From Reservoir to Microscopic Scales: What Do We Learn Through the Characterization of Tight Gas Reservoirs?

Jiajie Chen¹, Mark E. Petersen¹, and William Jamison^2
1Marathon Oil Company
²The Upper Crust Inc.

To objectively characterize tight gas reservoirs and formulate the most feasible field development plan, detail fracture analysis should be performed in conjunction with integrated reservoir modeling, deposition and deformation evaluation, as well as petrographic, cementation and structural diagenesis analysis. Here we present a case study from a Canadian tight gas field as an example to illustrate this concept.

The study area is located in the eastern margin of the Rocky Mountain Foothills belt in southwestern Alberta, Canada, with complex thrust faults and imbricate folds. With large area of distribution, the field and its nearby plays possess significant reserve potential. However, without detail analysis from field to microscopic scales, the resource potential and associated risks in economic development can not be fully understood. 

A comprehensive approach was applied to characterize this tight gas field. The method involves a number of steps, including construction of a full field static reservoir model to assess the total resource potential, deposition environment analysis to understand the influence of original sediment inputs, deformation evaluation to reveal the types, pattern and openness of various fracture sets, and fracture diagenesis analysis on microscopic scale to understand the cementation history and degradation of the fracture systems.

Total Reserve Assessment Based on 3D Static Reservoir Modeling

The model area is covered with several 2D seismic lines, including one strike line (S-Line-A) and four dip lines (D-Line-1,2,3 and 4). At the time of this study, four wells have been drilled in the area, abbreviated as W1, W2, W3 and W4 in this paper for convenience. There data were used to construct a general 3D reservoir geological model, from which the total resource potential in terms of original gas in place (OGIP) was estimated. In the study area, stacked faults have separated the depositional sequences into three fault sheets, namely sheets A, B and C from the base to the top. Our modeling effort has been mainly focusing on the third sheet where most of the reserve has been found. For geocellular model layers 38-107 that cover the main reservoir intervals, our results indicate that the OGIP is at the range of 850-900 BCF. 

However, not all the pore spaces that are not occupied by water are occupied by moveable gas. In reality, fluid only moves in the interconnected pores with sufficient permeability and pressure to support the flow. To define the interconnected gas zones, we applied a set of cut-off as below: porosity > 2.5%, water saturation < 85%, permeability > 0.0001 md, and Gas_Index < 0 (see text below for definition of Gas_Index). This modeling process results in identification of eight geobodies, and the total reserve contained in these geobodies is around 240 BCF.

Obviously, the reserve potential is great. The question is whether rock and fracture properties support economic flow rates, a point we will address below.

Influence of Original Depositional Environment

Gas produced in the field concerned is from the upper Cretaceous Belly River Group, a marginal marine-deltaic-fluvial clastic wedge comprised of mid-fine grained volcanic arenite. The major producing interval, the Basal Belly River with average thickness of only about 15 meters, is located at depth ranging from 2000 to 2500 meters. This interval accounts for 51% of total gas production. 

We observe two stratigraphic cycles below the Basal Belly River marker. Both cycles started with shallow marine shale at the base, and were represented by a coarsening-upward sequence. The lower one consists of prodelta turbidites of the Lees Lake Formation, followed by distributary channel of the Burmis Formation. The second one starts from the marine shale of the Milk River Formation, evolving upward to the marginal marine shale of the Pakowki Formation, with couple of prodelta sands embedded in the overall marginal marine shale interval. It further evolved to the distributary channel of the Basal Belly River interval. Such coarsening-upward cycles are typical of deltaic origin, often favoring gas migration and accumulation at the top of the cycles. This model differs from some previous studies that ascribed the Basal Belly River interval to the beginning of a fluvial cycle. 

Due to the presence of volcanic components, some of which have altered to clay, overall porosity and permeability are significantly reduced, making the sand intervals extremely tight. This presents a challenge to identify zones with relatively better reserve quality in this field.

Rock properties such as rigidity (µ) and “incompressibility” () can be derived from shear and compressive sonic velocities. A mathematical expression using µ, and density (), -µ, referred as Gas Index in this study, was used to help discriminate gas sand from wet sand and further from shales, as suggested in previous studies. Our results indicate that the Gas Index successfully predicts existence of gas-charged sands in the study area. Except for the few intervals identified with favorable Gas Index values, the matrix is so tight in most of the area that even if the rocks are fractured, there is no source in the matrix to feed the fractures with gas.

Evaluation of Deformation Characteristics

Fracture trends    Core and image log analysis suggests that distinct fracture trends exist in each thrust sheet. In sheet 3, shear fractures have a dominant trend ranging from ENE-WSW to ESE-WNW. Extension fractures trend predominantly NE-SW except for the W3 well, where the dominant trend is NW-SE.

Fracture densities    Fracture densities are generally low to moderate away from faults or the limb/hinges of tight folds. Near to faults and within the tight folds, fracture densities are very high but generally only shear fractures present. Extensional fractures are quite low throughout the entire structure, except for sheet 3 of the W2 well.

Infilling of fracture aperture    Shear fractures developed in the Belly River Sandstones typically are infilled with a combination of gouge and clay/carbonate cements. These same cements are also common in the apertures of extension fractures. The mineral cements and gouge significantly degrades the fractures as flow conduit. Extension fractures with only partial cement infill have the greatest potential for enhancing reservoir quality, but their occurrence is very limited. 

Fracture Diagenesis and Degradation

The deformation pattern analysis conducted under core and image log scales were further supplemented by fracture diagenesis analysis at microscopic scale. Petrographic analysis of four samples from different zones reveals that percentages of quartz are similar for all samples (within the range of 32-35%), suggesting that these samples have the same grain characteristics. However, further fracture diagenesis analysis indicates that it is the percentages of postkinematic cements in the mineral forms of calcite, dolomite and kaolinite that really impact the degradation of the fractures as a flow conduit. Results show that except for one of the four samples (GR004-V-481-11), all fractures have high degradation due to high degree of fracture infilling cementation. The sample with low degradation has higher secondary porosity, as the fractures are not completely sealed. These results provide additional evidence in line with our observation from regional depositional environment analysis and deformation evaluation.

Summary

The field evaluated in this study obviously possesses large resource potential. However, with unfavorable accumulation of volcanic arenite, the matrix of the reservoir rock has been extremely tight at its very beginning. Zones of strong deformation were strongly dominated by shear fractures. In this lithology, the shear fractures are not expected to carry flow, nor respond well to hydraulic fracture simulation. Natural fractures that are not completely occluded by mineral cement or gouge are small in number. As a result, fracture density alone can not determine the reservoir potential. Structural diagenesis analysis is an important tool in the determination of overall fracture connectivity. With the poor matrix properties, the creation of hydraulic fractures is necessary to establish an open fracture density adequate for a reasonably well interconnected fracture network that will support economic flow rates. However, hydraulic fractures by themselves do not appear to create an effect flow network. The original depositional environment must also be taken into account. These results provide important input on decision making during field development.

Copyright © 2004. The American Association of Petroleum Geologists. All Rights Reserved.