Completion designs for hydraulic stimulation of shale-gas reservoirs frequently accounts for vertical growth of the treatment volume in the formation. Where vertical growth is expected, wells are drilled near the base of the reservoir optimizing the distribution of proppant upwards. Other treatments may seek to transport treatment fluid across a lithologic barrier, effectively trying to "treat two formations for the price of one.?? Vertical growth needs to occur under controlled conditions, undesirable growth leads to a potential creation of pathways for treatment fluids to leak out of formation, or worse pathways allowing undesirable fluids to flow into formation. In either case, this could lead to a loss in optimization for production. To better understand vertical growth characteristics of hydraulic treatment volumes, microseismic monitoring arrays deployed downhole just above the formation provide a good discriminant for vertical growth of events. Further characterization of this growth can be accomplished through Seismic Moment Tensor Inversion (SMTI), when a sufficient angular distribution of multiple downhole arrays detects the microseismicity. SMTI can distinguish the source type of the mechanisms (e.g. openings, closures, shear, etc.) and the orientations of the activated structures, allowing for a more complete picture of the failure process. In the example provided, different stages of stimulation in the Marcellus shale formation are examined in the context of varying degrees of vertical growth. When vertical growth occurs, as identified through SMTI analysis, it appears to be related to the activation of sub-vertical natural joints whereas for vertically confined stages the primary fracture set is sub-horizontal suggesting delamination of fissile bedding planes is the dominant process. These differences, from stages in the same completion program, suggest that subtle background stress changes can result in very different behaviors. Full understanding of these mechanisms will lead to further optimization of these treatment programs to promote vertical growth to traverse structural barriers and retain containment of the treatment within zone.
Over the past decade, microsesimic monitoring has become the approach most often used to gain an in-situ understanding of the rock's response during hydraulic fracture stimulations. From initial monitoring performed in the Barnett Shale to monitoring currently being carried out for example in the Horn River and Marcellus formations, we review the evolution of microseismic monitoring from data collection (single versus multi-well array configurations, utilization of long lateral stimulation wells), to data analysis and the incorporation of microseismic parameters to constrain and validate reservoir models. Furthermore, we discuss the variations in microseismic activity for different stimulation programs (e.g. zipper-fracs) and stimulation fluids.
Generally, we have observed that overall fracture height, width and length, orientation, and growth vary from formation to formation and within each formation, thereby highlighting the ongoing necessity for microseismic monitoring. Additionally, through the use of advanced microseismic analysis techniques, such as Seismic Moment Tensor Inversion (SMTI), details on failure mechanisms have been used to assess stimulation effectiveness and define complex Discrete Fracture Networks (DFN). This information provides estimates of Enhanced Fluid Flow (EFF), which assist in calibrating and validating reservoir models. Utilizing spatial and temporal distributions in DFN and EFF, along with estimates of fracture interconnectivity and complexity, the role of pre-existing fractures and fault structures in the rock matrix can be established and used to provide more realistic estimates of stimulation parameters such as Stimulated Reservoir Volume (SRV).
Hydraulic fracture stimulations result in the re-distribution of stress adjacent to the treatment well that in turn may result in the triggering of seismicity. Generally, seismicity only occurs when the triggering source, such as the injection of proppants, acts to drive the pre-existing stress field closer to failure levels of susceptible fractures/faults. If we consider that the observed seismicity is a function of the stress state and its interaction with the fracture network, under some conditions, it might be possible that seismicity may then occur at distances beyond the volume related to or directly influenced by the stimulation. Currently, the extent of seismic influence, including the potential for remote mobilization of pre-existing structures, the potential for out-of-zone fluid communication, and the relative stress and stress transfer conditions leading to large event occurrence (M > 0) are not well understood.
In this paper, we investigate the spatial and temporal variations in microseismicity associated with hydraulic fracture treatment of as shale play in North America during which distant larger magnitude events were triggered. These events, in thiese examples, were located up to several hundred metres away from the stimulations, and in some cases in different formations from the reservoir host rock. These datasets, including calculated failure mechanisms and source parameters such as apparent stress drop and seismic moment, provide the basis for examining the conditions under which remote structures can be activated and how the distribution of microseismicty can be used to establish the region of stimulation influence. Based on these observations we postulate on the potential of triggering events remotely.
Completion programs for hydraulic fracture stimulations are planned to optimize the spacing of wells and perforation clusters such that the largest volume of the reservoir can be accessed through the promotion of a discrete fracture network in the reservoir. Such treatments also seek to minimize costs associated with pumping proppants and fluids down wells by ensuring that these injectants reach their target formations, stay in zone, and act to promote the stimulation of the reservoir. From this viewpoint, it is seen as desirable to minimize the overlap of treatment volumes between neighbouring wells and stages to avoid the preferential diversion of proppants and fluids into the previously stimulated volumes of the reservoir. However, it has also been argued that the creation of new fractures in a previously treated volume promotes a complex fracture network enhancing drainage. When these stimulations are monitored from multiple geophone arrays surrounding the treatment zone, seismic moment tensor inversion (SMTI) analysis offers the ability to test these hypotheses by inferring if the microseismic events are related to the opening of or closure of pre-existing natural or newly created fractures. In this paper, we discuss event clusters that occur with a significant degree of overlap between neighbouring stages in the Marcellus Shale. Because the events were monitored with multi-array sensor configuration, the SMTI calculations can be conducted with a high degree of accuracy. SMTI allows for the orientations of the underlying fractures to be determined, allowing us to construct a discrete fracture. Further analysis of the orientations of the underlying fractures also enable us to assess which fractures are being activated in relation to the pre-existing structures in the reservoir, and how the activation of those structures correlates to estimates of stimulated reservoir volume which we relate to the regions of the reservoir where an complex, intersecting fracture network is being activated by the stimulation.
Nine wells were drilled to test cyclic steam stimulation as a recovery mechanism in the diatomite reservoir in the Belridge field. Microseismic monitoring was proposed to evaluate steam chest and fracture growth. A series of models were constructed to determine both microseismic event detectability and locatability. The modeling indicated that poor signal-to-noise ratios would constrain the ability to locate events using a single array. As a result, three monitoring wells were installed. The purpose of this paper is to evaluate the microseismic event location results obtained from the 3-well solution and compare them with the solutions obtained when turning off one or two of the arrays. This first phase of investigation was performed on the sand-propped hydraulic fracture stimulation prior to the cyclic steam operations. This study may be applied to other areas including imaging hydraulic fracture stimulations in shale plays, reservoir steam monitoring, or in any area where location precision in microseismic monitoring is necessary.
It has been demonstrated in the field that the monitoring range of a microseismic system can be increased and the potential for locating microseismic events improved by installing multiple arrays having overlapping radii of observation. Innovative system design, deployment techniques and operational procedures, plus advanced multi-well processing strategies have all contributed to creating a growing data set with over 4,000 events recorded during the first 13 months of operation. Results of turning off arrays showed increased location error with two well solutions and a significant increase in error with one-well solutions notwithstanding the reduction in event location count due to lack of multi-phase signals on a single array. Azimuth errors in single microseismic observation well solutions result in disperse interpreted fracture geometry and in mis-interpretation.
Overall, our observations show that, for both multi- and single-phase events, the observed detection limits and size distribution of the seisms has far exceeded those originally predicted by the earlier modeling.
Over the past decade, microsesimic monitoring has become the approach most oftenused to gain an in-situ understanding of the rock's response during hydraulic fracture stimulations. From initial monitoring performed in the Barnett Shale to monitoring currently being carried out for example in the Horn River and Marcellus formations, we review the evolution of microseismic monitoring from the viewpoint of data collection (single versus multi-well array configurations, utilization of long lateral stimulation wells), to data analysis, to the incorporation of microseismic parameters to constrain and validate reservoir models.
Generally, we have observed that overall fracture height, width and length, orientation, and growth vary from formation to formation and within each formation, thereby highlighting the ongoing necessity for microseismic monitoring. Additionally, through the use of advanced microseismic analysis techniques, such as Seismic Moment Tensor Inversion (SMTI), details on rupture mechanisms have been used to assess stimulation effectiveness, define complex Discrete Fracture Networks (DFN) and provide estimates of Enhanced Fluid Flow (EFF), which assist in calibrating and validating reservoir models. Utilizing spatial and temporal distributions in DFN and EFF, along with estimates of fracture interconnectivity and complexity, the role of pre-existing fractures and fault structures in the rock matrix can be established and used to provide more realistic estimates of stimulation parameters such as Stimulated Reservoir Volume (SRV).
Hydraulic fracture completions seek to balance spacing of treatment wells and perforation clusters in order to minimize the costs of drilling wells and pumping fluids and proppant downhole while promoting the development of a discrete fracture network to connect even the most isolated pockets of hydrocarbon. To this end, numerous strategies for well completions have been proposed, such as avoiding the overlap of treatment volumes between adjacent wells and/or stages because of the risk that proppant and fluid will preferentially be diverted into earlier treated volumes. In counterpoint, it has also been suggested that the creation of new fractures in a previously treated volume promotes a complex fracture network enhancing drainage. When these stimulations are monitored from multiple arrays surrounding the treatment zone, seismic moment tensor inversion (SMTI) offers the ability to test these hypotheses by inferring if the events represent the opening of fractures or closure of pre-existing natural or newly created fractures. In this paper, we discuss two different completion programs. One common thread between the two data sets is that observed event clusters occur with a significant degree of overlap between neighbouring stages. Both completions were monitored with optimal multi-array configurations allowing for the calculation of SMTI with a high degree of robustness. The first stages in both examples showed significant opening components of failure. Neighbouring subsequent stages show closure events in the overlapping regions suggesting that the previously opened fractures were now closing due to local re-orientations of the stress-strain field stress induced by the later injection over-printing the region of overlap. Based on these analyses, it can be suggested that the moment tensor response can be used to identify the effective spacing for perforation clusters and establish optimal stimulation programs, which could include setting fracture ports farther apart.
The inclusion of fracture networks in reservoir models is generally based on the concept of failure associated with subvertical fractures. In general, it is surmized that fractures can grow irregularly in a stress field that is perturbed by a hydraulic fracture injection. It has also been considered that structural weaknesses in the rock such as pre-existing fractures and naturally occurring laminations commonly found in shale-gas reservoirs can be conduits for fracturing during stimulation and active pathways for fluid flow. We postulate that local stress perturbations through stress transfer allows for fractures to propagate and initiate failure along pre-existing fracture sets, which include sub-vertical and sub-horizontal fractures. Additionally, the degree of fracture interconnectivity and the type of fracturing will play a role in whether effective proppant transport is achieved. Through moment tensor inversion of microseismic events related to stimulation in the Horn River Basin utilizing well-conditioned geophone arrays, we have been able to define a three dimensional discrete fracture network consisting of sub-horizontal and sub-vertical fractures. Geologic data from the site provided corroborative evidence to the validity of the observed discrete fracture network, the presence of sub-horizontal fractures and fracture orientations in-line with current regional stress field. The fracture intensity and complexity appeared to be directly related to the degree of interaction between the sub-horizontal and sub-vertical fractures. Regions dominated by sub-horizontal fractures were also regions exhibiting poor fracture intensity and complexity. Based on these observations and moment tensor derived failure modes (opening component of failure), we were able to identify regions of enhanced fluid flow, further identifying regions of effective fluid transport. Regions with poor connectivity and dominance of sub-horizontal fractures also were identified as regions of poor fluid flow; these then become regions for potential re-stimulation. Based on these analyses, it can be suggested that sub-horizontal fractures can play an important role in the overall fracture development.