Stimulated reservoir volume (SRV) is a critical factor in resource plays and is used by engineers to assess the effectiveness of a fracture treatment, establish parameters for future frac stages, and compute decline curves for quantifying bookable reserves. However, estimates derived from widely used imaging approaches come with significant limitations and uncertainties and often result in SRV being overestimated. Another issue is that the same data analyzed by different processors can generate widely differing results, which leads some engineers to question the value and integrity of the information microseismic analysis is delivering.

Current weaknesses

To improve results, it is important to understand some of the critical weaknesses of current techniques used to process microseismic data.

Frac event location methods that assume a frac event creates a single arrival will routinely produce false events. This assumption can lead to major errors in calculating the volume of interest. In one instance, a single microseismic event generated as many as five arrivals due to multipathing. Projecting these arrivals to a subsurface location results in a “comet tail” of false events extending away from the one true event. The other four events imaged are artifacts of the processing. Even though scattered event clouds are now more likely to be delivered to users with error bars showing their likely imprecision, mislocations are only one source of error contributing to flawed SRV estimates. False positives are equally at fault.

Many processors rely on acoustic processing with single-component data. This can introduce a phantom location for every real event. In such cases, engineers interpreting frac event images derived from compressional (P)-wave-only processing need to keep in mind that events they are seeing nearer the array can be mishandled shear (S) arrivals.

Incorrectly located or spurious events can lead to costly operating decisions. A frac engineer who sees falsely positioned events developing in the overburden could curtail a treatment to avoid a breach, when in reality the frac has stayed well inside the target zone. This type of misguided decision could reduce the productivity of the stage or the entire stimulation program.

Full wavefield analysis

A superior approach to delivering more reliable information for optimizing production is full elastic wave-equation imaging of microseisimic data. This should be combined with instruments that sample the whole wave-field from both surface and borehole locations. When analysts acquire higher quality microseismic recordings and use an algorithm based on the most complete physics, they can avoid imposing assumptions on the data and minimize the need for operator interaction.

The first element of this solution is high-sensitivity three-component broadband surface and borehole receiver nodes, with instrument self-noise below ambient earth noise. Spectraseis deploys these instruments as part of its Ultra Sense broadband microseismic surface and borehole arrays, with a flat frequency response from 0.1 Hz up to 1,000 Hz delivering broadband datasets with improvements in event detection and usable bandwidth compared to conventional tools. Highly flexible acquisition geometries are achieved by nodal surveys free of the physical constraints of cabled arrays.

By adding a strong low-frequency response and greater sensitivity, these arrays are lowering the industry’s detection threshold for frac events recorded from both surface and borehole locations. The advantage to the operator is a flexible acquisition strategy combining the most advanced instruments, integrating real-time results as needed.

The multiples problem: A single perforation shot from a surface array, raw (left), is flattened on the first arrival (right). Intrabed multiples labeled 1-5 have high signal-to-noise and faster apparent velocity, generating four false events in the volume. Elastic wave equation imaging eliminates this problem. (Images courtesy of Spectraseis)

Modeling

Another important element is forward modeling to design the optimal acquisition geometry prior to field work to incorporate the objectives of the program into the survey design. The resulting solution could include arrays with any combination of surface, near-surface, and borehole deployments designed to optimize the frac imaging quality at a specific target location.

It should be no surprise that a one-size-fits-all approach to frac monitoring fails to meet expectations most of the time. Instead, the main aim in developing a microseismic survey geometry should be to get the operator the best possible dataset within its budgetary parameters while ensuring that the program goals will be met. Forward-modeling the survey geometry is an essential step given the wide variation in geological and environmental settings around unconventional plays. New methodology

While the array design and equipment are crucial to improving the reliability of microseismic data, further advancements are likely through the introduction of the first elastic wave-equation imaging algorithm for frac events. The Time Reverse Imaging (TRI) method, the subject of 12 issued and pending US patents, addresses pitfalls by eliminating the assumptions and simplifications of ray-based techniques.

A high-sensitivity three-component broadband sensor has a frequency response from 0.1-1,000 Hz. Spectraseis deploys these instruments as synchronized nodes, eliminating cable constraints and allowing highly flexible survey design.

TRI removes the need for intensive operator input, reducing processor bias and improving efficiency. The TRI processing flow provides robust, data-driven mitigation of acquisition artifacts, noise contamination, and false positives. In this way, it provides not only locations but meaningful amplitudes as well. Using transparent statistics, the workflow also provides statistical confidence that events are not false positives.

The radiation of P and S energy from a frac is imaged with the elastic wave equation on the left. The frac plane and radiation direction need only minor interpretation to arrive at the seismic moment tensor diagram depicted on the right.

Beyond providing accurate event locations, the real power of imaging the seismic frac radiation pattern is the ability to fully characterize the event failure mechanism. The radiation of P and S energy from the frac encodes the azimuth and dip of the frac into the wavefield. The moment tensor is directly retrieved from these images to measure the subsurface stress regime. Drilling direction and frac connectivity can be understood with this information.

To handle the compute-heavy wave-equation imaging, Spectraseis has developed software to run on graphical processing unit clusters that use thousands of cores in parallel. The runtime for a typical TRI volume has been reduced from weeks to hours as a result. Elastic propagation and imaging codes are running near theoretical throughput maximums. Having reached hardware limitations, the company is now working on domain-decomposition strategies that promise extremely high processing speeds for microseismic data, opening the way to TRI imaging in real time.

Elastic processing is the natural domain for handling multimode data generated by fracturing rocks. It is a natural evolution for microseismic to capitalize on the power of migration methods. The result is a step-change in the potential value of microseismic frac data to the working engineer.

By delivering richer, more accurate, and more robust imaging, this technology adds confidence to SRV calculations and reliable frac characterization results for more efficient and profitable resource development.

Comments