I summarize the presentations and talks during the coffee breaks on the EAGE Fourth Passive Seismic Workshop (WS) from the 18.03.2013-20.03.2013. The main focus was on optimizing development of unconventional reservoirs and the program was separated into five sessions with following topics: acquisition, surface monitoring, processing & moment tensor inversion, induced seismicity and surface characterization & interpretation.
Micro-seismic Event Detection (MED) is fast becoming a significant tool for offshore reservoir monitoring. The E&P industry, driven by field operator requirements for real-time data on hydraulic and geomechanical processes, presently needs more information to understand dynamic reservoir changes. MED is an important tool for hydraulic fracture detection, the mapping of reservoir stresses and seismic hazard analysis which are in turn key indicators for evaluating reservoir dynamics and associated environmental risks.
This technique is based on previous observation of very small earthquakes with magnitudes down to Mw = -4 or even lower generated by the production activities in a reservoir.
Micro-seismic events are distributed at specific locations and follow the fault structures in the reservoir. These structures act as flow channels and routes for premature fluid or gas breakthrough in the reservoir, destroying the cap-rock integrity with surface subsidence as result. MED typically identifies such reactivated fault structures with significantly greater accuracy than seismic-surveys. In addition to 4D seismic surveys MED can also be used to provide real-time 3D monitoring of the change in fluid-pressure distribution. Conventional MED approaches use vertical receiver-arrays in injection or production wells facilitating event-detection ranges of approximately 1Km from the well. Studies demonstrating the feasibility of offshore surface receiver for this technique were conducted from 1997 onwards.
Micro-seismic detection relies on the detection and analysis of large numbers of micro-earthquakes using surface receiver arrays to provide detailed information about the effects of hydraulic fracture, stress induced reactivation of faults and changes in pore pressure in the reservoir.
Micro-seismic detection is based on seismology techniques; hence the receiver position is not as important as the sensor-node quantity. Like standard earthquake seismology, the resultant micro-seismic event detection is more accurate the more receivers can be used for measurement with larger distance between each receiver. Short distance receivers will not increase the accuracy of the event location. In order to locate an event it is only important that it is situated within the receiver-array’s geometric boundary. Ray-tracing can help to estimate shadow-zones (caused e.g. by faults/ fractures) and to estimate the velocity model.
This WS was very inspiring and informative with interesting discussions, good presentations and an overview what is state-of-the-art. MED is a widely accepted industrial tool and is about to become an additional tool for permanent reservoir monitoring. In order to understand the dynamic changes in the reservoir/hydraulic fracture zone more information are needed than provided during a short measurement campaign. MED is also an important factor in hydraulic fracture detection, mapping reservoir stresses and seismic hazard analysis.
There is a common understanding in the WS community that any type of fluid injection will cause microseismic fractures in a reservoir or fracture zone, to what extent is still under discussion. The wave-form inversion calculation of the focal mechanism by the moment tensor can estimate tensile openings and closures. This is an important step to understand what really happens in the reservoir or overburden. The result can “easily” be plotted into a Hudson-diagram which shows the opening and closing of fractures during fluid injections over time.
To do a proper wave-form inversion it is necessary to have accurate velocity-models, estimated rock parameters (e.g. resonance frequency, Q-factor…) and an understanding of the source mechanism. All this parameters are difficult to get, and assumptions have to be made. The focal mechanism and moment tensor inversion was presented and discussed on the WS, but from my point of view nothing new. The same methods are used as presented on the last passive seismic WS and I think the visualization of the moment tensor, like beach-balls are made for scientists rather than engineers. But the engineers have to make their decisions on the results we provide and therefore easier understandable visualization methods should be developed. Some improvements already exist like the use of Hudson-diagrams and fracture volume-plots.
However, microseismic monitoring is an engineering tool and I got the impression that we are losing the reason why we are doing all this. I am still missing the link between fracture volume estimation (if possible) and microseismic events. How can we calculate e.g. the fracture size and orientation? How can we estimate the fracture volume? And how accurate are the results? This are questions to be answered, only to name some.
This WS showed that we have microseismic detection workflows in place to determine the source location and moment tensor as soon as we are able to detect microseismic events. But in order to make microseismic events detectable, we have to add some steps which are not well discussed yet. This steps are not in the main focus, but I think they are equally important than event detection, moment tensor inversion and estimation of the focal mechanism.
The first step should be the a more or less standardized feasibility study, but this will require access to a geophysical model with high accurate velocities models as well as the Q estimation and the receiver characteristics. With some assumptions with respect to the characteristics of microseismic events, we can calculate the attenuation of the event with distance from its location. The measured event Magnitude at the receiver position can be evaluated. Furthermore the signal to noise ratio (SNR) improvement with different number of receivers used can be predicted. The result of this feasibility study will be the evaluation of the maximum achievable detection ranges for different receiver configurations. The feasibility study should also include some synthetic experiments using different microseismic sources of different sizes which are triggered at specific locations along a fracture/fault plane in the model in order to quantify detectability with size, depth and noise level of the source.
There will be not one workflow to fix all requirements, but the useful combination of preprocessing methods will be the key to improve the measured data.