Abstract
The Besa River Formation mainly comprises argillaceous sediments, and the upper shale bed of the Formation has been considered as a prospective shale gas reservoir. The maturity and quantity of organic matter found in source rock are considered as factors for evaluating resource value. Original hydrogen index (HIo) and total organic carbon (TOCo) are considered as properties to estimate generated gas volume in shale gas reservoirs. Generally, TOCo and HIo are determined via geochemical analysis of organic matter and calculated from well-log data like density, sonic data, and resistivity. In this study, these properties were measured or calculated by using geological analysis data and bulk-density log, and then the prospective area from the geochemical perspective in Liard Basin is defined by realizing a 3-dimensional static geochemical model. To validate the static model, the gas-in-place (GIP) (496 Bcf/section) of the production well in this study area was compared with GIP (262 Bcf/section) in the static model. Expulsion efficiency was considered as 0.6. The result implies that this model has a method of informative assessment with regard to undeveloped shale gas resources. The static model provides spatial information for generated and retained gas volume in the Besa River Formation, Liard Basin.
1. Introduction
Shale resources in several unconventional fields such as Horn River, Montney, Cordova Embayment, and Liard in British Columbia have been developed or produced. Among these, the Liard Basin has been evaluated as a gas-in-place (GIP) resource of 848 trillion cubic feet (Tcf) and initial raw gas reserves of 0.1 Tcf in the Exshaw and Patry beds [1, 2]. The GIP of the Liard is almost double that of the Horn River Basin, and its ratio of reserves per resource shows the least value among the basins [1]. This indicates that the Liard Basin has a higher potential than the other fields.
The shale gas resource is source rock and reservoir rock; for this reason, the evaluation method for the source rock can be applied for the unconventional resource [3–6]. Many researches have been evaluating shale gas reservoirs by using geochemical analysis and petroleum system modeling. The system modeling gives information about evolution of petroleum system and thermal maturity of regional source rock [7, 8].
However, the model needs various and abundant data ranging from geochemical data to depositional history [7, 8]. Moreover, the generative potential of shale gas resources can be estimated by initial geochemical properties such as the original hydrogen index (HIo) and original total organic carbon (TOCo) [9, 10]. For these reasons, the 3-dimensional static geochemical model will be suggested by a method to define promising spots in the basin, such as Liard Basin, where there is not enough data for petroleum system modeling. In this study, geochemical analysis and well-log data are used to estimate the generative potential based on the original geochemical data in the Besa River Formation, Liard Basin.
2. Study Area and General Geology
The Liard Basin spans across the British Columbia, Yukon, and N.W. Territories [11]. The study area lies in British Columbia; specifically, the area is Series number 94 and Area codes N to O on the NTS (National Topographic System) coordinate system (Figure 1). It stretches 32 km in the east to west direction and 70 km in the north to south direction and has an area of 1,947 km2.
The Besa River Formation was deposited during the middle Devonian to early Carboniferous [11]. The Formation comprises predominantly argillaceous rocks, which are massive to bedded in texture, grey to black in color, and calcareous to noncalcareous [11–14]; minor sandstone and siltstone are interbedded locally [1, 11, 13, 14]. The depositional environment is interpreted as the marine slope into the deep marine environment [11] under anoxic condition [12, 15, 16].
3. Data and Methods
Experimental analysis data by the Rock-Eval pyrolysis and XRF (X-ray fluorescence spectrometer) of four boreholes (Figure 1) were based on data of Hong et al. (A-038-B/094-N-08) [17], Kim et al. (A-068-D/094-O-05) [18], and Choi et al. (B-023-D/094-O-05 and B-D003-K/094-O-12) [19] (Table 1).
The structural frame of the model considered the geological interpretation of the reservoir in order to realize the 3D distribution for geochemical properties (Figure 2). Hence, the shale bed was classified in terms of hydrocarbon generation potential. The correlation of prospective shale beds in the formation nine boreholes was conducted using gamma-ray (GR) log. The distance between the boreholes was considered as the boundary of the study area.
In this study, the original total organic carbon (TOCo) and original hydrogen index (HIo) are determined to estimate the hydrocarbon (HC) generative potential. These were calculated based on mineralogical and geochemical analysis data [9, 10, 15]. The TOCo data cannot be directly obtained from the rock sample directly; hence, the present-day total organic carbon (TOCpd) was determined using bulk density log [20, 21] and Rock-Eval pyrolysis data. Geochemical property models were realized in the structure grid model.
This study undertook 3D shale gas reservoir characterization considering the geochemical properties in the prospective shale beds of the Besa River Formation, Liard Basin. From the model, generated gas volume was estimated, and the volume was compared with the GIP of a production well in this study area and the well location in the model.
3.1. Structure Modeling
To realize the spatial distribution of organic matter, horizontal grid size was considered by distance between the boreholes (); the vertical interval of the grid was considered as 10 m. Hydrogen index (HI) is a generally used indicator to predict thermal maturity for source, but it is difficult to predict thermal maturity exactly in mature and overmature source rocks using estimated present-day HI. For this reason, the original hydrogen index (HIo) is considered to more accurately estimate generated gas volume by TOCo; it was calculated by Equation (1) [9, 10, 15] with data [17–19].
The prospective shale beds were classified as upper and lower beds, termed as A and B shale beds, based on the HIo value. Generally, the stratigraphic correlation between boreholes is based on well-log data; the data have information about lithology and other characteristics for beds. Stratigraphic interpretation in boreholes is based on reference data such as well-header information, mud logging data, and research articles on the same study area. Furthermore, in accordance with subject of the research or project, referred data can be switched or changed with the others. In this study, the top of the prospective shale bed was correlated with the well-log and information. In addition, abrupt vertical variation of HIo in the boreholes was considered for maturity in each bed, because HI is a value to expect thermal maturity in source rock. The A shale bed (upper) shows a number less than 600 mgHI/TOCg and B shale bed (lower) shows a number over than 600 mgHI/TOCg [16] (Figures 3 and 4). The B shale bed is thicker than the A shale bed and tends to get thicker from south to north gradually (Figure 5). Among the nine boreholes having well-log data, only four have geochemical data (Figures 1 and 4).
(a)
(b)
3.2. Original TOC (TOCo) Calculation
The initial generative capacity within the source rocks can be expected in terms of TOCo and estimated to determine the generation potential of the source rocks [10]. The TOCo, which is the initial total organic carbon, is calculated by the empirical Equation (2) [9, 22]. In Equation (2), present-day generative organic carbon (GOCpd) is organic carbon existing in the rock, and it can be possibly transformed to hydrocarbon (HC). In contrast, present-day nongenerative organic carbon (NGOCpd) is organic carbon which does not have any potential to transform the HC; Equation (2) [9, 22] is as follows:
The properties of Equation (2) can be calculated or obtained by the results of the Rock-Eval pyrolysis; however, the data is only available for the four boreholes. Hence, in order to calculate TOCo for the boreholes lacking geochemical data, Equation (2) must be modified. For this reason, the HIo was applied as the average value in the A and B shale beds. Average values for HIo are 447 mgHC/gTOC in the A bed and 670 mgHC/gTOC in the B bed.
TOC consists of present-day generative organic carbon (GOCpd) and present-day nongenerative organic carbon (NGOCpd). The GOCpd is the organic carbon remaining in the present day and has the capacity to transform into hydrocarbons. On the other hand, the NGOCpd rarely transforms into hydrocarbons [19]. The GOCpd and NGOCpd were estimated by using the data of Hong et al., Kim et al., and Choi et al. [17–19]. Matured to overmatured source rock in the A and B beds [12] includes little GOCpd (Figure 6); the value is too low to affect the estimation of TOCo. Hence, GOCpd was not considered in Equation (2).
The relationship between NGOCpd and TOCpd shows a high coefficient of determination over 0.9 (Figure 7). Therefore, NGOCpd substituted the TOCpd value, in Equation (3).
3.3. Calculation of TOCpd from Well-Log Data
Generally, TOCpd is measured from the source rock by geochemical analysis such as the Rock-Eval pyrolysis method. Many researches have been conducting estimation of TOCpd using only well-log data without further geochemical analysis [21, 23, 24].
The Besa River Formation mainly consists of shale with little pyrite, which is considered as a heavy mineral in shale [11–14]. Schmoker’s experimental equation can account for pyrite [20, 21]; hence, the equation was used to consider the influence of the mineral in the Besa River Formation.
Schmoker’s TOCpd was calculated using the bulk-density log of the boreholes having TOCpd via geochemical analysis and then validated against the measured TOCpd of the boreholes. The relationship between measured TOCpd and Schmoker’s TOCpd shows a high coefficient of determination in the range 0.540–0.753 (Figure 8). Thus, Schmoker’s TOCpd was applied to calculate the TOCo. The measured and calculated TOCpd of boreholes are shown in Figure 9.
3.4. 3-Dimensional Geochemical Static Modeling
The 3D distribution of TOCpd was realized in the structure grid model using the moving average method (Figure 10). The moving average method is one of the interpolation methods; this uses an average, which is weighted by inverse of the distance, of surrounding samples to the specific point. The number of boreholes in this study is limited to analysis spatial distribution such as variogram modeling; for this reason, the method is used to realize 3D property modeling. The TOCo model was generated based on the TOCpd model, using Equation (3), where in the average HIo value used was 447 mgHI/gTOC for the A shale bed and 670 mgHI/gTOC for the B shale bed (Figure 11).
(a)
(b)
(a)
(b)
Models for 3D geochemical properties—TOCpd (Figure 10) and TOCo (Figure 11)—were generated for prospective shale beds in the Liard Basin. From TOCo volume and Equation (4) [10], the original generation potential of the shale gas beds could be estimated. The potential value indicated the hydrocarbon content in the rocks; thus, the potential was converted to the generated gas volume (Figure 12).
(a)
(b)
4. Results and Discussion
This study shows the realized 3D distribution of estimated gas volume using geochemical data and well-log data in the Liard Basin. The GGV in the B shale bed is higher than that in the A shale bed (Figure 12). Figure 12 shows that generated gas volume without considering for expulsion of gas volume. This indicates that the B bed has higher generative potential than the A bed. Expulsion efficiencies have to consider estimating accurate gas volume in the shale gas reservoir in the subsurface. Hence, several studies have researched predicting the expulsion efficiency [25–30]. Therefore, in this study, the expulsion efficiency is considered to accurately estimate the retained gas volume which is considered as the GIP in the static model; the efficiency was assumed as 0.6 [10].
The retained gas volume is compared with GIP of the production well (200/C-045-K/094-O-05/2), which is located in the study area (Figure 1). The GIP about the well is 496 Bcf (billion cubic feet)/section, and the retained gas volume is 262 Bcf/section in the well location on the model. Although the order of the GIP value of the well and the model is similar, the value in the model is much lower than that of the well. This indicates that the GIP model estimated from the geochemical static model could suggest the approximate prospective area where there is an undeveloped place in the basin. The gap between GIP values may be caused by the difference in volume of the estimated model and the affected volume due to hydraulic fracturing on the subsurface. It is difficult to predict the reservoir scale of induced cracks around a wellbore due to fracturing without microseismic data.
Furthermore, the model represents the GIP in the study area; hence, it cannot explain petroleum generation, migration, and accumulation in the shale bed because the static model does not mimic changes from basin evolution. On the other hand, petroleum system modeling is another method to estimate generated and retained gas volumes in the source rock; moreover, it can suggest generation and migration time and place [29, 31, 32]. Hence, in order to understand the geological processes, data and information such as temperature, pressure, burial history, thermal maturity, and stratigraphy are required, followed by additional research for petroleum systems and processes.
The initial state of the properties is emphasized by researches [9, 10, 30, 33]; however, these original properties can be calculated indirectly via data by laboratory experiments. This geochemical static model was based on these original geochemical properties, though assumptions and constant values in the overall estimation process should be revised for the entire Liard Basin.
5. Conclusion
This study determines the spatial prospective area via shale gas reservoir characterization for the prospective shale beds of the Besa River Formation in Liard Basin. The stratigraphic correlation among the wells realizes the stratigraphy and structural interpretation results for beds using HIo and gamma rays of the wells. Spatial TOCo distribution is realized by TOCpd and a structure grid model by the correlation result. Furthermore, correlation of determination between measured TOCpd by the Rock-Eval analysis and calculated TOCpd by bulk density log shows from 0.540 to 0.753. So the bulk density log of the boreholes without rock samples provides information regarding the estimated TOCpd value. From the measured HIo and estimated TOCo, generated and retained gas volume (262 Bcf/section) is calculated; this static model may aid in Basin exploration and development plans and function as a preliminary or reference model for the Basin.
Data Availability
The raw data of the geochemical analysis performed in this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP); the Ministry of Trade, Industry & Energy, Republic of Korea (MOTIE) (No. 20178510030880); and the Korea Institute of Geoscience and Mineral Resources (KIGAM) (No. GP2020-006). The authors are thankful to the British Columbia Oil and Gas Commission (BCOGC) for providing valuable field data and the great help for publishing this research article. In particular, the authors are grateful to Mr. Filippo Ferri and Dr. Hamed Senai for their helpful comments on this article.