Difference Map

Key Takeaways

• 4D seismic repeatability is the critical quality requirement for meaningful difference maps, because any seismic amplitude difference between the baseline and monitor surveys that is caused by acquisition geometry differences, processing inconsistencies, or non-repeatable noise (rather than genuine reservoir changes) will appear in the difference map as false anomalies that cannot be distinguished from real fluid or pressure changes without additional interpretation effort: the normalized root mean square (NRMS) difference calculated in areas known to have no reservoir changes (for example, in the overburden above the reservoir or in tight seal formations) provides a measure of the acquisition and processing repeatability, with NRMS values below 20 to 30 percent typically considered acceptable for 4D seismic interpretation and values above 40 to 50 percent indicating poor repeatability that would make the difference map unreliable for quantitative reservoir monitoring; improving 4D repeatability requires using the same acquisition parameters (cable geometry, source parameters, tow depth) for the monitor survey as for the baseline, processing both surveys with identical workflows, and controlling for seasonal and oceanographic noise sources that differ between surveys acquired in different seasons or sea states; permanent ocean bottom cable (OBC) or ocean bottom node (OBN) systems that leave the receivers in place on the seafloor between baseline and monitor acquisitions eliminate the receiver positioning repeatability problem entirely, producing 4D difference maps with NRMS values of 5 to 15 percent that approach the theoretical noise floor of seismic acquisition.
• Fluid substitution effects on seismic amplitude and the magnitude of the 4D signal in difference maps depend on the acoustic properties of the reservoir fluids being displaced and the Biot-Gassmann relationship between fluid bulk modulus, rock frame compressibility, and total rock acoustic impedance: replacing brine with gas in a sand reservoir produces a large decrease in acoustic impedance (gas has much lower bulk modulus than brine, reducing the P-wave velocity and density of the gas-saturated rock) that creates a strong negative amplitude anomaly in the difference map at the gas-swept zone relative to the brine-saturated baseline; replacing oil with brine (as water injection sweeps the oil from a sand reservoir) produces a smaller and sign-dependent amplitude change that depends on the density and bulk modulus difference between the formation oil and the injection brine, with higher-gravity oils (closer to brine density) producing smaller 4D signals that are more difficult to detect above the noise floor; replacing oil with a lighter oil (from gas injection or pressure decline releasing dissolved gas) can produce either positive or negative amplitude changes depending on the relative acoustic properties; the magnitude of the expected 4D signal for a specific reservoir is estimated from Gassmann fluid substitution modeling using the reservoir's rock and fluid properties, which determines whether the expected signal is detectable above the acquisition noise level before the time-lapse survey program is committed.
• Pressure effects on 4D difference maps create a separate class of seismic anomaly from fluid substitution effects, because changes in reservoir pressure alter the effective stress on the reservoir rock and thereby change the frame elastic moduli independently of fluid content: pressure depletion (decrease in pore pressure) increases the effective stress on the rock frame, stiffening the rock and increasing P-wave velocity, which produces a positive amplitude anomaly in the difference map even if the fluid content has not changed; pressure increase from injection (water or gas injection above the initial reservoir pressure) decreases the effective stress, softening the rock and potentially decreasing P-wave velocity; in chalk reservoirs (which have very large rock compressibility compared to most siliciclastics), the pressure effect on seismic amplitude can be much larger than the fluid substitution effect, making the 4D signal a more reliable indicator of pressure changes than of fluid movement; separating the pressure effect from the fluid substitution effect in a difference map requires either Gassmann-based inversion of the 4D data (which requires accurate rock physics calibration) or using P-wave and S-wave 4D data simultaneously (since S-wave velocity is relatively insensitive to fluid changes but sensitive to pressure changes), with the P-to-S ratio change helping to isolate the fluid effect from the pressure effect in areas where both have changed.
• Difference map interpretation workflow in reservoir management integrates the seismic 4D result with production data, pressure data, and reservoir simulation models to determine the most consistent geological explanation for the observed amplitude changes and to update the reservoir model for improved production forecasting: the 4D interpreter overlays the difference map on the reservoir simulation model's predicted fluid saturation changes between the baseline and monitor survey dates, comparing where the simulation predicts fluid movement with where the 4D difference map shows amplitude anomalies; agreement between the simulation prediction and the 4D observation validates the simulation model's representation of reservoir connectivity and sweep efficiency, while discrepancies between them (areas where the 4D shows changes the simulation did not predict, or where the simulation predicted changes that the 4D does not show) identify errors in the simulation model that should be corrected by adjusting the permeability distribution, fault transmissibility, or connectivity assumptions; the iterative process of updating the simulation model to match the 4D difference map observations (history matching to 4D data) produces a more geologically realistic reservoir model that makes more accurate production forecasts, with the improvement in forecast accuracy from 4D-constrained history matching demonstrably reducing the uncertainty in remaining reserves estimates and optimal well placement decisions.
• Difference map applications beyond waterflooding include monitoring of steam-assisted gravity drainage (SAGD) and cyclic steam stimulation (CSS) in heavy oil reservoirs, where steam injection creates large temperature changes that dramatically alter the heavy oil viscosity and acoustic properties, producing strong 4D seismic signals that delineate the steam chamber extent and guide decisions about where to drill new producer wells to drain the thermally mobilized oil most efficiently: the steam chamber in a SAGD operation has dramatically lower acoustic impedance than the surrounding cold bitumen (steam and hot oil are much less dense and have lower bulk modulus than cold bitumen), creating strong negative amplitude anomalies in 4D difference maps that outline the heated zone with clarity not achievable by any other remote sensing method; time-lapse difference maps acquired every 6 to 18 months during a SAGD operation track the upward and lateral growth of the steam chamber over the producing pair's lifetime, identifying stagnant areas where the chamber has not grown as expected (indicating barriers to steam migration such as shale baffles or tight heterogeneities) and areas of steam breakthrough to surface or to adjacent well pairs; the SAGD 4D program at fields including Christina Lake, Foster Creek, and Surmont in the Alberta oil sands has demonstrated significant value in optimizing steam injection rates, infill well placement, and producer well management by providing direct imaging of the steam chamber geometry that cannot be obtained from the production data alone.