Reservoir surveillance during production is a key to meeting goals of reduced operating costs and maximized recovery. Differences between actual and predicted performance are typically used to update the reservoir's geological model and to revise the depletion strategy.
The changes in reservoir fluid saturation, pressure and temperature that occur during production also induce changes in the reservoir acoustic properties of rocks that under favorable conditions may be detected by seismic methods.
The key to seismic reservoir surveillance is the concept of differential imaging using time-lapse, or 4-D measurements.
Time-lapse seismic methods are usually based on differences in seismic images that minimize lithologic variations and emphasize production effects. The concept is illustrated in figure 1, where a base 3-D survey acquired before production is compared with a monitor 3-D survey acquired at a later time, dependent on the recovery process to be monitored.
The difference between the seismic surveys can then be interpreted in terms of the production-related changes in reservoir properties.
Time-lapse seismic data have been shown to increase reserves and recovery by:
- Locating bypassed and undrained reserves.
- Optimizing infill well locations and flood patterns.
- Improving reservoir characterization — identifying reservoir compartmentalization and permeability pathways.
Four-D also can decrease operating costs by:
- Reducing initial development well counts.
- Optimizing phased developments using early field-wide surveillance data.
- Reducing reservoir model uncertainty.
- Reducing dry holes and targeting optimal completions.
As a result of these benefits, many oil companies are aggressively pursuing the application of time-lapse seismic data.
The Physical Basis
Seismic velocity and density changes in a producing reservoir depend on rock type, fluid properties, and the depletion mechanism. Time-lapse seismic responses may be caused by:
- Changes in reservoir saturation. Displacement of oil by gas cap expansion, gas injection or gas exsolution resulting from pressure decline below bubble point; these decrease velocity and density. Water sweep of oil increases velocity and density.
- Pore fluid pressure changes during fluid injection or depletion. Injection will increase fluid pressure, decreasing the effective stress acting on the rock frame and lowering seismic velocities. Compaction during depletion reduces porosity and increases velocity and density.
- Temperature changes. An increase in temperature increases fluid compressibility, and as a result decreases reservoir seismic velocities and density.
Reservoir factors that affect the seismic detectability of production changes can be evaluated in order to determine which geological settings and production processes are most suited for reservoir monitoring. Each field is unique, and modeling of the seismic response to production, based on reservoir flow simulation, is used to evaluate the interpretability of seismic differences and to determine how early in field life a time-lapse survey can be used to monitor reservoir changes.
The optimal times for repeat seismic surveys depend on detectability and the field's development and depletion plan. Planning for repeat surveys in the context of field surveillance will maximize the value of the data.
The difference between two seismic surveys is not only sensitive to changes in reservoir rock properties, but also to differences in acquisition and processing.
As suggested in figure 2, the chance of success for a 4-D project depends on both detectability and seismic repeatability. Some of the factors that affect repeatability include:
- Acquisition geometry differences such as sail line orientation and heading, source-receiver spacing, streamer feather, and coverage due to obstructions.
- Near surface conditions resulting in statics and receiver coupling variations.
- Sea level, sea state and swell noise, water temperature and salinity.
- Residual multiple energy.
- Ambient and shot-generated noise.
- Geological factors such as shallow gas and steep geological dip.
4-D Seismic Acquisition, Processing
The objective of 4-D seismic acquisition and processing is to minimize differences in the seismic data that are unrelated to production — and to preserve and resolve those differences in the reservoir that are due to production.
Four-D repeat survey acquisition attempts to match both the source and receiver positions and signatures of the baseline survey. Positional repeatability ensures the same raypaths for base and monitor surveys. Tolerance to geometry deviations depends on the complexity of the overburden; where there is rapid lateral change or anisotropy in the overburden, raypaths need to be more similar.
A number of strategies have been developed to maximize acquisition repeatability for both land and marine data. And permanent monitoring systems — such as the BP's installation at Valhall — can result in high repeatability. While there is a large up-front cost associated with fixed receivers, these systems can permit the acquisition of lower-cost monitor surveys with short repeat intervals or "on demand."
Four-D processing is best described as co-processing or parallel processing of base and monitor surveys. This implies:
- Controlled amplitude and phase.
- Early equalization of geometry to facilitate QC comparisons.
- Application of the same algorithms and parameters where appropriate.
A key to successful time-lapse processing is continual comparison of the base and monitor surveys to ensure repeatability is not being compromised. Often, "fast track" data (e.g. decimated, post-stack migrated and/or using parameters based on earlier processing) are used to evaluate the processing flow and refine interpretation concepts.
And the objective to maximize repeatability may be at the expense of other processing objectives, such as high-resolution imaging. As a result, it is not uncommon that separate flows are used for time-lapse data.
The interpretation of time-lapse seismic differences in terms of reservoir changes requires integration of the data with detailed reservoir characterization, fluid flow simulation, petrophysics and conventional reservoir surveillance data.
Many companies use a model-based 4-D interpretation workflow, where seismic differences are compared to predicted differences based on seismic modeling of history-matched reservoir flow simulations. The interpretation process is one of comparing, contrasting, reconciling and validating these two images of the production process.
This approach is used because 4-D seismic interpretations are non-unique.
- A lack of change between the baseline and monitor seismic surveys can be interpreted as unswept reservoir or as an area of no reservoir.
- Four-D measurements taken once every few years can be aliased in time. Rapid changes in saturations and pressures found in some recovery processes can require more rapid seismic repeat intervals.
An example of 4-D interpretation is from the North Sea Jotun Field, where oil is being depleted through a strong natural water drive. Water sweep in the reservoir results in a 10-12 percent increase in the seismic impedance.
Figure 3 compares the results of inverting the seismic difference acquired after three years of production to obtain impedance change with the oil saturation change predicted by the reservoir flow simulation. At this location, the simulator suggests that the reservoir is fully swept — but the seismic data show that only one reservoir zone has been swept and that internal shales act as barriers or baffles to flow. This results in a flank rather than bottom water drive.
Infill or sidetrack opportunities are found where there is no change in the seismic data, and where reservoir characterization suggests there is high net-to-gross sand. As a result of the 4-D survey at Jotun, three successful infill wells were drilled and a potential dry hole was avoided.
Other published 4-D case studies show that seismic data can image production changes in a variety of geological settings and production scenarios, including water and gas sweep, pressure changes and compaction, and enhanced recovery. And 4-D interpretation is evolving toward a more quantitative analysis of the data. By incorporating time-lapse shear wave information, either from AVO analysis, elastic inversion or PS data, it is possible to estimate saturation and pressure changes in the reservoir. These estimates can be a strong history match constraint on reservoir simulations.
More predictive simulations will result in more efficient reservoir management.