![]()
Summary Commercially-applied algorithms commonly employ the first-arriving P and S waves to locate microseismic events that are recorded by borehole geophone arrays. Depending on the velocity structure, the first-arriving wave can be a direct wave, which arrives at the sensors directly from the source, or a headwave which is created by a high velocity layer above or below the sensor array. By using only the headwave signal throughout the whole sensor array, the depth accuracy of the event will decrease significantly, as it reduces the apparent aperture of the sensor array. In areas where both direct waves and headwaves are recorded as first arrivals on the sensor array, the headwaves increase the apparent aperture and improve the depth accuracy. When only direct waves are recorded the depth accuracy is relatively stable over the whole monitoring range. In addition to the headwaves and direct waves, the recorded seismograms contain additional phases, e.g. reflections and conversions. Incorporating these phases into the localization algorithm has the potential to increase the accuracy of the event location if the underlying velocity structure is thoroughly understood.
Introduction Microseismic mapping of hydraulic fracture treatments is widely used to determine the geometry of the created fracture network. But limitations on the available sensor distribution, usually a single downhole toolstring, decrease the location accuracy if only the first-arriving P- and Swaves are used in the localization. This is especially true if high velocity layers in the area create headwaves that arrive at the sensor array at very similar angles. The localization algorithm used in this study is grid-search based. After the P- and S-arrivals are identified, their traveltimes are compared with the theoretical arrivals on a grid. The grid point with the smallest residual, i.e. the best fit between observed and modeled arrival times, is considered the most likely location. Assessing location accuracy To assess the effectiveness of the different localization methods, it is necessary to quantify the location accuracy. This area of microseismic monitoring has received some heightened interest (e.g. Zimmer et al. (2009), Maxwell (2009), Eisner et al.(2009), Kidney et al.(2010), Grechka (2010)) since the quantitative interpretation of microseismic event locations requires at least an estimate of their individual location uncertainties. The distribution of the traveltime residuals can be used to visualize the location-accuracy for different signal qualities. The distribution map of traveltime residuals shows how the event location would move if the traveltime picks were varied by a certain amount. In Figure 1, 1ms residual contours are used to visualize the error space. For good quality data, the accuracy of the traveltime can be considered higher than 1 ms; this degree of error is indicated by the innermost (dark blue) contour. For more marginal data, the residuals would be larger and the error space would be more accurately illustrated by the larger green contours. The highest uncertainty is introduced when using only headwaves, as shown in location 3. The difference between these angles can be described as the effective aperture.
Site News