Reed A.Johnson, R. Timothy Bartshe
Energy Foundation Inc.
Lakewood, Colo
Recent interest in U.S. horizontal drilling has largely focused on vertically fractured plays such as the Bakken shale and Austin chalk.
The Upper Cretaceous Niobrara formation, the chronological equivalent of the Austin chalk, has recently been targeted as a candidate for horizontal drilling in the Denver basin and other areas of the Rocky Mountains.
A primary key to success in such plays is to predict the occurrence and distribution of oil bearing fracture systems. Much emphasis is placed on theoretical aspects of fracture origin and prediction.
Remote sensing techniques (e.g., seismic, satellite image analysis) have gained wide use in the search for fractured reservoirs. While these methods are important elements of an integrated exploration effort, they lack the benefit of direct detection of open, oil saturated fracture systems.
In the areas of the Denver basin in which the Niobrara is oil prone, certain resistivity responses are indicative of the proximity of oil bearing fractures to the well bore. This provides an extremely useful technique in areas of pre-existing well control penetrating the Niobrara section.
As such, the Denver basin is an ideal area due to the large number of penetrations to the Lower Cretaceous D and J sandstones that underlie the Niobrara.
The following discussion will examine the factors affecting resistivity response, methods of calculating anomalous resistivity, and the empirical corroboration of the effect in areas productive from the Niobrara.
NIOBRARA PETROLEUM GEOLOGY
Production from the Niobrara occurs in two general areas (Fig. 1).
Gas is produced at depths ranging from less than 1,000-ft to 3,000 ft along the gently dipping eastern flank of the Denver basin in eastern Colorado and western Kansas.1 Gas accumulations are associated with local structural traps. Porosity tends to be very high and permeability relatively low.
Local fracturing may enhance production characteristics, and hydraulic fracturing is generally required to stimulate production. The origin of the gas is generally considered to be related to biogenic decomposition of organic matter.2
In the areas of the Denver basin that are or were more deeply buried, the Niobrara produces thermogenic oil and wet gas condensate.3 4 Production of this type is generally at depths greater than 6,000 ft with the exception of fields bordering the Front Range uplift.
Fields of this type that have been productive from the Niobrara include Surrey field in Adams County, Colo.; Spindle, Wattenberg, Roggen, and Lost Creek fields in Weld County, Colo.; Berthoud, Loveland, and Fort Collins fields in Larimer County, Colo.; Silo field in Laramie County, Wyo., as well as numerous more isolated locales.
The following discussion will be concerned primarily with the deeper, thermogenic, oil prone areas of the Denver basin.
The Niobrara formation in the Denver basin consists of alternating cycles of marine limestone (chalk) and calcareous, organic shales. The chalks and calcareous shales are relatively rich in preserved organic carbon.
The organic content of the overall Niobrara ranges from 1-6 wt %.2 5
The Niobrara is divided into two members, the Fort Hays limestone member at the base of the formation and the overlying Smoky Hill member (Fig. 2). The Smoky Hill member is further subdivided into three carbonate intervals, referred to as the A, B, and C chalks, in descending order. Each chalk interval is underlain by a calcareous shale interval.
Following deposition, the Niobrara formation was buried coincident with the subsidence of the Denver basin. The majority occurred during late Cretaceous through early Eocene time. The diagenetic effects of burial consist primarily of mechanical and chemical compaction and the thermal maturation of organic matter.6 7
The combined result of the compaction processes is a progressive reduction in porosity with depth. At the shallow depths associated with biogenic gas production (less than 3,000 ft), porosity is 25-45%. At depths greater than 6,000 ft, porosity of the chalks is generally less than 10%.
Permeability varies linearly with porosity such that chalks with 10% porosity or less (greater than 6,000 ft depth) generally exhibit permeability of less than 0.1 md. Therefore matrix permeability is not a significant factor in production of Niobrara oil, and fractures are required to transmit fluid to the well bore.
When buried to a sufficient depth for a sufficient period of time (or when exposed to anomalously high heat flow), the organic matter within the Niobrara formation becomes thermally mature and generates hydrocarbons.
Based on vitrinite reflectance studies,8 9 time-temperature index calculations,10 and clay geothermometry studies,6 11 the Niobrara is thermally mature over much of the deeper portion of the Denver basin (Fig. 1).
In this area, the Niobrara is generally within the oil generation window, with the exception of Wattenberg field area, which is in the wet gas/wet gas condensate window."
Based on burial history reconstruction and depth vs. maturation relationships, the Niobrara in the axis of the basin most likely began generating oil in the Eocene. Because the Niobrara is self-sourcing and the generated oil migrates only locally, the area of thermal maturity (the "maturation envelope") defines the boundary of the exploratory area of interest (i.e., for fracture-mediated oil production).
The most likely fracture systems to have been available for oil saturation are those that were either generated prior to oil expulsion and remained open, or were generated during or after oil expulsion.
The common occurrence of large Laramide structures within the maturation envelope with no apparent Niobrara production (e.g. Chivington, Golden Prairie, and Borie fields, Laramie County, Wyo.) suggests that the latter case is more likely. Such large structures undoubtedly generated fractures in the Niobrara, but the absence of oil production suggests that the fractures were not open at the time of Niobrara oil expulsion.
The essential concepts important in exploring for fractured Niobrara oil reservoirs include 1) the Niobrara is a self-sourcing reservoir, 2) production is mediated by vertical (or sub-vertical) fracture systems, 3) matrix pore systems may provide storage volume but minimal transmissivity, and 4) oil migration is likely very localized, therefore the exploratory area of interest is largely defined by the area of thermal maturation (Fig. 1).
RESISTIVITY CONCEPT
The basic factors that affect measured resistivity are temperature, mineralogy, porosity, pore geometry, pore fluids, and geometric effects.
The relationship of many of these factors is expressed in a variation of the 1942 Archie formula as follows:
[SEE FORMULA]
where Rt is the resistivity of rock, Rw is the resistivity of formation water, Sw is water saturation, f is porosity, m is cementation factor, and n is the saturation exponent.
Resistivity of a rock increases with increasing water resistivity, increasing oil saturation (decreasing water saturation), and decreasing porosity. Pore geometry is expressed by the cementation factor m.
As the path through the pore system becomes increasingly restricted and convoluted, the value of m increases and results in a higher value of measured resistivity.
Temperature has a profound effect on resistivity in rocks that contain water, either in movable or bound form. The relationship between the resistivity of aqueous solutions (NACI) and temperature is given by the Arps formula:13
[SEE FORMULA]
where Rft stands for resistivity at formation temperature, R75 is resistivity at 75 F., and Tf is formation temperature.
As temperature increases, resistivity of the solution decreases. Similarly, for a rock containing water in its pore system or bound to the mineral surfaces, resistivity of the rock decreases as temperature increases.
In general, unless a localized heat flow anomaly exists, temperature is not a significant factor when comparing resistivity values from a single zone within a local area (i.e. on the scale of a small field).
However, when comparing resistivity measurements in a regional area where significant changes in formation temperature do occur, temperature becomes a very important factor. In order to make such comparisons, the resistivity values must be corrected to a common temperature.
Mineralogy may also affect the value of measured resistivity. The most common mineralogic effect is the reduction of resistivity as occurs with increased clay content or a high concentration of a conductive mineral such as pyrite.
The Niobrara may exhibit widely ranging resistivity values in the cleaner chalks on many scales; regional, local (i.e. between adjacent wells), and within a single chalk bed. The range of values may cover more than an order of magnitude.
The causes of variability must account for the magnitude and the various scales of variability. In general, changes in water resistivity, porosity, and pore geometry cannot account for the resistivity variability on all scales, especially the latter two. The most probable causes are changes in oil saturation and geometric effects (discussed below).
Elevated resistivity values associated with thermal maturation have been recognized in many self-sourcing strata, including the Bakken shale,14 Woodford shale,15 and Austin chalk.16
Resistivity values are observed to shift to markedly higher values once the threshold of thermal maturation is exceeded. The mechanism of this effect has been ascribed to the progressive displacement of matrix pore water by generated hydrocarbons as thermal maturation proceeds.
Under this scenario, resistivity could theoretically approach infinity as oil saturation approaches 100%. However, unless the rock becomes oil wet, some water will remain in the pore system, thereby creating an upper limit of oil saturation.
The limit of oil saturation is generally lowest (higher residual water saturation) in fine grained rocks such as in Niobrara chalks. It should be expected then that there will be an upper limit of resistivity values due to matrix oil saturation well below infinity.
Smagala et al. have compared resistivity and the degree of thermal maturation in the Niobrara formation in the Denver basin. Vitrinite refectance values (Ro) measured or interpolated for the Niobrara were compared to maximum resistivity values from the Smoky Hill C chalk throughout the basin.
Resistivity generally increases with higher values of vitrinite reflectance. The mechanism of this effect was postulated to be due to the progressive increase in matrix oil saturation in the chalk as pore water was displaced by oil migrating from adjacent source layers.
The relationship between resistivity and vitrinite reflectance was interpreted to be linear (Fig. 3A); therefore, very high resistivities were considered to result solely from increased matrix oil saturation associated with increasing thermal maturation. However, their conclusions are in doubt due to the fact that the data set covers a wide range of depth and temperature, and the resistivity values were not corrected for temperature variation.
Due to the strong influence of temperature on measured resistivity, in comparing resistivity values in a region with widely varying depth and formation temperature all values must be corrected to a reference temperature.
Temperature correction of resistivity values read from well logs requires three steps. First, the measured bottom hole temperature must be corrected because it reflects a disequilibrium condition due to the cooling effects of mud circulation.14 17
Second, the Niobrara formation temperature is calculated from the corrected bottom hole temperature using the mean annual surface temperature and the calculated geothermal gradient.
Third, the resistivity measured at the corrected formation temperature (i.e. read from a log) is corrected to a common reference temperature by using the Arps formula.
Once corrected to the reference temperature, resistivity values from widely separated locales may be compared.
The data from Smagala et al. have been corrected to a reference temperature of 75 F. and replotted on a linear scale (Fig. 3B). The corrected data exhibit markedly different trends from the single linear trend interpreted from the uncorrected data.
Four distinct trends are evident. At low levels of thermal maturation (Ro < 0.6%), resistivity increases linearly with R. Because R, increases linearly with burial depth (in the absence of heat flow anomalies12), this relationship reflects the progressive loss of porosity with initial burial and the resultant increase in resistivity.
At Ro values of 0.6%-O.7%, resistivity shifts abruptly to higher values, reflecting the onset of thermal maturation. The large increase in resistivity over a narrow range of Ro is caused by the displacement of matrix pore water by indigenous oil.
For Ro values greater than 0.7%, two trends are evident. One trend displays resistivity that increases linearly with increasing thermal maturation. This trend parallels that of initial burial and is related to the combination of continued loss of porosity with further burial and the progressive increase in matrix oil saturation with continued maturation. This trend defines the baseline resistivity due to matrix oil saturation at various levels of thermal maturity.
The fourth "trend" consists of an area of anomalously high resistivity values relative to the degree of thermal maturation. The random scatter of points and the large resistivity values that exceed those defined by the linear, matrix oil saturation baseline suggest a mechanism unrelated to a simple, progressive increase in matrix oil saturation.
The points in this area are attributed to the presence of oil bearing fractures in close proximity to well bores (i.e. within the radius of investigation of the resistivity tools).
Note that no anomalous points occur in the thermally immature area (Ro < 0.6%).
The matrix oil saturation baseline of Fig. 3B defines the 75 F. threshold resistivity (Rth75), or the temperature-corrected resistivity that corresponds to matrix oil saturation at a given level of thermal maturation.
Resistivity values in excess of Rth75 indicate oil bearing fractures.
If the degree of thermal maturation for a given area is known, Rth75 may be estimated from Fig. 3B.
Correction of Rth75 to formation temperature (using the Arps formula) gives the thresholds resistivity (Rth) as recorded on a log.
The threshold resistivity is not a constant in all portions of the basin but is a function of formation temperature (present depth, geothermal gradient) and the degree of thermal maturation.
The mechanism of the anomalously high resistivity response to oil bearing fractures is twofold. The first mechanism is related to the volumetric effect of oil bearing fractures on the resistivity tool.
Matrix pores contain both oil and water and result in measured resistivity equal to or lower than the threshold value.
The presence of a significant pore fraction of 100% oil saturated fractures within the radius of investigation of the resistivity tool will cause an elevated resistivity response due to the averaging of partially saturated matrix and fully saturated fractures. That this occurs is not always evident because fracture porosity is commonly not detected on neutron and density porosity logs due to the limited area of investigation.
The second mechanism is related to the geometric confining effect of oil saturated fractures. Resistivity values are calculated by the logging panel from measured values of voltage potential and known values of current (e.g. for conductive resistivity tools).
Implicit in the calculation is an assumed geometry of investigation that is expressed as a sonde constant.
The assumed geometry of investigation consists of the radius of investigation of the tool and the electrode spacing. For the measured value of resistivity to reflect true resistivity, the tool must be able to investigate this volume in an unconfined manner.
The presence of an oil bearing vertical fracture within the radius of investigation of the tool will act as a planar insulator and effectively confine the area of investigation. The degree of confinement increases as the fracture approaches the well bore. Resistivity calculated under this circumstance will not be a true resistivity but an apparent resistivity.
The effect of such confinement will be an increase in the measured voltage potential, proportional to the degree of confinement, and an increase in the computed value of apparent resistivity.
To the extent that oil bearing vertical fracture systems act as barriers to electrical current flow, the magnitude of the increased resistivity will be related to the degree of insulation provided by the fracture system and the distance of the fracture system from the well bore. In this regard, the resistivity response may be considered semi-quantitative.
The limit of proximity that can be detected is determined by the radius of investigation of the resistivity tool. For most deep-reading tools the radius is effectively limited to 20 ft or less.
MEASUREMENT OF ANOMALOUS RESISTIVITY
Given the concept of threshold resistivity (Rth) and the definition of resistivity values in excess of the threshold resistivity as anomalous, how are resistivity anomalies best measured for comparison purposes and mapping?
Resistivity anomalies consist of three components; the threshold resistivity, the maximum resistivity in excess of the threshold, and the thickness of section that exceeds the threshold (Fig. 4). The latter two components are measured relative to the threshold resistivity and are referred to as "relative components."
Single component methods of measurement consider only one of the relative components or some associated derivative, e.g. thickness of section in excess of the threshold, maximum resistivity, or the ratio of maximum resistivity to threshold resistivity. Single component methods suffer from incomplete representation of the data.
Dual component methods combine both relative components in some fashion. Perhaps the most meaningful method of combining the relative components is a measure of the area of the resistivity curve in excess of the threshold value.
Due to the nonuniform shape of this area, the exact calculation would be time consuming and cumbersome. As an approximation of the area, the area of a triangle with a base equal to the thickness of zone exceeding the threshold (t) and a height equal to the difference between the maximum resistivity (Rmax) and the threshold resistivity (Rth) may be calculated (Fig. 4). This is referred to as the triangulated resistivity anomaly (TRA):
TRA = 1/2 t(Rmax - Rth)
For logs with depths measured in feet, the units of TRA are expressed as ohm-meter-feet. If multiple, separate zones in a single well exceed the threshold resistivity, each area may be calculated separately and added together.
CONFIRMATION: 3 BERTHOUD STALE
A cored Niobrara interval from the Coquina Oil Corp. 3 Berthoud State, NE SW 16-4n-69w, in Berthoud field, Larimer County, Colo., directly demonstrates the relation between oil bearing fractures and resistivity. Anomalously high resistivity is correlative to a highly fractured interval that contains oil stain in open fractures and recovered significant free oil during drillstem tests.
The entire Niobrara formation was cored in the 3 Berthoud State (Fig. 5). Fractures are evident in the core in specific intervals,18 principally in the cleaner chalks.
The fractures in all but two intervals consist of vertical to subvertical, calcite cemented hairline fractures with no visible evidence of oil. The single heavily fractured interval from approximately 3,070-3,106 ft (C chalk) contains partially open, oil stained fractures.
These fractures are oriented from vertical to near horizontal, are partially cemented by calcite, and exhibit evidence of fault movement. The uncemented aperture of these fractures ranges from a fraction of a millimeter to more than 4 mm.
Five drillstem tests were run in the cored interval (Fig. 5). Two tests (DSTs No. 3 and 4) that covered the heavily fractured C chalk interval each recovered 500 ft or more of gas-cut oil. Tests covering intervals exhibiting only hairline fracturing recovered substantially less oil and, in the case of DST No. 2, no free oil.
The fact that some intervals containing hairline fractures recovered some oil indicates that the hairline fractures likely have some permeability.
The heavily fractured C chalk interval exhibits anomalously high resistivity whereas the other fractured intervals do not. Using a vitrinite reflectance value of 0.7% (Crysdale and Barker, 1990) and a corrected formation temperature of 120 F., the calculated threshold resistivity for Berthoud field is 42 ohm-m.
The measured resistivity of the C chalk exceeds the threshold value from 3,074-3,106 ft and reaches a maximum resistivity of 120 ohm-m, 78 ohm-m in excess of the threshold. The resultant calculated TRA value is 1,248 ohm-m-ft.
Anomalously high resistivity correlates to the interval that is most heavily fractured and from which the largest volumes of oil were recovered in drillstem tests. As previously discussed, the increase in resistivity is unlikely to result from changes in pore geometry, matrix pore fluid chemistry, or matrix water saturation.
A crossplot of porosity vs. resistivity indicates no causal relationship (Fig. 6).
Three populations are apparent on the crossplot that are inferred to represent clayrich facies (low porosity, low resistivity), purer chalks containing oil in the matrix only (moderate porosity, resistivity < Rth), and fractured chalks with oil in matrix and fractures (moderate porosity, anomalously high resistivity).
Note that the point of separation between the latter two populations occurs between 40 and 50 ohm-m (Rth = 42 ohm-m). The cause of the anomalously high resistivity is confirmed to be directly related to the presence of oil bearing fractures.
End Part 1 of 2.
(Next: Application to Silo field.)
Copyright 1991 Oil & Gas Journal. All Rights Reserved.
