Although fine-grained, clay-rich rocks with particle sizes less than 1/16 mm account for about 70% of sedimentary rock mass, not all of these are effective source rocks. Petroleum is made up of hydrocarbon molecules originating in a rock in which organic matter (kerogen) is dispersed in the sedimentary formation and has been buried and thus heated enough to transform kerogen to hydrocarbons (i.e. thermal maturity). In addition, the rock formation should have considerable volume (thickness of meters and length of kilometers) in order to generate large amounts of hydrocarbons. Rock volume becomes especially important for exploring shale plays, in which the source, reservoir and seal rock are the same.
The organic richness of source rocks is measured by total carbon content (TOC), which is described in weight percentage of the sample. Rocks are ranked from poor to excellent; higher percentages of TOC indicate organic-rich sedimentary rock.
Direct methods of measuring TOC are made by the Leco carbon analyzer. About a gram of powdered sample is treated with acid to remove the inorganic carbon (carbonate). The dried residue is then mixed with a metal accelerator (iron and copper), and combusted (rapid reaction with oxygen) using a high-frequency induction furnace (1,200–1,400°C). The mass of the carbon dioxide gas thus formed is measured in a non-dispersive infrared detection cell and converted to percent carbon based on the dry sample weight. TOC values can be underestimated if there is loss of immature organic carbon in the acid treatment or overestimated if carbonate is not completely removed prior to analysis.
Indirect methods of estimating TOC include pyrolysis (see below) and well logs. A high content of organic matter in sedimentary rocks gives positive (high) feedback in gamma-ray and neutron logs, and results in high transit time in sonic log, high resistivity and low density (if free hydrocarbon molecules occupy pores). Recently, some geophysicists have also attempted to locate very high TOC sedimentary layers by their significantly low acoustic impedance on seismic data (Løseth et al., 2011).
In interpreting TOC values for prospect evaluation, one needs to consider sample lithology, sampling procedure and the TOC measurement method. For surface sampling, the specimen should not come from the exposed surface, although even then, outcrop samples will often yield lower TOCs compared to samples of the same formation collected from cores because of weathering of surface rocks. Several regularly spaced samples from the entire thickness of a formation should be collected – a single sample from an organic-rich or organic-poor layer of the formation could be misleading.
TOC data based on a large number of sample analyses from around the world provide some guidelines for prospect evaluation, suggesting average TOC of all shales is 0.9%; for shale source rock it averages 2.2%; for calcareous shales 1.9%; for carbonate sources 0.7%; and the average for all source rocks is 2.2% (Tissot and Welte, 1984). In general, dark, laminated clay-rich sedimentary rocks with no bioturbation are indicative of organic carbon content and anoxic depositional environments, thus often making good source rocks.
Thermal Maturity Indicators
Burial temperature ‘cracks’ the complex kerogen molecules into simpler hydrocarbons. There are several methods of estimating the thermal maturity of a source rock. Burial history (stratigraphic time vs. burial depth curve) together with reasonable assumption of geothermal gradients informs us whether the target layer ever resided in the oil or gas window zone.
A common method for estimating sample paleotemperature (maturity) is vitrinite reflectance (VR or Ro). Vitrinite is a type of maceral particle derived from the cell-wall or woody tissue of plants, and is found in coal and kerogen. It has a shiny (vitreous, from the Latin virtum, glass) appearance which irreversibly increases with increasing temperatures (due to aromatization).
VR analysis may be performed on whole rock powder or on kerogen extracts from the rock. The sample is mounted on epoxy resin or similar material and is observed under a reflected light microscope. The reflectance (%Ro = % reflectance in oil-immersion objective lens) of vitrinite particles can be used to evaluate the thermal maturity of rocks. The VR scale has been calibrated empirically and experimentally, and is widely used in the petroleum industry.
At least 50 vitrinite particles should ideally be measured for each sample. The data are plotted in the form of a histogram that gives the minimum, maximum and mean values of Ro as well as the number of data points and standard deviation. Sometimes the histogram gives a complicated distribution with more than one cluster, in which case results should be interpreted considering sample conditions as well as data distribution. Lower Ro values probably indicate contamination of the sample by caving, (i.e. immature kerogen from shallower levels caving into the borehole). Alternatively, higher values may come from reworked kerogen eroded from a more mature rock.
The onset of oil generation corresponds to Ro values of 0.5–0.6% and the maximum generation of oil to 0.85–1.1%. The onset of gas generation (wet gas and condensate) generally corresponds to Ro values of 1.0–1.3% and the gas window continues for Ro values of 3% for dry gas. Temperatures and corresponding Ro values will vary for different types of kerogen present in the source rock.
VR cannot directly be used for samples older than Devonian because land plants only flourished on Earth after then. For Early Paleozoic rocks, graptolite reflectance or conodont alteration index (CAI) can be used. VR scale has also been calibrated for reflectance of bitumen (solid hydrocarbon), so samples lacking vitrinite particles can be measured for bitumen reflectance, and ‘VR equivalent’ values are then obtained. VR equivalent values can also be calculated from pyrolysis (see below).
Another important technique for estimating thermal maturity is the Thermal Alteration Index (TAI) based on color changes (from light yellow through orange and brown to black) in spores and pollen as a result of the burial heat the sample has experienced. Several TAI scales relating the observed color under the microscope to temperature have been formulated by scientists. The oil window corresponds to TAI values of 2.6–3.3.
For samples which have been analyzed by pyrolysis the results include a parameter called Tmax, which is also a paleo-temperature indicator. The oil window corresponds to Tmax values of 435–470°C.
Petrographic methods such as VR provide information about the maximum temperature that the rock experienced in the past but they do not tell us about timing. A comparison of VR or similar data with the burial curve is one way to assess the rock’s thermal history. Another is fission-track thermochronolgy of apatite minerals separated from sandstone layers adjacent to the source rock. Uranium fission tracks in apatite shorten at temperatures of 60–120°C, which correspond to the oil window temperatures. Therefore, by determining the fission-track age and the track lengths in apatite, one can reconstruct the time-temperature pathway of the rock sample (see GEO ExPro Vol 7. no. 1, pg 10, World Leaders in Thermal History).
Pyrolysis is the process of heating a rock or kerogen sample in the laboratory in order to measure the type, richness and maturity of hydrocarbons by thermal decomposition in the absence of oxygen. This can be done either in the presence of water (hydrous) or in its absence (anhydrous). RockEval™, developed by the Institut Français du Pétrole, is a programmed instrument for anhydrous pyrolysis of rock samples. Similar instruments include the Source Rock Analyzer™ by Weatherford Laboratories Instruments and HAWK by Wildcat Technologies.
In pyrolysis, nearly 100 mg of crushed, dried rock sample is progressively heated (~25°C/minute) up to at le ast 550°C under an inert helium or nitrogen atmosphere. The first step is to heat it to 300°C to volatize the pre-existing free hydrocarbons in the sample. The amount of these hydrocarbons is measured from a peak area (S1). The next step is to pyrolyze the kerogen present in the sample at higher temperatures and measure it at peak area S2. The carbon dioxide generated from the cracking of kerogen (up to 390°C) is collected as S3. S1 and S2 are measured by the Flame Ionization Detector (FID), and S3 by the Thermal Conductivity Detector (TCD). Residual carbon in the sample is also collected and measured as S4 (at 600°C). All of these peak areas are expressed in units of mg/g rock. High S1 indicates an active source rock while S2 values measure the remaining hydrocarbon-generating potential of the rock. Tmax is measured from the S2 peak.
Data from pyrolysis are used to calculate the following parameters:
Hydrogen Index (HI) = (S2 x 100) mg/g / %TOC
HI values of >600 indicate oilprone kerogen type I while those of <200 indicate gas-prone kerogen type III
Oxygen Index (OI) = (S3 x 100) mg/g / %TOC
OI values of >50 are indicative of immature kerogen
Production (Generation) Index (PI or GI) or Transformation Ratio (TR) = S1 / (S1/S2)
PI values of 0.1-0.4 correspond to the oil window. Lower values are for immature source rock and higher values indicate gas generation to destruction
Potential Yield (PY) or Generative Potential (GP) = S1 + S2 (mg/g or kg/ton of rock)
PY values of <2 indicate poor yield, those between 2-4 fair and >6 very good
Calculated Vitrinite Reflectance = 0.018 x Tmax – 7.16 (Jarvie, 2012).
Calculated TOC = [0.082 (S1+ S2) + S4] / 10 (Espitalié, Deroo and Marquis et al., 1985)
Calculated TOC is the sum of carbon obtained in pyrolysis of the sample
Kerogen is categorized into four types:
Type I is derived from lipid-rich algal material (alginite) preserved in anaerobic environments, especially lacustrine and similar marine conditions. This kerogen is abundant in aliphatic (non-aromatic) compounds, and is highly oil prone.
Type II, rich in liptinie macerals, is derived from waxy and resinous parts of plants including extinite (skins of spores, pollen, and cuticles of leaves and herbaceous plants) and degraded, amorphous phytoplankton from marine environments. This type commonly has higher amounts of sulfur than other kerogens. It is mainly alicyclic (naphthenic) and produces moderate amounts of both oil and gas.
Type III is relatively hydrogen-poor, predominantly vitrinite macerals, derived from terrestrial woody and fibrous plant fragments and structureless, colloidal humic matter. It is rich in aromatic compounds, and mainly gas prone, although liptinite kerogen, if present >15% in the sediments, also generates oil. Kerogen type III is deposited in terrestrial or shallow marine environments. Coals usually contain this type of kerogen.
Type IV refers to extremely hydrogen-poor, carbon-rich residual (reworked from older sediments) or oxidized kerogen. It is the chemical equivalent of inertinite maceral group, and is considered as dead carbon with no potential for generating oil and gas.
The Dutch chemist Dirk Willem van Krevelen (1914–2001) noted that various kerogens can be recognized on the basis of their atomic H/C versus O/C plots. In the 1970s French petroleum chemists Bernard Tissot and J. Espitalié extended van Krevelen’s work from coal to kerogen and substituted the elemental parameters with Hydrogen Index and Oxygen Index obtained from pyrolysis (a pseudo- or modified Van Krevelen diagram). In this analysis, kerogen type I has H/C >1.5 and HI >600; type II has H/C of 1.0 to 1.5 and HI of 200 to 600; type III has H/C of 0.7 to 1 and HI of 50 to 200; and kerogen type IV has H/C of <0.7 and HI of <50 (Peters and Cassa, 1994).
Geochemical Logs and Basin Modeling
Geochemical data from single or multiple source rock horizons can be plotted as geochemical logs for wells in order to compare and contrast the data vertically and horizontally. If there are considerable data from various wells, specific parameters, such as TOC or VR, may also be plotted on formation top maps and contoured to decipher lateral variations. Basin modeling indeed began with source rock burial-maturity-generation modeling in the early 1980s, and source rock data still constitute the critical input data for basin modeling software packages.