Apparent Matrix
Apparent matrix is a petrophysical concept describing the effective solid-framework properties inferred by combining two or more porosity-sensitive wireline log measurements to determine the dominant mineral or mineral mixture in a formation without requiring core samples. When a density log and a neutron log (or a sonic log) are both available, each tool responds to the porosity of the rock and to the mineralogy of the solid matrix in predictable but different ways: the density tool measures bulk electron density (which reflects grain density and pore fluid density), while the neutron tool responds primarily to hydrogen index (which reflects bound and pore water, clay water, and hydrocarbon content). By plotting both responses together and assuming specific fluid properties, the analysis can solve for the combination of grain density (rho_ma) and neutron response (phi_N_ma) that accounts for the observed log readings at the measured porosity. The resulting point in density-neutron or sonic-density space is called the apparent matrix point, and its position relative to known pure-mineral endpoints (quartz sandstone, calcite limestone, dolomite, anhydrite) identifies the lithology or mineral blend. The concept is foundational to quantitative log interpretation because it allows lithology identification in real time from logs alone, which is essential for computing accurate porosity, water saturation, and reservoir volume estimates throughout the global oil and gas industry.
Key Takeaways
- The density-neutron crossplot is the primary tool for locating the apparent matrix point in log analysis: On a standard crossplot, the x-axis shows the neutron log reading (phi_N, calibrated to limestone) and the y-axis shows the density log reading (rho_b or the density-derived porosity phi_D, also calibrated to limestone). For a given fluid assumption (typically freshwater, rho_f = 1.0 g/cc, or saline formation water at 1.1 g/cc), data points from a clean formation plot along a trend line connecting the fluid point (100 percent porosity) to the apparent matrix point (zero porosity). Pure mineral endpoint lines for quartz sandstone (rho_ma = 2.648 g/cc, phi_N_ma approximately -0.02 in limestone units), calcite limestone (rho_ma = 2.710 g/cc, phi_N_ma = 0.00), dolomite (rho_ma = 2.870 g/cc, phi_N_ma approximately 0.02 to 0.04), and anhydrite (rho_ma = 2.960 g/cc, phi_N_ma approximately -0.01) are plotted as reference. A formation composed purely of quartz sandstone at 20 percent porosity will plot near the quartz endpoint trend; a limestone will plot near the calcite trend; a dolomite will appear between the calcite and dolomite lines. Mixed lithologies (such as a calcareous sandstone or a dolomitic limestone) plot between the pure mineral lines, and the crossplot position identifies both the dominant mineral and the volumetric proportions of the mixing minerals by interpolation.
- Clay minerals shift the apparent matrix point off the clean-formation trends in ways that fingerprint clay type and volume: Smectite (montmorillonite), illite, kaolinite, and chlorite each have characteristic density and neutron responses that differ substantially from clean quartz or carbonate minerals. Kaolinite has a grain density of approximately 2.41 to 2.44 g/cc and a high bound-water neutron response of 35 to 40 percent in limestone-calibrated neutron units; illite has a grain density of approximately 2.52 to 2.58 g/cc and a neutron response of 20 to 30 percent; chlorite has a grain density of 2.6 to 2.9 g/cc (variable with iron content) and a neutron response elevated by structural hydroxyl and interlayer water. When clay-bearing sandstone is plotted on the density-neutron crossplot, the data points lie off the clean quartz line toward the clay mineral endpoint, and the direction of displacement identifies the clay type: displacement toward high neutron and low density indicates smectite or mixed-layer illite/smectite; displacement primarily toward high neutron with modest density change suggests kaolinite; displacement toward high neutron with near-normal density suggests chlorite (which has a high photoelectric factor). In the Viking Formation of central Alberta, chlorite-coated grains cause a systematic neutron elevation of 5 to 10 percent above the clean sand trend, so apparent matrix identification on density-neutron crossplots must account for chlorite endpoint before the porosity can be corrected for the clay effect.
- The Matrix Identification (MID) crossplot extends apparent matrix analysis to three-mineral systems using sonic, density, and neutron: The MID plot (developed by Schlumberger in the 1970s) uses the apparent matrix density (rho_ma*) and apparent matrix travel time (DTma*) computed from two independent crossplot pairs: (a) the density-neutron pair gives rho_ma* = (rho_b - phi_D_N × rho_f) / (1 - phi_D_N), where phi_D_N is the porosity from the density-neutron crossplot; and (b) the sonic-density pair gives DTma* = (DT - phi_D_N × DTfluid) / (1 - phi_D_N). Plotting rho_ma* on one axis and DTma* on the other creates a two-dimensional space in which the four main mineral endpoints define a quadrilateral: quartz (2.648 g/cc, 55.5 us/ft), calcite (2.710 g/cc, 47.5 us/ft), dolomite (2.870 g/cc, 43.5 us/ft), and anhydrite (2.960 g/cc, 50.0 us/ft). Data from a formation cluster near the mineral endpoint that matches its composition; a reservoir with 70 percent dolomite and 30 percent limestone plots between the dolomite and calcite endpoints in proportion to the mixing fractions. In Nisku carbonate reservoirs of the Pembina area, where alternating dolomite and limestone laminations are common, MID crossplots from FMS formation scanner or standard triple-combo log suites are used to determine the dolomite/calcite ratio and compute a corrected porosity using the actual mineral-weighted grain density rather than an assumed 100 percent limestone or 100 percent dolomite matrix.
- Apparent matrix analysis is used to correct porosity calculations that would otherwise assume the wrong mineral and produce systematically biased results: Porosity from the density log is computed as phi_D = (rho_ma - rho_b) / (rho_ma - rho_f), where rho_ma is the input grain density. If the logging software assumes quartz sandstone (rho_ma = 2.648 g/cc) but the actual rock is dolomite (rho_ma = 2.870 g/cc), the computed porosity will be too low by approximately (2.870 - 2.648) / (2.870 - 1.0) times the actual porosity: for a 15 percent porous dolomite, the sandstone assumption would compute phi_D = (2.648 - 2.481) / (2.648 - 1.0) = 0.101, an error of nearly one-third of the true porosity. The apparent matrix crossplot identifies the correct rho_ma before computing porosity, and in mixed or uncertain lithologies, the MID-derived rho_ma is used directly in the density-porosity calculation to avoid this systematic bias. In Leduc reef carbonates of the Rimbey-Meadowbrook trend, where massive dolomitisation of an original calcite reef fabric is incomplete and variable, well-by-well MID analysis of the triple-combo log suite routinely shows rho_ma* values ranging from 2.72 to 2.84 g/cc within a single well, requiring a depth-varying grain density input to the porosity computation rather than a fixed single-mineral assumption.
- Photoelectric factor (PEF) provides an independent apparent matrix indicator that is sensitive to atomic number and therefore highly discriminating for calcite versus dolomite versus sandstone: The photoelectric factor, measured by the litho-density tool's low-energy detector window, is approximately proportional to the fifth power of the average atomic number divided by twice the average atomic mass (Z/A), making it sensitive to heavy elements such as calcium, magnesium, iron, and barium but nearly insensitive to porosity and fluid type when formation water salinity is moderate. Pure quartz has PEF approximately 1.81 barns per electron, calcite 5.08, dolomite 3.14, anhydrite 5.05, and siderite 15.0. By combining PEF with the apparent matrix density from the density-neutron crossplot, a ternary crossplot (PEF on one axis, rho_ma* on the other) can distinguish calcite from dolomite (which have similar densities but very different PEF values) and identify siderite or pyrite-cemented zones by their anomalously high PEF readings. In the Glauconitic Formation of the Medicine Hat area, the presence of glauconite (a green clay mineral with iron content giving PEF of approximately 6.5 to 7.5) causes PEF values elevated above both quartz and calcite, and the PEF-versus-rho_ma* crossplot immediately identifies glauconitic intervals that require modified porosity and saturation equations relative to the clean sandstone zones in the same well.
Apparent Matrix in Formation Evaluation, Lithology Identification, and Reservoir Characterisation
The practical workflow for apparent matrix analysis begins with quality-controlled log curves: the density log corrected for borehole washout using the density correction curve (delta_rho), the neutron log corrected for temperature and salinity, and (where available) the PEF curve and the sonic compressional slowness. In intervals where the density correction exceeds 0.15 g/cc (indicating significant borehole rugosity), the crossplot data from those depths is flagged as unreliable and excluded from the matrix identification, as washed-out borehole conditions introduce positive errors in the density-derived porosity that will place the data point in the wrong position on the matrix identification crossplot and produce an erroneously low apparent grain density. This quality screening is standard practice in any formation evaluation study in the WCSB because the Montney and Duvernay formations frequently have significant silica, dolomite, and clay volumes that interact with the crossplot in complex ways, and contaminated density data would produce incorrect lithology calls that propagate into the reserve estimate.
In shaly sand formations such as the Viking of central Alberta, apparent matrix crossplots show a systematic trend from the clean quartz endpoint (for the highest-resistivity, lowest-GR intervals at the top of fining-upward sand packages) toward a clay endpoint (for the most shale-prone intervals at the base of the packages). The slope and direction of this trend identifies the dominant clay type in the formation: a trend toward low density and high neutron (typical of smectite or mixed-layer illite/smectite) differs from a trend toward moderate density elevation with high neutron (typical of illite) or from a trend toward density values near quartz but with substantially elevated neutron (characteristic of chlorite). This clay-typing from the crossplot controls the choice of shale correction model applied to the water saturation equation: formations with large CEC (cation exchange capacity) clays such as smectite require a dual-water or Waxman-Smits correction, while chlorite-dominated formations may respond adequately to a simple linear Vsh correction in the saturation equation because chlorite has a relatively low CEC.
In carbonate reservoirs, the apparent matrix crossplot is the primary non-core method for quantifying dolomitisation and identifying secondary porosity. The replacement of calcite by dolomite increases grain density from 2.710 to 2.870 g/cc and reduces the sonic travel time from 47.5 to 43.5 us/ft; the MID crossplot tracks this systematically as a data trend from the calcite to the dolomite endpoint, and the position along this trend gives the dolomite volume fraction (V_dol). Secondary vugs in dolomite reservoirs respond differently to density and neutron than the primary intergranular porosity: vugs appear on the neutron-density crossplot as data points below the primary matrix trend (the vug porosity adds to total porosity but contributes less to the neutron than to the density response), allowing the interpreter to estimate the ratio of vug porosity to total porosity and apply the appropriate acoustic-to-total porosity transform for reservoir simulation inputs. In Devonian Leduc reef carbonates at Redwater, Alberta, secondary vug porosity constitutes 20 to 40 percent of total porosity in the core-calibrated dolomite facies, and apparent matrix crossplot analysis was the tool that first identified this from logs before the full core programme was available.