You are on page 1of 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317156213

Well Log Based Geomechanical and Petrophysical Analysis of the Bakken


Formation

Conference Paper · June 2017

CITATIONS READS

12 1,070

5 authors, including:

Alan Alexeyev Mehdi Ostadhassan


University of North Dakota Northeast Petroleum University
5 PUBLICATIONS   12 CITATIONS    102 PUBLICATIONS   407 CITATIONS   

SEE PROFILE SEE PROFILE

Rehan Ali Mohammed Bailey Bubach

4 PUBLICATIONS   12 CITATIONS   
University of North Dakota
32 PUBLICATIONS   157 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

NMR study View project

Petrophysical rock typing View project

All content following this page was uploaded by Alan Alexeyev on 25 May 2017.

The user has requested enhancement of the downloaded file.


ARMA 17-942

Well Log Based Geomechanical and Petrophysical


Analysis of the Bakken Formation
Alan Alexeyev, Mehdi Ostadhassan, Rehan Ali Mohammed, Bailey Bubach,
Seyedalireza Khatibi, Chunxiao Li, Lingyun Kong
University of North Dakota, Petroleum Engineering Department, Grand Forks, ND, USA

Copyright 2017 ARMA, American Rock Mechanics Association


This paper was prepared for presentation at the 51st US Rock Mechanics / Geomechanics Symposium held in San Francisco, California, USA, 25-
28 June 2017. This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and
critical review of the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA,
its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written
consent of ARMA is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may not be
copied. The abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: The detailed knowledge of reservoir properties in the first step in developing a reservoir. This paper discusses the
derivation of reservoir and geomechanical properties from well logs in the unconventional reservoir of Middle Bakken in Williston
Basin, ND. The study was performed on the middle member of the Bakken Formation in the Blue Buttes Field. Reservoir analysis
consisted of determining the lithology and petrophysical properties such as porosity, permeability, water saturation, and shale volume
calculations. We included XRD analysis for more precise analysis of the reservoir behavior. Geomechanical properties were
calculated using existing correlations and relation. Those properties included the Young’s modulus, Poisson’s ratio, tensile strength,
UCS, and pore pressure.

permeability, water saturation, shale volume, lithology,


1. INTRODUCTION and mineralogy. Various methods and approaches were
used to acquire those properties.
The advancements in hydraulic fracturing and horizontal
drilling enabled production from the low porosity, low
2. GEOLOGIC SETTING
permeability Middle Bakken reservoir. The focus of this
study is to add to the pool of information on the Bakken
The Williston Basin is located in the center of North
that will help with understanding of the properties and
America. In particular, it overlies the areas of US states
create a methodology for future similar studies. This case
and Canadian provinces of North Dakota, Montana,
study will help with getting an additional information on
Saskatchewan, and Manitoba. The Bakken Formation lies
the Bakken Formation and geomechanical properties in
slightly below 3100 meters (10,000 feet). The
the Blue Buttes Field in particular as a major producing
distinguished feature of this formation is that it comprises
field. Middle Bakken is the producing part of the
of three highly recognizable parts – two shale members
formation; therefore, it has been the attention of many
and one mixed lithology member in between those. They
geologists and reservoir engineers in order to characterize
are formally called Upper, Middle, and Lower Bakken.
it in great details. Knowing these properties is important
Bakken is part of the Mississippian/Devonian period,
as they are used in the beginning of the oilfield
originated about 350M years ago. It was deposited in the
development for reservoir simulation and geomodelling
offshore marine environment. In recent years, various
purposes as well as the geomechanical and Mechanical
studies have been done on various geological and
Earth Modelling (MEM).
petrophysical properties of the Middle Bakken (Simenon,
2010; Sonnenberg, 2009; Alexandre, 2011; Kowalski,
Well log based modeling is the efficient way of getting
2016). The general conclusions are that the porosity
different reservoir properties in the absence of actual core
exhibits the range between 3 and 10%, permeability
measurements, and it is considered more cost effective to
between 0.001 mD and smaller to 0.01mD and slightly
get the data compared with conventional core
higher, and the lithology is mainly a dolomitic sandstone,
measurements. The work started with well logs gathering
with parts of feldspar, calcite, and clay (Webster, 1984;
from the designated field. The reservoir properties to
Meissner, 1991, LeFever, 1991; Gerhard, 1982).
identify were: porosity, permeability, effective

1
3. RESERVOIR PROPERTIES coefficient values, thus more attention while using the
relationship may be required. We observed that when
To begin the reservoir characterization, basic rock
using Coates formula, the Kc permeability coefficient
properties need to be evaluated and understood. In this
needs to be set at 70 to 100 in value in order to get the
study, we analyzed permeability, porosity, effective
curve that would at least be in the same permeability
porosity, shale volume, and water saturation using a set of
range as some cored values. It is not possible for sure to
commercial software and utilizing the different well logs get the exact curve for permeability, as this would require
available at the North Dakota Industrial Commission,
a coring procedure for every single well; however, we can
Department of Mineral Resources, Oil and Gas Division
get it to match the permeability range of the Middle
database website.
Bakken, which according to core lab results and relevant
Gamma Ray, Compensated Neutron Density log,
studies (Alexandre, 2011; Kowalski, 2016; Simenson,
Borehole Compensated log, and Resistivitty logs were 2010) is within 0.001-0.01 mD. Table 1 presents the
collected from the wells that completely penetrated
results from the core analysis from well #16652.
Bakken and Three Forks formations. The initial logs were
The irreducible water saturation was calculated based on
in the image TIFF format that required them to be
the formula:
digitized in order to be used in the software.

3.1. Permeability
There is no direct relation between porosity and 𝑆 = (3)
permeability, yet via the empirical research some
researchers were able to find some correlations, which Where Swirr is the irreducible water saturation, BVW is
would enable us to calculate permeability using only bulk volume water, 𝜙 is porosity from Neutron Density.
porosity, and, in some cases, irreducible water saturation Figure 2 is the irreducible water saturation that is
together with combination of various coefficients (Timur, calculated. The value of Swirr varies from 0.25 to 0.35. The
1968; Coates and Dumanoir, 1973; Wyllie and Rose, upper bound of 0.35 was taken as the Swirr value to be
1950; Morris and Biggs, 1967). Mainly, two methods input in the permeability calculations. Using the
were tried to estimate the permeability: Wyllie-Rose and procedures described, the permeability ended up being
Coates. equal between 0.001 to 0.01 mD.

Wyllie-Rose is the following:

𝐾=𝐾 ∗ (1)

Where K is permeability, ϕ is porosity, e, d are empirical


coefficients, and Kw is permeability coefficient. Two
different sets of values for these coefficients were
developed by Timur (1968) and Morris-Biggs (1967).
Timur coefficients e=2, d=4.4, and Kw=3400 are
empirically based on 155 sandstone samples. Morris-
Biggs coefficients are equal to e=2, d = 6 and Kw = 62500.
Porosity is taken from Density Porosity log.
Coates formula includes the effective and total porosities,
as well as irreducible water saturation:
Fig. 1. Permeability curves comparing Wyllie-Rose and
Coates methods.

𝐾 =𝐾 ∗𝜙 ( ∗
) (2)

Where K is permeability, Kc is permeability coefficient,


ϕ_eff is effective porosity, ϕ_total is total porosity from
neutron-density, and S_wirr is irreducible water saturation.
As shown in figure 1, the permeability curves from both
methods are quite similar, however, the Wyllie-Rose
method tends to overestimate the permeability when
calculating using the default Timur or Morris-Biggs

2
Fig. 3. Profile of the porosity and effective porosity
from Neutron Density calculation.

Fig. 2. Irreducible water saturation profile.


3.3. Water Saturation
Table 1. Porosity and permeability form well #16652 Water saturation was calculated using two methods:
standard Archie, 1942 and Simandoux, 1963. Archie
Porosity, Permeability found the relations between resistivity and water
Depth, ft % to air, mD saturation by knowing m and n coefficients. Below is the
10665.6 5.66 0.004 Archie water saturation formula:
10666.25 5.49 0.004

10674.1 6.68 0.034 𝑆 = (4)

10675.2 6.18 0.014
10676 7.4 0.005
Where Sw is the water saturation, m is cementation
10676.9 6.58 0.004 exponent, n – saturation exponent, a – tortuosity factor,
10678.2 6.09 0.008 Rw – water resistivity, Rt – formation resistivity, 𝜙 is
10679.1 6.18 0.004 porosity from density log.
10680.1 6.2 0.004 The m and n exponents usually equal to each other and
10681 6.27 0.004 are set at 2. However, for some formations it is advised to
10682 6.32 0.004 modify the values in order to get more accurate results.
For this formation, we found that m and n are equal to
10683.2 6.82 0.006
slightly less than 2 (Simenson, 2010; LeFever, 1991), and
10684.1 5.68 0.003 through constructing a Pickett plot m and n are equal to
10685.2 4.36 0.001 roughly 1.75.
10686.1 3.44 0.001 Simandoux formula incorporates the shale resistivity and
10687.1 5.96 0.003 effective porosity, and is usually used in shaly
formations as below:
10688.1 3.33 0.021
10689.1 5.89 0.002 ∗𝑆 + ∗𝑆 − =0 (5)
10690 5.96 0.002
10691.1 6.81 0.032
Where 𝑅 is deep formation resistivity, 𝑅 is resistivity
of formation water, 𝑅 is resistivity of shale, 𝜙 is
3.2. Porosity effective porosity, 𝑉 is shale volume, 𝑎 is tortuosity
The results for the calculated effective porosity are shown factor, m is cementation exponent and n is saturation
in figure 3. Table 1 lists the core porosity measurements. exponent.
As can be seen, the porosity varies between 3 and 7%. The
well log based calculated effective porosity, which is
removing the shale volume, was very close to those
values, varying from 4 to 8%.

3
This method was used here for comparison purposes. As 3.5. Lithology
seen in figure 4, the two methods produce quite similar To get a better idea of lithology and mineralogy X-Ray
saturation profiles, resulting in water saturation being Diffraction (XRD) test is done. For XRD, the core
equal to 40% or less. Similar findings were also reported samples from three Blue Buttes wells (#2820, #16433,
in the study of the Bakken by Simenson, 2010 and #16652) were taken to the lab and prepared for the test.
Alexandre, 2011. The results from the XRD are summarized in the table 2.
Table 2. XRD Results
Component, #2820 - #2820 #16433 #16433 #16652 #16653
in % 10592.5' - - - - -
10598' 10612' 10638' 10663' 10678'
dolomite 33 13 40.5 13.3 11.1 15.3
calcite 8.6 14.9 9 10.2 6.2 10.7
(limestone)
quartz 31.3 60.8 50.5 41.6 45.45 50.9
(sandstone)
Muscovite 13.3 11.8 11.3
Sanidine, 12.8 22.6 25.9
ferrain
pyrite 1 10 0.5
K 1.2
(potassium)
Fig. 4. Water saturation profiles using Archie and Simandoux Illite 23.1
methods.

3.4. Shale Volume The mineral composition of the cores is shown in the
The common way to calculate the shale volume is percentages. The largest element in each core is the quartz
through the Gamma Ray Index, using the following (sandstone), comprising to about a half. The mineral
formula: content is quite heterogeneous, with some elements
present in one sample and missing in the other. Dolomite
𝐺𝑅 = (6) is the second largest element, varying from 11 to 40%.
Calcite is very minimal at roughly 10% in each sample.
Feldspar based minerals appear in most cores, comprising
Where GR is the total Gamma Ray values, without
about 30%.
considering spectral gamma ray. The non-linear shale
volume computation method developed by Larionov Similar results were found from the mineralogy modeling
(1969) was used in this study. The formula for older rocks using Quanti.Elan module. Based on XRD analysis from
is the following: few samples, muscovite-mica and K-Feldspar were
chosen in the model. Other minerals included calcite,
𝑉 = ∗ (2( )
− 1) (7) dolomite, quartz, and illite. Figure 6 represents the well
log based estimated minerals. We can observe that quartz
Below is figure 5, which shows the profiles of shale
(sandstone) always was the largest component, with
volume calculated using the abovementioned method.
dolomite being the second most abundant, and then
Shale volume distribution does not change significantly
feldspar and calcite. The overall lithology can be
and stays at 10-20%.
interpreted as dolomitic sandstone, since the sandstone
and dolomite compose the largest portions in the cores.
Similar findings were found in the study of LeFever,
1991.

Fig. 5. The profiles of shale volume across the wells.

4
Poisson’s Ratio of the Bakken formation that could be
used for comparison.
These geomechanical properties can later be used to
design hydraulic fracturing operation, which is essential
for production in the Bakken. Thus, calculating these
properties is a crucial step in characterizing a reservoir.
From the set of well logs, the following curves were taken
for further analysis: GR, NPHI, DT, RHOB. After that,
we calculated the following properties: DT Shear (travel
time), UCS, E (Young’s modulus), G (shear modulus), K
(bulk modulus) in static and dynamic forms, friction
angle, Poisson’s Ratio, pore pressure, and tensile strength.
The following formulas represent the correlations that
were used in order to extract properties from the logs:

Young’s modulus, dynamic:

𝐸 = ∗ 1.34 ∗ 10 (𝑝𝑠𝑖) ∗
10 ∗ 6.895 (𝐺𝑃𝑎) (9)
Fig. 6. Mineralolgy and lithology of the sample well.
Poisson’s Ratio:
4. GEOMECHANICAL PROPERTIES
𝛥𝑡2𝑠 −2𝛥𝑡2𝑝
Calculation of geomechanical properties is the next step 𝜈 = 0.5 (10)
𝛥𝑡2𝑠 −𝛥𝑡2𝑝
after reservoir properties have been established and well
logs had been collected. Not always are the cores Shear modulus, dynamic:
available for testing for Uniaxial Compressive Strength 𝐸
𝑑𝑦𝑛
(UCS) or other properties, therefore the empirical 𝐺𝑑𝑦𝑛 = 2(1+𝜈 )
(11)
correlations exist that can calculate the missing properties
based on existing data and logs. Those correlations were
achieved based on numerous testing from different types Bulk modulus, dynamic:
of formations with different lithology and different parts 𝐸
𝑑𝑦𝑛
of the world. One such relation is to get a shear wave 𝐾𝑑𝑦𝑛 = 3(1−2𝜈 )
(12)
velocity (Vs) from compressional-wave velocity (Vp). For
our study, we used the relationship developed by Brocher Young’s Modulus, static:
(2005), which is the following:
𝑉𝑠 = 0.7858 − 1.2344𝑉𝑝 + 0.7949𝑉2𝑝 – 0.1238𝑉3𝑝 + 𝐸𝑠𝑡𝑎𝑡𝑖𝑐 = 0.414 ∗ 𝐸𝑑𝑦𝑛 − 1.05 (13)
0.0064𝑉4𝑝 (8)
There is another correlation by Castagna (1985), but it Shear modulus, static:
is found to be valid for Vp between 1.5 and 4.25 km/s, 𝐸
𝑠𝑡𝑎𝑡
𝐺𝑠𝑡𝑎𝑡 = 2(1+𝜈 (14)
which is not the case in our study, as most of Vp values )
are above 5 km/s.
Other researchers tried to evaluate mechanical properties Bulk modulus, static:
too. Liu, et al (2016) used nanoindentation to evaluation 𝐸
𝑠𝑡𝑎𝑡
𝐾𝑠𝑡𝑎𝑡 = 3(1−2𝜈 (15)
mechanical properties in the Bakken Formation and )
correlated the results with XRD and SEM. Azadpour, et
al. (2015) used Weakley’s approach to calculate pore For UCS, the correlation by Bradford (1998) was used,
pressure in a carbonate formation. Chang, et al. (2006) the following is the formula:
reviewed the derivation of mechanical properties in
various lithologies. Xu, et al. (2016) used mineralogy 𝑈𝐶𝑆 = 2.28 + 4.1089 ∗ 𝐸 (16)
approach for brittleness and rock strength assessment.
The paper contains the profiles of Young’s Modulus and

5
Tensile strength was taken as 10% of UCS. 7. REFERENCES
Friction angle:
1. Alexandre, C. S., S. A. Sonnenberg, and J. F Sarg, 2011.
𝜑 = 26.5 − 37.4(1 − 𝜙 − 𝑉 ) + 62.1(1 − 𝜙 −
Reservoir Characterization and Petrology of the Bakken
𝑉 ) (17) Formation, Elm Coulee Field, Richland County, MT
(Doctoral dissertation, Colorado School of Mines).
Eaton’s method was used to calculate the pore pressure: 2. Archie, G. E. 1942. The electrical resistivity log as an
∆ aid in determining some reservoir
𝑃 = 𝑂𝐵𝐺 − (𝑂𝐵𝐺 − 𝑃 )( ) (18) characteristics. Transactions of the AIME, 146(01), 54-

62.
Where OBG is overburden stress gradient, 𝑃 is
3. Azadpour, M., N.S Manaman. 2015. Determination of
hydrostatic pore pressure gradient, n and log refer to Pore Pressure from Sonic Log: a Case Study on One of.
normal and measured values of sonic ΔT at each depth. Iran Carbonate Reservoir Rocks. In Iranian Journal of
We first had to establish a normal compaction trend. Once Oil& Gas Science and Technology.
the relation has been established, multiplying it by the 4. Bradford, I. D. R., J. Fuller, P. J. Thompson, and T. R.
depth allowed to calculate the pore pressure. Figure 7 Walsgrove, 1998, January. Benefits of assessing the
represents the ΔT curve versus depth – we can observe the solids production risk in a North Sea reservoir using
normal trend in the beginning before it starts to deviate at elastoplastic modelling. In SPE/ISRM Rock Mechanics
deeper depths. in Petroleum Engineering. Society of Petroleum
Engineers.
In a paper by Xu and Sonnenberg (2016), they performed
5. Brocher, T. M. 2005. Empirical relations between elastic
UCS testing on a well nearby from the Blue Buttes Field. wavespeeds and density in the Earth's crust. Bulletin of
They listed the UCS, Young Modulus, Poisson’s Ratio for the Seismological Society of America, 95(6), 2081-2092.
the middle Bakken core. Given that those are the cored
values, we compared our well log based calculations with 6. Brocher, T. M. 2008. Key elements of regional seismic
their measurements, and the values were found to be velocity models for long period ground motion
simulations. Journal of Seismology, 12(2), 217-221.
similar, giving us additional reassurance that the
correlations we used are within acceptable accuracy. 7. Castagna, J. P., M. L. Batzle, and R. L. Eastwood. 1985.
Relationships between compressional-wave and shear-
wave velocities in clastic silicate rocks. Geophysics,
5. CONCLUSIONS 50(4), 571-581.
8. Chang, C., M. D. Zoback, and A. Khaksar. 2006.
The reservoir and geomechanical properties were
Empirical relations between rock strength and physical
evaluated using well logging and empirical correlation, properties in sedimentary rocks. Journal of Petroleum
with very few core results used only for the reference to Science and Engineering, 51(3), 223-237.
make sure our results are within the acceptable range.
Having only the initial well logs and geomechanical 9. Coates, G. R. and J. L Dumanoir. 1973, January. A new
approach to improved log-derived permeability.
correlations it’s possible to prepare large amounts of data
In SPWLA 14th Annual Logging Symposium
and evaluate the results of well logging analysis.
10. Gerhard, L. C., S. B. Anderson, J. A. Lefever and C. G.
We showed lithology, reservoir and geomechanical Carlson. 1982. Geological development, origin, and
properties of the Bakken Formation, and this dataset is energy mineral resources of Williston Basin, North
ready to be used for further geomodelling, simulation, and Dakota. AAPG Bulletin, 66(8), 989-1020.
geomechanical studies.
11. Kowalski, B. 2016. Quantitative mineralogic analysis of
the middle Bakken member, Parshall Field, Mountrail
County, North Dakota (Doctoral dissertation, Colorado
6. ACKNOWLEDGEMENTS School of Mines).
The authors are thankful to Aldjia Boualam and Sofiane 12. LeFever, J. A. 1991. History of Oil Production from the
Djezzar for helping with reservoir properties and software Bakken Formation, North Dakota.
guidance, Kouqi Liu and Xiaodong Hou for XRD setup
13. Liu, K., M. Ostadhassan, B. Bubach, and H. Jabbari.
and data gathering, Schlumberger and Neurolog for 2016, June. Bakken Formation Shales Nano-Scale
software use. Analysis to Understand Mechanical Parameters. In 50th
US Rock Mechanics/Geomechanics Symposium.
American Rock Mechanics Association.
14. Meissner, F. F. 1991. Petroleum Geology of the Bakken
Formation Williston Basin, North Dakota and Montana.

6
15. Morris, R. L. and W. P. Biggs. 1967, January. Using log-
derived values of water saturation and porosity.
In SPWLA 8th Annual Logging Symposium. Society of
Petrophysicists and Well-Log Analysts.
16. Simandoux, P. 1963. Dielectric measurements on porous
media, application to the measurements of water
saturation: study of behavior of argillaceous formations.
Revue de l’Institut Francais du Petrol,
18(supplementary issue), 93-215.
17. Simenson, A. 2010. Depositional Facies and
Petrophysical Analysis of the Bakken Formation,
Parshall Field, Mountrail County, North Dakota
(Colorado School of Mines).
18. Sonnenberg, S. A. and A. Pramudito. 2009. Petroleum
geology of the giant Elm Coulee field, Williston Basin.
AAPG bulletin, 93(9), 1127-1153.
19. Timur, A. 1968, January. An investigation of
permeability, porosity, and residual water saturation
relationships. In SPWLA 9th annual logging symposium.
Society of Petrophysicists and Well-Log Analysts.
20. Webster, R. L. 1984. Petroleum source rocks and
stratigraphy of the Bakken Formation in North Dakota.
21. Wyllie, M. R. J. and W. D. Rose. 1950. Some theoretical
considerations related to the quantitative evaluation of
the physical characteristics of reservoir rock from
electrical log data. Journal of Petroleum Technology,
2(04), 105-118.
22. Xu, J., and S. Sonnenberg. 2016, August. Brittleness and
Rock Strength of the Bakken Formation, Williston
Basin, North Dakota. In Unconventional Resources
Technology Conference, San Antonio, Texas, 1-3 August
2016 (pp. 2237-2254). Society of Exploration
Geophysicists, American Association of Petroleum
Geologists, Society of Petroleum Engineers

8. APPENDIX
Below are the results of the geomechanical calculations.
Figure 7 represents the normal compaction trend for
compressional slowness for five different wells in the
Blue Buttes Field. Figures 8 and 9 represent the
calculated geomechanical properties using the formulas
presented earlier.

7
Fig. 7. Normal compaction trend (ΔTn) for ΔT (compressional slowness) profiles for different wells.

8
View publication stats

Fig. 8. Geomechanical properties for the well #7572.

Fig. 9. Geomechanical properties for the well #9558.

You might also like