimentidis, R., Sinha, S., 2010. From oil-prone source
rock to gas-producing shale reservoir and petrophysical characterisation of unconventional shale gas reservoirs. In: Paper SPE-131350–MS, Presented at the
International Oil and Gas Conference and Exhibition, Held in Beijing, China, 8-10
June.
Patrick, D.D., Gerilyn, S.S., John, P.C., 2002. Outcrop based reservoir characterisation: a
composite phyll-id-algal mound, Western Orogrande basin (New Mexico). AAPG
(Am. Assoc. Pet. Geol.) Bull. 86, 780.
Petronas, 1999. Petroleum Geology and Resource of Malaysia, Technology and
Engineering Unit.
Pickett, G.R., 1966. A review of current techniques for determination of water saturation
from logs. JPT (J. Pharm. Technol.) 18 (11), 1425–1433.
Pickett, G.R., 1973. Pattern recognition as a means of formation evaluation. Paper SPESPWLA-. In: Presented at the SPWLA Fourteenth Annual Logging Symposium,
Lafayette, Louisiana, USA, 6–9 May.
Rider, M.H., 2011. The Geological Interpretation of Well Logs, third ed. Rider-French,
Sutherland.
The authors would like to thank the Ministry of Higher Education,
Malaysia, and Universiti Teknologi Malaysia, for supporting this research through Research Management Grant Vot. No. R.
J130000.7846.4F946 and PETRONAS for providing the subsurface
dataset.
References
Agi, A.A., Junin, R., Gbonhinbor, J.R., Onyekonwu, M.O., 2017. Exploitation of thin oil
rims with large associated gas cap. Int. J. Petroleum Engineering 3 (1), 14–48.
Archie, G.E., 1942. The electrical resistivity log as an aid in determining some reservoir
characteristics, transactions of american institute of mining. Metallurgical and
Petroleum Engineers 146, 54–62.
Charles, T., Guemene, J.M., Corre, B., Vincent, G., Dubrule, O., 2001. Experience with the
quantification of subsurface uncertainties. In: Paper SPE 68703, Presented at the SPE
Asia Specific Oil and Gas Conference and Exhibition Held in Jakarta, Indonesia,
17–19 April.
Cuddy, S., Allison, G., Steele, R., 1993. A simple convincing model for calculating water
saturations in southern north sea gas fields. In: Paper H, SPWLA 34th Annual Logging
Symposium, 13–16 June.
Dai, J., Snyder, F., Gillespie, D., Koesoemandinata, A., Dutta, N., 2008a. Exploration for
gas hydrate in the deep water, northern gulf of Mexico: Part I. A seismic approach
based on geologic model, inversion, and rock physic principles. Mar. Petrol. Geol. 25,
830–844.
Dai, J., Banik, J., Gillespie, D., Dutta, N., 2008b. Exploration for gas hydrate in deep
water, northern gulf of Mexico: Part II. Model validation by drilling. Mar. Petrol.
Geol. 25, 845–859.
De Hemptinne, J.C., Zhou, H., Morachinni, G., Brunnella, I., 1997. Influence of water on
estimation of gas in place. SPE Reservoir Eng. 207–2010.
Demirmen, F., 2007. In: Reserve Estimation: the Challenge for the Industry, Paper SPE103434, Distinguished Authors Series, pp. 80–89.
Ellis, D., Singer, J., 2007. Well Logging for Earth Scientists, second ed. Springer,
Dordrecht, The Netherlands.
Etris, E., Crabtree, N.J., Dewar, J., 2002. True depth conversion: more than a pretty
picture. CSEG Recorder 26 (9), 11–22.
Ezeh, S.C., Mode, W.A., Ozumba, B.M., Yelwa, N.A., 2016. Sedimentary and ichnology of
neogene coastal swamp deposits in the Niger delta basin, Nigeria. Geologos 22 (3),
191–200.
Fui, H.K., 1978. Stratigraphic framework for oil exploration in Sarawak. Geol. Soc.
Malaysia. Bulletin 10, 1–13.
Gbadamosi, O.A., Kiwalabye, J., Junin, R., Agi, A.A., 2018. A review of gas enhanced oil
recovery schemes used in the north sea. J. Petrol. Explor. Prod. Technol 1–15.
https://doi.org/10.1007/s13202-018-0451-6.
Grechka, V., Tsvankin, I., Cohen, J.K., 1997. Generalized Dix equation and analytic
treatment of normal-movement velocity for anisotropic media. In: Paper Presented at
the 59th EAGE Conference, Geneva.
Haldorsen, H.H., Damsleth, E., 1990. Stochastic modelling. JPT (J. Pharm. Technol.) 42,
404–412.
Harrison, B., Jing, X.D., 2001. Saturation heights methods and their impact on volumetric
hydrocarbon in place estimates. In: SPE Paper 71326, 30th September- 3rd October.
264
Journal of Natural Gas Science and Engineering 57 (2018) 238–265
M.A. Yusof et al.
Challenge the norms: fluvial channel in shallow marine environment. It is Possible? A
case study of A-1 well in central luconia, Off-shore Sarawak, Malaysia-An integrated
approach combining seismo-stratigraphy and seismic attributes analysis. In: AAPG
Annual Convention and Exhibition, Pittsburgh, Pennsylvania, 19–22 May.
Wang, X., Wu, S., Lee, M., Guo, Y., Yang, S., 2011. Gas hydrate saturation from acoustic
impedance and resistivity log in the shenhu area, south China sea. Mar. Petrol. Geol.
28, 1625–1633.
Wu, W., 2001. Advantages and significance of prestack migration, paper SEG-2001-1037.
In: Presented at the SEG Annual Meeting Held in San Antonio, Texas, USA, 9-14
September.
Yi, B.Y., Lee, G.H., Kang, N.K., Yoo, D.G., Lee, J.Y., 2017. Deterministic estimation of gashydrate resource volume in a small area of the ulleung basin, east sea (Japan sea)
from rock physics modelling and pre-stack inversion. Mar. Petrol. Geol. https://doi.
org/10.1016/j.marpetgeo.2017.11.023.
Yilmaz, O., 2001. Seismic Data Analysis, Processing, Inversion and Interpretation of
Seismic Data. Society of Exploration Geophysicists, Tulsa, USA.
Rijks, E.J., 1981. Baram delta geology and hydrocarbon occurrence. Geol. Soc. Malaysia
14, 1–18.
Scherer, F.C., 1980. Exploration in east Malaysia over the last decade. In: Halbouty, M.t.
(Ed.), Giant Oil and Gas Fields of the Decades 1968-78. Am. Assoc. Petrol. Geol. Mem,
30, 423–440.
Sheriff, R., Geldart, L., 1995. Exploration Seismology, second ed. Cambridge University
Press, New York.
Siddiqui, N.A., El-Ghali, M.A., Abd Rahman, A.H., Mijinyawa, A., Ben-Juwah, J., 2013.
Depositional environment of shallow marine sandstone from outcrop gamma ray
logs, belait formation, meragang beach, Brunei, darussalam. Res. J. Environ. Earth
Sci. 5 (6), 305–324.
Skelt, C., Harrison, R., 1995. An integrated approach to saturation heights analysis. In:
Paper NNN, SPWLA 36th Annual Logging Symposium.
Skelt, C., 1996. A relationship between height and saturation, permeability, and porosity.
In: Paper E018, 17th European Formation Evaluation Symposium (SPWLA),
Amsterdam, 3-7 June.
Wafa, S.S., Mustafa, M.S., Kamarudzaman, K.A., Sapri, D.H., Mahmud, O.A., 2013.
265
Документ
Категория
Без категории
Просмотров
1
Размер файла
14 928 Кб
Теги
2018, 038, jngse
1/--страниц
Пожаловаться на содержимое документа