Summary We discuss a method to assess if a particular seismic survey may be suitable as a base survey for a time-lapse seismic monitoring project, and to predict if anticipated changes in the reservoir's acoustic properties could be interpreted on a repeat survey. Our approach is to generate a simulated repeat survey using a realistic noise realization estimated from the candidate base survey. This repeat survey contains seismic changes that have been modeled by integrating rock and fluid property data, as well as results from the flow simulator. The method can be used to reduce the risk of shooting a repeat survey on which meaningful seismic changes cannot be interpreted. It can also help in deciding the proper timing for an eventual reshoot. Furthermore, the method is valuable for assessing data quality before quantitative interpretation studies. We show an example from the Middle East illustrating a survey that is unfavorable as a base survey. Furthermore, we show examples from the North Sea where application of the method was key for deciding upon a repeat survey and upon the proper timing of a possible repeat survey. Introduction Time-lapse seismic monitoring is an emerging technology with potentially very large commercial value. By repeat two-dimensional (2D) or three-dimensional (3D) seismic surveys, it aims to monitor seismic changes related to fluid and stress changes during the production of a field.1,2 Hence, this technology has the potential to allow field monitoring between and away from wells. In favorable circumstances, displacement of the fluid fronts may be seen.3,4 Pressure changes may also be detectable showing compartmentalization in a field.5 Improved knowledge of saturation and/or pressure distributions will result in improved dynamic reservoir models, which will help to optimize recovery.6 Sonneland et al.7 and Jenkins et al.8 describe two cases where time-lapse seismic monitoring has led to commercially significant decisions. The technique will work with a high probability of success on fields where the presence of hydrocarbons is already clearly indicated through reflection amplitude anomalies visible in the seismic data. In favorable circumstances, one may recognize waterflood, steamflood, gas cap formation, or bypassed oil. The probability of success of the technique heavily depends on many factors, like reservoir parameters (depth, rock, and fluid properties; pressure; etc.), nature of the recovery processes,9 and on business drivers.10 Repeatability of the different seismic surveys is another important factor; for instance, positioning errors or different offset distributions may result in failure if not accounted for correctly.11 A quick and quantitative approach in assessing the risk of a time-lapse seismic project has been described by Lumley et al.12 This paper focuses on another critical factor, namely, how the non-repeatable noise of the processed seismic data compares itself to the anticipated changes caused by production. The character of this noise may vary in time, as well as in space. Given the quality of the base survey, it can be assessed if seismic monitoring between wells is feasible and, furthermore, after how many years of production one may be able to pick up changes in seismic response. This information can have a high business impact. In specific cases, it may become clear that seismic monitoring is only feasible if the signal-to-noise ratio (SNR) of the seismic data can be increased via more sophisticated processing. We illustrate the application of the method using data from an onshore field in the Middle East (Yibal), and from two offshore fields in the North Sea (Brent and Draugen). Outline of the Method Approach. Our aim is to assess if a particular seismic survey may be suitable as a base survey in a time-lapse seismic monitoring project. Therefore, we need to know if expected acoustic changes (due to production from a reservoir) could be interpreted on a repeat survey, given the quality of the candidate base survey. Our approach is now to create a simulated, but realistic repeat survey consisting of three components:signal,noise, andmodeled seismic change. The signal and noise follow from the candidate base survey and are obtained as follows. Using a sparse spike inversion (SSI) method (explained below) we estimate reflectivities (spikes) that represent the reflectivity of the Earth. These reflectivities are subsequently used to determine the highest-resolution signal that can be determined from the base survey. This signal is component (i) for the simulated repeat survey. The difference between that signal and the base survey yields an estimate of the noise for that survey. Component (ii) is now obtained by changing all noise traces to other grid locations; hence, one obtains a new, realistic noise realization. Component (iii), the expected change, is now modeled as the seismic response of a 3D-wedge model representing expected changes in rock properties. Such a wedge model is useful to study for which thicknesses one could detect which contrasts in acoustic parameters. To this end, one can vary those contrasts in the lateral direction perpendicular to the direction of varying thickness. (For more details on the applications of wedge models, refer to, e.g., Refs. 13 and 14.) Alternatively, the expected change can be modeled as the difference between synthetic data sets computed from two flow-simulator runs corresponding to the acquisition years of the base survey and the anticipated repeat survey. Fig. 1 illustrates the approach described above (and visualizes a 3D-wedge model). The final step is now to determine if it is possible to interpret the seismic changes, in the presence of noise, between the candidate base survey and the simulated repeat survey. To this end, one can perform time-shift measurements between the base and simulated repeat surveys, as well as measurements on seismic attributes (like maximum amplitude, area, or width of the seismic wiggle between two consecutive zero crossings) using any proposed interpretation techniques. Detailed Description and Illustration of the Method In this section, we discuss the method in detail and we illustrate it using examples from the onshore Yibal field in Oman and the Brent and Draugen fields in the North Sea.
In many, but not all cases, time-lapse (4D) seismic data can be used with commercial advantage to monitor reservoir changes between and away from the wells. In order to avoid disappointment we need to recognise the favourable cases before starting a time-lapse seismic monitoring project. We aim to assess if a particular seismic survey may be suitable as a base survey and we discuss a 4D-feasibility method, which can be used to predict if predicted changes in the reservoir's acoustic properties could be interpreted using a repeat seismic survey. To this end, seismic, rock and fluid property data, as well as results from the flow simulator are integrated. The method can be used to reduce the risk of shooting a repeat survey that would not show what was being sought, and can also help in deciding on the proper timing for an eventual reshoot. Furthermore, the method is also valuable for assessing data quality before quantitative interpretation studies. The method is illustrated by examples from the Middle East and the North Sea. P. 307
No abstract
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.