A Method for Prediction of the Planar Distribution of Liquid

Hydrocarbons in Shale Gas: A Case Study in Duvernay Shale in the

West Canadian Sedimentary Basin

Houqin Zhu

, Yuzhong Xing

and Xiangwen Kong

Research Institute of Petroleum Exploration and Development, PetroChina, Beijing.

Email: zhuhouqin@petrochina.com.cn

Keywords: Liquid hydrocarbon, pyrolysis experiment, hydrogen index, planar distribution, Duvernay Shale, West

Canadian Sedimentary Basin (WCSB)

Abstract: The pyrolysis experiment data and production data are applied to predict the planar distribution of liquid

hydrocarbons in shale gas. On the basis of the variation of pyrolysis experiment parameters of samples

before and after extraction, the quantitative relations of △S2 with Ro, △S2/S2 with Ro, △HI with Ro and

△HI with HI

unextracted

are established and HI

unextracted

is corrected. Combined with the production data, the

quantitative relation of HI with CGR is established. The planar distribution diagram of HI values is plotted,

and based on the division standard for different grade of liquid hydrocarbons, the planar distribution of

different grade of liquid hydrocarbons is predicted quantitatively. The results indicate that, (1) the pyrolysis

comparison experiments before and after extraction indicate that △S2, △S2/S2 and △HI have a good

correlation with Ro, △HI also has good correlation with HI

unextracted

, and these quantitative relations can be

used to correct HI

unextracted

. (2) The corrected HI has a very good correlation with CGR, and the planar

distribution of all grades of liquid hydrocarbons can be predicted according to HI. (3) The dry gas is

distributed mainly in the structural deformation belt and Leduc ring reef belt. The northeast region of the

study area is mainly a rich oil region. Gas condensate, rich gas condensate and very rich gas condensate are

distributed in NW-SE banded shape from Simonette to Willesden Green.

1 INTRODUCTION

According to the practices in the North America,

quantitative evaluation on planar distribution of

liquid hydrocarbons and enhancement of well fluid

deliverability are effective ways to improve the

economic profit of shale gas exploration and

development, in addition to technical progress. As is

known, prior to shale gas exploration and

development, shale was always considered as the

source rock for conventional reservoir. Thus there

are plenty of geochemical and logging data.

Typically, the pyrolysis data can be acquired easily,

fast and accurately with low cost. With the pyrolysis

data, together with production performance data, the

planar distribution of liquid hydrocarbons can be

predicted quantitatively. Then, in combination with

the shale facies research and the existing

drilling/completion technologies, cost reduction and

benefit improvement can be achieved in shale gas

exploration and development.

Rokosh et al. (2012) divided the Duvernay shale

in the WCSB into four regions, i.e. Gas Maturity,

Liquid Maturity, Oil Maturity and Immature,

according to the kerogen thermal genetic theory and

the Ro data (Rokosh et al., 2012). However, the

thermal maturity can simply be used to qualitatively

judge whether there is a potential of liquid

hydrocarbon in the region, while the shale oil and

gas operators and investors care more about whether

the shale is dominated by liquid hydrocarbon or gas

and how much liquid hydrocarbons are endowed.

The liquid hydrocarbon refers to condensate oil

herein.

The main purpose of this paper is, with the

Duvernay shale in WCSB as an example, to

establish the relations of HI vs. CGR, based on the

pyrolysis and production performance data, and use

these relations to predict the planar distribution of

Zhu, H., Xing, Y. and Kong, X.

A Method for Prediction of the Planar Distribution of Liquid Hydrocarbons in Shale Gas - A Case Study in Duvernay Shale in the West Canadian Sedimentary Basin.

In Proceedings of the International Workshop on Environment and Geoscience (IWEG 2018), pages 261-266

ISBN: 978-989-758-342-1

261

liquid hydrocarbons. On the basis of this, a set of

quantitative evaluation method is proposed

quantitative evaluation on liquid hydrocarbons in

shale.

2 GEOLOGIC SETTING

The West Canadian Sedimentary Basin (hereinafter

abbreviated to WCSB) is a typical wedge-shaped

foreland basin and covers an area of 170×10

WCSB is located between the Canadian Shield and

Cordillera fold mountain system. The Rocky

Mountain piedmont thickness is larger than 6000m

in Albert and British Columbia (Jarvie et al., 2010;

Shannon et al., 1989). WCSB can also be divided

into two parts such as Albert Basin and Williston

Basin. The study area is located in the central west

of Albert and the slope belt of WCSB, where the

strata are gentle and their burial depth is

1400~5500m. The study area is located in the

central west of Albert and the slope belt of WCSB,

with gentle formations buried at 1400~5500m

(Figure 1).

Figure 1: Regional location map of the study area

(modified as per reference Jarvie et al., 2010 (

Jarvie et al.,

2010

Duvernay shale is a set of dark brown or black

organic-rich shale deposited in the maximum

transgression period of Upper Devonian Woodbend

Group. The shale and Leduc reefs deposited

simultaneously. Large scale transgression occurred

in the sedimentary period of Woodbend Group,

apron reefs were mainly developed near the Peace

River structural belt, the lithology inside the basin

was deepwater limestone and shale, and Leduc reefs

were characterized by patch reefs, and appeared

mainly in the shallow water area on the carbonate

platform and the west margin of the basin. At the

end of the sedimentary period of Woodbend Group,

the approximate 325km long NE-trending Rimbey-

Meadowbrook reef belt across Edmonton Region

divided Albert Basin into east and west shale basins

(Mossop and Shetsen, 1994).

The WCSB has evolved in three stages, i.e.

stable craton platform from pre-Cambrian to Middle

Jurassic, foreland from Middle Jurassic to Eocene,

and intra-craton basin from Eocene to the present.

3 DATA AVAILABLE AND

METHODS

3.1 Data Available

For this study, pyrolysis data of 40 wells (including

Beaton data in 2010) (Beaton et al., 2010) were

acquired, including 3 wells with data both before

and after extraction (whole core pyrolysis

experiment for 2 wells, and cutting pyrolysis

experiment for 1 well), and 37 wells with data

before extraction, and production data of 67 wells

were acquired. Among these wells, 32 wells have

both pyrolysis data and CGR data.

3.2 Methods

According to the geochemical theory, with the

pyrolysis, together with the production performance

data, the planar distribution of hydrocarbons at

different grades in the Duvernay shale is

quantitatively predicted. Based on the variation of

pyrolysis experiment parameters of samples before

and after extraction, the quantitative relations of

△S2 with Ro, △S2/S2 with Ro, △HI with Ro and

△HI with HI

unextracted

are established, and then

unextracted

is corrected. Combined with the

production data, the quantitative relation between HI

and CGR is defined, and the HI division standard for

liquid hydrocarbons at different grades is clarified.

According to the division standard and the planar

distribution map of HI, the planar distribution of

liquid hydrocarbons at different grades is predicted

quantitatively (△S2=S

2unextracted

-S

2extracted

△HI=HI

unextracted

-HI

extracted

IWEG 2018 - International Workshop on Environment and Geoscience

262

4 RESULTS AND DISCUSSIONS

4.1 HI Correction

Espitalie et al. (1977) firstly proposed a rock

pyrolysis method for obtaining S1 and S2 which

respectively denote the amount of free or adsorbed

hydrocarbons in source rocks and the amount of

hydrocarbons generated from kerogen pyrolysis so

as to reflect the hydrocarbon generation capacity of

mature source rocks (Espitalie et al., 1980). S1 is

corresponding with the hydrocarbons volatilized

during heating to ≤300℃ in a Rock-Eval experiment,

and they are basically C7-33 hydrocarbons; S2 is

corresponding with the hydrocarbon yields from

pyrolysis during heating to >300℃ in the Rock-Eval

experiment. Delvaux et al. (1990) made some

modifications on the definition and normalized the

experiment results of the amount S1 of free or

adsorbed hydrocarbons and the amount S2 of

hydrocarbons generated from kerogen pyrolysis

(Delvaux et al., 1990). Dan Jarvie (1987), Lafargue E.

et al. (1998), Behar F. et al. (2001)

(Dan, 1984;

Lafargu et al., 1998; Behar et al., 2001) believed that

S2 was hydrocarbons released during the pyrolysis

of kerogen between 300 and 550 or 600 degrees C

with a linear temperature gradient usually between

25 and 30 degrees C per minute. Wang Anqiao et al.

(1987) found that the value of S2 after chloroform

extraction was less than that before chloroform

extraction through the comparison of a direct

pyrolysis experiment on a source rock sample with a

pyrolysis experiment on it after chloroform

extraction (Wang and Zheng, 1987). This indicates

that there are some liquid hydrocarbons in S2; due to

the adsorption and swelling action of organic

matters and too high boiling point of part liquid

hydrocarbons (boiling point of n-C18 302℃), these

liquid hydrocarbons cannot be evaporated out

at<300℃ in a Rock-Eval experiment. Delvaux et al.

(1990) also obtained the same conclusion from their

studies. According to the conclusion, part

macromolecular substances belonging to free

hydrocarbons S1 such as asphaltene and colloid in

crude oil have similar pyrolysis hydrocarbon

temperature during heating in a conventional sample

pyrolysis experiment, so that the experiment value

of free hydrocarbons S1 is relatively low while that

of pyrolysis hydrocarbons S2 is relatively high. The

determination of S2 is complicated by the retention

of some of the generated hydrocarbons by the rock

matrix, and thus HI (hydrogen index=S2/TOC×100)

will not give the true ratio of pyrolyzable

hydrocarbons to organic carbon unless appropriate

correction is made (Langford et al., 1990). The best

treatment method is to resample and conduct a post-

extraction pyrolysis experiment, but this will waste a

lot of original pyrolysis data. In this paper, the

regularity of all analytical data has been discussed

on the basis of the analysis of core pyrolysis

experiment data on two wells, and relevant formulas

have been fitted to correct HI.

The comparative experiments on the full cores of

two wells before and after extraction show a large

difference, the S2 difference (△S2) is 0.75~3.52mg

HC/g Rock and the HI difference (△HI) is

20.54~82.97mg HC/g TOC. △S2, △S2/S2 and △HI

show power decrease with Ro with correlation

coefficient of 0.9, 0.9 and 0.93 respectively (Figure

2a~2c), i.e. the larger Ro is, the smaller △S2,

△S2/S2 and △HI are. △HI increases exponentially

with HI

unextracted

(Figure 2d) with a correlation

coefficient of 0.98, the larger HI

unextracted

is, the

larger △HI is. According to the above-mentioned

correlations, it is feasible that HI can be corrected by

fitting relation.

Based on the formula HI=S2/TOC×100(mg

HC/g TOC) and the above analysis, multiple

methods can be used to correct HI.

4.1.1 To Calculate HI by Correcting S2

According to the above analysis, S2=S

unextraced

-△S2,

where △S2 can be calculated from the regression

formula of △S2 with Ro, i.e. Ro = 1.4844×△S2

0.248

In addition, S2 can also be calculated from

△S2/S2. Let △S2/S2=K

S2recovery coefficient

, and then

S2=S2

unextracted

- KS

2recovery

coefficient

×S2

unextracted

=S2

unextracted

×(1-K

S2recovery coefficient

where K

S2recovery coefficient

can be calculated according

to the regression formula of Ro with △S2/S2, i.e.

Ro= 0.9632×K

S2recovery coefficient

0.413

. Substitute the

corrected S2 into the HI calculation formula to

calculate HI.

4.1.2 Direct HI Correction

Calculate △HI according to the regression formula

of △HI with HI, i.e. △HI=4.05× HI

unextracted

0.304

correction

=HI

unextracted

-△HI.

A Method for Prediction of the Planar Distribution of Liquid Hydrocarbons in Shale Gas - A Case Study in Duvernay Shale in the West

Canadian Sedimentary Basin

263

4.2 Relation of HI with CGR

According to the viewpoints of Tissort et al. (1977),

the higher the thermal evolution degree, the smaller

the residual kerogen amount and the lower the

content of hydrocarbons S2 generated from

pyrolysis (Tissot and Welte, 1984). Therefore, the

ratio of S2/TOC can reflect the fact that the higher

the maturity, the smaller the value, the worse the

residual hydrocarbon generation capacity and the

lower the hydrocarbon generation amount. The oil

and gas produced from shale are the residual

hydrocarbons generated from thermal evolution of

organic matters in shale and not migrated, and the

amount of the residual hydrocarbons is affected by

conversion and hydrocarbon generation of organic

matters. The smaller the hydrocarbon generation

capacity of the residual organic matter is, the more

the shale oil and gas reserves may be. The HI of

kerogen which has experienced thermal evolution to

some extent is just used to calibrate the level of

liquid-rich hydrocarbons in shale based on the

characteristic of shale oil and gas such as self-

generating and self-preserving so as to finally

achieve the purpose of quantitative research.

Therefore, the relation of HI with CGR can be

established using the method of combining the HI of

kerogen with the shale oil and gas production of a

single well. The relation of HI with CGR of

Duvernay shale is established according to the

collected CGR and the corrected HI (Figure 3). As

shown in the figure 3, CGR has a linear relation with

HI, their correlation coefficient is up to 0.91,

showing a very good correlation.

4.3 Prediction of Planar Distribution of

Liquid-Rich Hydrocarbons

The shale oil/gas in the study area is mainly

condensate oil, thus the grading of liquid-rich

condensate can forecast the highly economic liquid-

rich region. Due to difficult recovery, complicated

operation modes and high E&D cost, the

conventional liquid hydrocarbon grading is no

longer suitable for the expensive shale oil/gas

exploration evaluation. The liquid hydrocarbon

grading in this paper is defined with consideration to

the economic profit in field practices. Specifically,

the fluids produced from a well are divided into five

grades according to CGR, i.e. dry gas, gas

condensate, rich gas condensate, very rich gas

condensate and rich oil. The quantitative relation

between HI and CGR is used to confirm the

threshold value of kerogen HI for the five grades of

liquid hydrocarbons (Table 1).

The planar distribution of liquid hydrocarbons at

all grades can be plotted on the basis of the planar

map of HI, and according to the liquid hydrocarbon

CGR division standard, or plotted directly with the

planar map of HI and according to the liquid

hydrocarbon HI division standard. The latter method

is adopted in this study, because the correlation

between the parameters and HI is good but not

100%, which will cause large error after multiple

calculations.

Figure 2: Relation chart of pyrolysis experiment

parameters.( (a) Relation of △S

with Ro; (b) Relation of

△S

with Ro; (c) Relation of △HI with Ro; (d) Relation

of △HI with HI)

Figure 3: Relation chart of CGR with HI.

Table 1: Division standard for grades of liquid hydrocarbons in Duvernay shale.

Parameter Grade of liquid-rich hydrocarbons

Dry gas Gas condensate Rich gas condensate Very rich gas condensate Rich oil

CGR(bbl/MMcf)

<5 5~65 65~140 140~250 >250

HI(mg/g) <10 10~18 18~27 27~40 >40

* From Duvernay reserves and Resources Report (2016)

IWEG 2018 - International Workshop on Environment and Geoscience

264

The kerogen HI increases from the deformed belt

in the southwest to the northeast (Figure 4). Pinto

along the Leduc ring reef belt is in the CH4

generation stage due to relatively high thermal

evolution degree, and the corresponding HI in the

area is low, generally less than 10 mg HC/g TOC.

Controlled by the deformed belt, Edson and local

areas in South-eastern Willesden Green show high

thermal evolution degree, thus the residual kerogen

pyrolysis hydrocarbon is less in the shale. In

Simonette and mostly Willesden Green, the organic

matters are in the condensate gas-wet gas generation

stage, showing the largest rate of conversion from

kerogen to hydrocarbon.

Figure 4: Planar distribution of HI values and all levels of

liquid hydrocarbons.

All five grades of liquid hydrocarbons are

observed in the study area, and they distribute

distinctively in zones from the deformed belt to the

northeast of the study area (Figure 4). In the region

close to the deformed belt and the Leduc ring reef

belt, dry gas is dominant. In the region within the oil

generation threshold in the northeast, rich oil is

dominant. The condensate oil, rich condensate oil

and very rich condensate oil distribute as bands in

NW-SE from Simonette to Willesden Green. The

dry gas region can serve as the reserve area, which

can be developed in case of allowable economic and

technical conditions. The rich oil region where

liquid oil is dominant is not an ideal target due to

proximal high clay mineral content, difficult drilling,

high recovery cost and low recovery rate. The rich

condensate oil and extra-rich condensate oil regions

are the primary targets in the near future. The

prediction results basically match the actual

production performance of active wells.

5 CONCLUSIONS

Based on the geochemical theory, together with the

pyrolysis data and well production data, the prediction

of the planar distribution of liquid-rich hydrocarbons in

Duvernay Shale is conducted. Some conclusions are

made as follows.

According to the pyrolysis comparison

experiments before and after extraction, △S2,

△S2/S2 and △HI have a good correlation with Ro,

△HI also has good correlation with HI

unextracted

, and

these quantitative relations can be used to correct

unextracted

The corrected HI has a very good correlation

with CGR. The quantitative division standard for

liquid hydrocarbons has been established according

to production data, and the planar distribution of all

levels of liquid hydrocarbons can be predicted

according to HI.

The dry gas region is distributed mainly in the

tectonic deformation belt and Leduc ring reef belt.

The northeast region of the study area is mainly a

rich oil region. Gas condensate, rich gas condensate

and very rich gas condensate are distributed in NW-

SE banded shape from Simonette to Willesden

Green.

ACKNOWLEDGMENT

This work was supported by China National Science

and Technology Major Project (Grant No:

2011ZX05028-002) and CNPC Key Science and

Technology Project (Grant No: 2013E-050102). We

are grateful to authors for their research findings

cited in the paper. We also thank the anonymous

reviewers and the editors for their constructive

comments and suggestions, which improved the

manuscript.

A Method for Prediction of the Planar Distribution of Liquid Hydrocarbons in Shale Gas - A Case Study in Duvernay Shale in the West

Canadian Sedimentary Basin

265

REFERENCES

Beaton A P, Pawlowicz JG and Anderson S D A 2010-07

Organic Petrography of the Montney Formation in

Alberta. Shale Gas Data Release. ERCB/AGS Open File

Report

Behar F, Beaumont V And Penteado H L De B 2001 Roch-

Eval 6 technology: Performances and

development.Oil&Gas science and technology 56(2) 111-

134

Dan Jarvie 1984 Application of the Rock-EVAL III oil show

analyzer to the study of gaseous hydrocarbons in an

Oklahoma gas well. Presented at the 187th American

Chemical Society National Meeting, Geochemistry

Division. St. Louis, Mo

Delvaux D, Martin H and Leplat P 1990 Comparative rock-

eval pyrolysis as an improved tool for sedimentary

organic matter analysis. Orangic Geochemistry 16(4-6)

1221-12294

Espitalie J, Madec M and Tissot B 1980 Role of mineral

matrix in kerogen pyrolysis: influence on petroleum

generation and migration. AAPG 64 59-66

Jarvie D M, Philp R P and Jarvie B M 2010 Geochenical

assessment of Unconventional Shale resource plays, North

America Special Issue on Shale Resource Plays due out

2nd quater

Lafargue E, Marquits F and Pillot D 1998 Rock-eval 6

applications in hydrocarbon exploration, production, and

soil contamination studies. Revue de l’institut français du

pétrole 53(4) 421-437

Langford F F and Blanc M M 1990 Interpreting rock-eval

pyrolysis data using graphs of pyrolizable hydrocarbons

Vs. total organic carbon. AAPG 74(6) 799-804

Mossop G and Shetsen I 1994 Atlas of theWestern Canada

Sedimentary Basin (Calgary: Canadian Society of

Petroleum Geologists and the Alberta Research Council)

chapter 3-12 pp 22-202

Rokosh C D, Lyster S, Anderson S D A 2012-06 Summary of

Alberta's Shale- and Siltstone-Hosted Hydrocarbons,

ERCB/AGS Open File Report

Shannon P M and Naylor D 1989 Petroleum Basin

Studies(London) 41-57

Tissot B and Welte D H 1984 Petroleum formation and

occurrence, Springen, Berlin Heideberg New York 131-

160

Wang anqiao and Zheng Baoming 1987 Calibration of

analytic parameters for pyrolytic chromatography (in

Chinese with English Abstract). Experiment Petroleum

Geology 9(4) 342-350.

Yao T, Walter B A and William D M 2013 The Eagle ford

shale play, South Texas: Regional variations in fluid types,

hydrocarbon production and reservoir properties.

IPTC16808 1-12

IWEG 2018 - International Workshop on Environment and Geoscience

266