First EAGE Passive Seismic Workshop - Exploration and Monitoring Applications

We look at the benefits of placing permanent seismic sensors in a wellbore and recording data over time, either passively or with dedicated surface sources. Implementation involves placing an array of sensors permanently in a well along with the completion string. Challenges include recording data over fluid-flow noise in the well (production or injection) and reliability of the sensors and associated hardware. Applications include 4D imaging and monitoring of reservoir properties, both stand-alone and in conjunction with surface seismic, and mapping of faults and other features using micro-seismic events. Recently, a fibre-optic based system has been created for permanent emplacement in wellbores. This system has been tested with three field trials including a 2006 production well installation. Surveys have been conducted with simultaneous acquisition of permanent seabed and borehole arrays using both active surface sources and passive noise.

The authors discuss more than twenty years experience using passive microseismic monitoring techniques to monitor the behaviour of oil and gas reservoirs, underground openings (both mining and structural), and natural/artificial slopes. In the majority of these instances logistical and cost constraints have necessitated the dispersal of groups of sensors and their associated acquisition units. Under some circumstances this can give rise to complex problems when the data has to be recombined prior to processing. However, these problems are relatively minor when weighed against the cumulative benefits which accrue to the operations as a result of the utilization of high quality microseismic data acquired from an extensive sensor array operated for an extended period.

Worldwide one has observed narrow-band, low-frequency (1.5-4 Hz) tremor signals on the surface over hydrocarbon reservoirs (oil, gas and water multiphase fluid systems in porous media). These ‘hydrocarbon tremors’ possess remarkably similar spectral and signal structure characteristics, pointing to a common source mechanism, even though the depth (some hundreds to several thousands of meters), specific fluid content (oil, gas, gas condensate of different compositions and combinations) and reservoir rock type (such as sandstone, carbonates, etc.) for each of those sites are quite different. However, the physical mechanisms underlying these observations are presently not fully understood. Therefore we propose a scientific strategy for a better understanding of those phenomena. Using well-known rock physical relationships we have identified on macro-, meso- and microscale different mechanism which can induce anomalies in the low-frequency band. Using different numerical approaches we are able to compare these mechanisms with observations in the field.

The use of seismicity for understanding the dynamics of fracturing during changes in the rock stress state has been ongoing for over 20 years in the mining industry. In particular for the last 15 years, seismicity data has been used in coal mining as a hazard or rock fall indicator during underground mining. This paper highlights the first trial of the use of seismicity in an underground coal mine in Australia and the first indicator that dynamic seismicity data can be used to track the onset of fracturing which may then be used to determine the rock stress state during dynamic loading and unloading conditions in an underground long-wall mine. The results of this case history assists the understanding of how seismicity can be used during production.

Modifications of the seismic background noise spectrum were observed above hydrocarbon reservoirs (‘hydrocarbon tremors’). Other effects, such as for example the presence of soft soil layers, can also amplify the background noise at certain frequencies. The effect of soft soil amplification is widely used to derive resonance frequencies of the site as well as the S-wave velocity and thickness of the soil layer. One of the most popular techniques to determine those values is the so called Nakamura’s H/V spectral ratio technique. We observe soil related modifications in our measured data and we are able to reproduce those effects with a 2D finite difference wave propagation model. The synthetic results match very well the analytical predictions. Such simulations help to clean measured data from soft soil related effects and thus to improve the identification of hydrocarbon related information in the microtremor spectra.

ing place within the reservoir as steam is introduced. The introduction of fluid and varying temperatures during cyclical steaming cycles have direct influence on the pre-existing physical properties of the rock and as a result, the velocity model in use needs to be calibrated accordingly. Since passive source waveforms are influenced by the dynamic nature of the reservoir properties it is sufficient to say that their associated event locations need to account for these changes. To this end, we employ a Particle Swarm Optimization (PSO) approach to invert for both the velocity model and the spatial locations of the events. Based on this approach, we observed P-wave velocity reductions up to 16% at the reservoir level for any given cycle. The re-located microseismic events appear ‘well behaved’ in the context of reservoir processes, and the location uncertainty of the re-located events dropped by approximately 21%. The approach further opens opportunities to examine the anisotropic conditions in the reservoir and assist in re-defining velocity derived lithologic boundaries.

No summary available

While drilling a deep target Saga Petroleum had an uncontrolled situation were gas from the target were injected into sandlayers above through a leak in the casing. Over a period of approximately two weeks seismic data were registered from a small array of geophones located on the sea-bottom close to the platform.
Several periods of high underground seismic activity were identified. These are most likely induced small earthquakes. With such a sea bottom seismic it is possible to determine the direction to the underground acoustic source, but one has only limited control on the distance from the receivers, and thereby also limited control on the source depth. It is, however, possible to discriminate between activity close to the surface and genuine underground seismic events. The observed events are not located to the well path. Almost all of the major events taking place in the recording period are located west of a line going South-southeast to North-¬northwest. There is possibly a tendency for the events to move outward with time. The events could basically be explained as being caused by the moving front of hydrocarbons into shallow sandlayers. Systematically repeated phenomena not hitherto understood are, however, also observed.

A microseismic feasibility study is presented to monitor the injection of gas in the Arab-D reservoir of an offshore Middle East field. Full waveform modeling and detection sensitivity analysis have shown that the proposed network of installing sensors at reservoir depth in observation wells has high detection capabilities. However the modeling results reveal that interferences would exist between direct and refracted/reflected waves making it difficult to identify first P-wave and S-wave arrivals from microseismic events. Large location uncertainties are also expected with the proposed network. As a result the feasibility study recommends placing in each observation well some additional shallower sensors ~400 meters above the reservoir. Based on these feasibility results a decision will be made whether to proceed with a field trial survey.

Rural and remote passive seismic monitoring sites are often challenged by the lack of power and communication utilities. Sites without power lines require costly on-site generation (from diesel, gasoline, propane) or self-sustaining power sources (solar, wind). In all cases, electrical power is a key component in the capital and operating cost of surface-based monitoring systems.

An instrument was designed for environmental and oilfield passive seismic monitoring with very low power consumption. The device was designed to use either a wired or wireless network link for data communication. Designed for sample rates up to 4 KHz, the instrument is ideally suited for connection to seismic-band sensors such as geophones or accelerometers. A low-noise, high-gain, programmable gain preamplifier provides the necessary amplification for the A/D converter.

Several of these instruments have been installed at two remote sites using wireless data transmission. Running on solar power alone, the instruments have operated very successfully over the past two years. Reliability and the availability of remote diagnostic facilities are found to be key components in a successful remote passive seismic monitoring program.