Thursday, March 29, 2018

Seismic Surveillance for Monitoring Reservoir Changes

Time-lapse images and acoustic listening devices are the latest tools of the trade for tracking reservoir fluid movement. These techniques, done with both active and passive seismic surveys, are presenting new ways of getting more from the reservoir. For years operators have developed sophisticated models of how fluid moves in reservoirs. However, checking and calibrating them against field measurements often meets with limited success. Sample points may be far apart or data hard to interpret. Now seismic surveillance offers high-quality data that might dramatically improve reservoir management. 

Seismic surveillance is a way to "watch" and "listen" to movement of reservoir fluids far from borehole. Knowing how fluid distribution changes over time is important for more effective management decisions. For example, tracking fluid contacts during production can confirm or invalidate flow models and thereby allow the producer to change recovery schedules. Mapping steamflood fronts during enhanced oil recovery (EOR) can point to bypassed zones that may become targets of remedial operations. Mapping hydraulic fractures reveals the local stress field, which can govern permeability anisotropy - vital input for well placement, stimulation treatment and waste disposal. Seismic monitoring sheds light on each of these problems by taking advantage of some traditional and some not-so-traditional techniques.

The methods for watching and listening to movement of reservoir fluids depend on the rate at which the movement takes place. For changes over months or years, as in the case of a moving gas-oil contact (GOC), the method of choice is time-lapse seismic, sometimes called four-dimensional (4D), differential, repeated, seismic. Images from traditional seismic surveys taken before and after production are compared, and the difference attributed to moved fluids. 

For changes taking place over microseconds to minutes -as when fractures are induced or when fluid flows through natural fractures -the technique is to use borehole sensors to localize the cracking noise produced by fluid movement. These sensors record signals from events similiar to tiny earthquakes, and much of the data analysis is borrowed from earthquake technology. 

As reservoirs are exploited, pore fluid undergoes changes in temperature, pressure and composition. For example, enhanced recovery processes such as steam injection increase temperature. Production of any fluid typically lowers fluid pressure. Gas injection and waterflooding mainly change reservoir composition. These fluid changes affect the bulk density and seismic velocity of reservoir layers. Changes in velocity and density combine to affect the amplitude and travel times of reflected waves. Most surface seismic monitoring is based on amplitude changes rather than travel time. The key is to have amplitude changes big enough to be seen between the baseline survey and subsequent surveys.

The expected change in seismic amplitude can be estimated from laboratory data. Most of the amplitude change comes from fluid effects on rock velocity rather than on rock density. Laboratory experiments on fluid-filled rock show that the greatest changes in velocity arise from two different phenomena - introduction of gas into a liquid-filled rock or an increase in temperature of a hydrocarbon-filled rock. Both cause a decrease in seismic velocity. Even a small amount of gas decreases velocity dramatically by making the fluid compressible. At higher temperatures, hydrocarbons become less viscous, reducing overall rigidity. Both effects are prominent at low pressures, such as in shallow, unconsolidated sands.

Monitoring EOR processes such as steamflooding and in-situ combustion using 4D seismic was first performed in the 1980s. The progression of temperature and gas fronts away from injection wells was mapped using seismic amplitude difference sections- a seismic section created by subtracting the baseline survey image from the monitor survey image. Cores taken after treatment provided independent verification on the extent of flooding - the steam had caused additional hardening and cementation.

These studies established the basic technique used today. The baseline and subsequent surveys must be acquired and processed in the same way. However, repeating a survey with the same sources, receivers and positioning is nearly impossible. Additional processing, therefore, is used to correct for these and other undesirable diffirences in the two surveys caused by weather, access or surface structures. To intepret amplitude differences in terms of fluid movements requires a calibration, obtained by matching reflection amplitudes outside the reservoir. Interpreting amplitude changes in terms of reservoir fluid changes is then a visual, qualitative step.

A more quantitative way of using 4D seismic to monitor fluid movement is to include sythetic seismic sections and reservoir fluid flow simulations in the calibration and interpretation. Norsk Hydro is using this technique to track a gas-oil contact during gas injection. Two 3D surveys were acquired 16 months apart in the Oseberg field, Norwegian North Sea. Special care was taken to preserve true amplitudes during processing. The base survey served several purposes : it provided structure for the initial reservoir model and input for further drilling and development decisions as well as acoustic properties of lithologies and fluids. 

Injecting gas into the Oseberg field pushes the GOC deeper and farther from the injection well. This displacement agrees with that shown on surface seismic and with the simulated reservoir fluid flow for the period between the seismic survey. 

Borehole seismic images from the field support the interpretation. Time-lapse verticaal seismic profiles (VSPs) from the same well also show downward movement of GOC.
The sequence of seismic data interpretation, modeling and comparison with reservoir fluid flow models forms a loop, taking advantage of feedback at many stages. Using these steps, Norsk Hydro geophysicst have been able to validate the existing reservoir model and develop a methodology for future monitoring.

 A monitoring technique that is just emerging is comparison of amplitude variation wtih offset (AVO) in repeated surveys. AVO analysis has been shown to be a powerful fluid discriminator. Scientist at Elf Aquintatine have mad some progress in testing this technique for monitoring waterfront movement in the Frigg field, Norway. The Frigg has producing wells at the crest of the structure, but little other than seismic data to describe the rest of the reservoir. Accurate estimates of gas reserves are crucial so that delivery contracts can be honored. 

Real-Time Monitoring

Real-time monitoring is a different kind of seismic surveiilance that goes by three names: microseismic monitoring, acoustic emission monitoring, or passive seismics. Pressure jumps caused by fluid injection, depletion or temperature change are rapidly transmitted to surrounding rocks, causing stress changes. Stresses are released by movement at fractures or zones of weakness through microseismic events - tiny earthquakes. Typically, microseismic events have very small magnitudes , from -6 to -2 on the Richter scale, a logarithmic measure of energy released in a seismic event. Most microseismic events have a million times less energy than the smallest earthquake that can be felt by a human at the Earth's surface, which is +2 to +3. However, some reservoirs have a history of felt seismicity induced by human intervention.

The transient nature of pressure changes requires a different approach to monitoring. Sensor are deployed in boreholes, adapting technology first developed to monitor earthquakes, then for fractures created by hydraulic stimulation of hot, dry crystalline rocks for extraction of geothermal energy. In the case of hydraulic fracturing for wellbore stimulation and waste injection, passive seismics can provide information about orientation and now height and length of induced fractures, depicting fracture containment. With orientation information, well locations can be optimized to take advantage of permeability anisotropy associated with fractures, and waste containment can be assured for regulatory purposes.

 How does microseimic monitoring work? Although the mechanism of hydraulic fracturing is usually understood to be tensile failure , many microseismic experts believe that recorded events are caused by shear failure along fractures. Energy radiates as a compressional (P) wave traveling at the P-wave speed followed by a shear (S)wave traveling at the slower S-wave speed. A conveniently located triaxial borehole receiver records signals that may be analyzed to locate the source of the emissions.

Location is determined by distance and direction from the receiver. Distance to the event can be obtained by knowing the velocites of P waves and S waves and the lag between their arrivals. Direction is known from polarization or particle motion of the P wave, which is along the path connecting the event and the receiver. 

Most microseismic monitoring experiments are conducted in dedicated observation wells near an injection well and away from noisy pumps.



No comments:

Post a Comment