- Home
- Conferences
- Conference Proceedings
- Conferences
81st EAGE Conference and Exhibition 2019 Workshop Programme
- Conference date: June 3-6, 2019
- Location: London, UK
- Published: 03 June 2019
41 - 60 of 93 results
-
-
Low Frequencies Unlock Visco-Acoustic Full Waveform Inversion Capability in Trinidad
Authors C. Theriot, T. Fox and S. BarronSummaryA large Ocean Bottom Node (OBN) survey acquired off the east coast of Trinidad has allowed Shell the opportunity to deploy another application of its visco-acoustic Full Waveform Inversion. The setting is ideal, with pervasive shallow gas covering the area resulting in dispersion and attenuation. A proper and accurate earth model including the Q anomalies in this area is more accurately determined with visco-acoustic FWI over tomographic methods given shallow water depths resulting in incoherent shallow images and migrated gathers. In addition to the Q-field, the inversion will also derive the velocity field including the resulting slow-downs from these gas zones and ultimately improve image quality. The higher quality low frequencies recorded by the OBN has resulted in a more accurate and geologically reasonable model where previous Full Waveform Inversions using streamer data have left inaccuracies. And from this more accurate model, we present a substantial image uplift.
-
-
-
Broadband FAZ Land Data: an Opportunity for FWI
Authors O. Hermant, A. Sedova, G. Royle, M. Retailleau, J. Messud, G. Lambare, S. Al Abri and M. Al JahdhamiSummaryThere are very few applications of full waveform inversion (FWI) on land data. This is commonly attributed to data-specific challenges. However, modern broadband full-azimuth (FAZ) land surveys offer an extraordinary opportunity for applying FWI. They have dense surface and offset sampling X and Y directions, and contain very low frequencies down to 1.5 Hz. We demonstrate in this study and in other real data examples from the Sultanate of Oman that it is possible to benefit from the broadband spectrum of modern land acquisitions to obtain a high resolution velocity model reliably using FWI.
-
-
-
Does Broadband Address the Cycle Skipping in Complex Areas?
By D. VighSummaryFull-waveform inversion (FWI) is a high-resolution model building technique that uses the entire seismic record content to build the earth model but have struggles with cycle skipping . Conventional FWI usually utilizes diving and refracted waves to update the low-wavenumber in other words the background components of the model; however, the update is often depth-limited due to the limited offset range acquired. To extend conventional FWI beyond the limits of the transmitted energy, we must use reflection data as well with broad band preprocessing.
-
-
-
Towards Lateral Broadband
Authors R. Soubaras and B. GratacosSummaryNew acquisition techniques and the evolution of broadband processing in the past ten years have enabled the extension of the frequency bandwidth from the conventional three octaves bandwidth [10Hz–80 Hz] to a six octaves broadband bandwidth [2.5Hz–160Hz]. Despite this impressive achievement, some problems still remain:
- The broadband processing sequence has become very complex.
- This processing sequence makes a heavy use of sparse tau-p transforms in steps like receiver deghosting and regularization. However, the underlying assumption that a shot point can be locally decomposed in a few linear events can be questionable.
- The lateral resolution has not increased in the same proportion as the vertical resolution.
In order to solve these problems, we show that we can obtain a significant increase in lateral resolution by using for the final imaging a least-squares migration with ghost and multiple modeling, allowing the deghosting, regularization and multiple attenuation being handled by the inversion. This is assessed on a real 3D dataset with depth-slices showing an increase in wavenumber bandwidth similar to the increase already obtained in frequency bandwidth.
-
-
-
Imaging Including Internal Multiples: Influence of Broadband Acquisition
By E. VerschuurSummaryNowadays there is a strong trend towards considering multiples as genuine part of the seismic response and, therefore, including this in the imaging process.
For surface multiples, this has already shown successful in various applications over the last decade.
For the correct imaging of internal multiples, there is a debate whether removing internal multiples can be more fruitful than trying to image them. In this paper we will show the added value of properly including internal multiples in the imaging stage, where the transmission effect is also being accounted for. In this way the imprint from a multiple-generating overburden is also minimized.
Finally, it will be demonstrated that acquiring data with a broadband acquisition set-up, the effect of internal multiples is already greatly reduced, which inceases the abilities to treat them properly in the imaging stage.
-
-
-
Thermal Modelling of Magmatic Geothermal Systems: the Role of Deep-Seated Heat Sources
Authors G. Gola and A. ManzellaSummaryWe present the results achieved in the framework of different research projects, i.e. the Geothermal Atlas of Southern Italy, the Image, the Descramble and the Gemex Projects, mainly focussing on the thermal aspects of four geothermal fields developed in magmatic setting. We applied an integrated method in order to set-up numerical models able to simulate the conductive-convective thermal structure of the Ischia Island (southern Italy), Long Valley Caldera (eastern California), Acoculco caldera complex (eastern Mexico) and Larderello-Travale (central Italy) geothermal systems. We propose a numerical approach implemented in a Finite Element environment capable to evaluate the contribution of the main variables that characterize the magmatic heat source and the geothermal reservoir. The final 3D thermal models were achieved via the optimization of the available temperature measurements in deep boreholes tacking into account the thermal effects of the interplay between the free convection and the topographically driven groundwater flow, the reservoir permeability and the thermal load released by the parametrized heat source. Our results contribute to better understand the relationship of magmatism to geothermal resources in continental settings.
-
-
-
Laboratory Studies of Organic and Inorganic Geothermal Tracers at Superhot and Supercritical Conditions
Authors Muller Jiri, Sissel Opsahl Viig and Helge StraySummaryLaboratory studies have been performed at testing stability of organic and inorganic tracers at super-hot and supercritical conditions. Both static and dynamic tests have been performed at specially constructed equipment which can tolerate such hard conditions. In some cases these tests indicate no rapid thermal degradation of the tested tracer candidates within the time frame of the performed stability test (2 months). In other cases the experiments indicate interactions between the rock material and the tracer candidates.
-
-
-
Seismic AVO Inversion for Geothermal Reservoir Characterisation
Authors E. Dalgaard, K. Bredesen, A. Mathiesen and N. BallingSummaryA field case is demonstrated to show how 2D seismic AVO inversion together with well log analysis can aid reservoir characterization of a geothermal play in the northern Zealand of Denmark. From the seismic inversion it is possible to interpret different lithologies and estimate porosities via links established at well logs. Several connected high porosity sands were predicted, and with an expected temperature of around 50C in the target zone this gives room for a potential good geothermal reservoir. With this specific field case it is demonstrated how seismic AVO inversion can be applied where geothermal reservoir characterisation is needed in order to obtain a better understanding of potential geothermal plays.
-
-
-
Geothermie 2020: Exploration and Development of Geothermal Energy in Geneva
More LessSummaryThe deployment of renewable energy sources for both power and heat production is accelerating in Switzerland, a trend that will continue, thanks to the 2050 Swiss Energy Strategy that aims at gradually phasing out nuclear power by reducing the energy consumption and increasing heat and electric power generation from renewable energy sources. Geothermal energy will be an important resource to supply heat and power for industrial, agricultural and domestic use.
-
-
-
Exploration of Geothermal Reservoirs: an overview and future opportunities
By P. JoussetSummaryAn overview of geophysical exploration methods for geothermal reservoir is proposed. Focus is made on the integration of seismic attributes and resistivity structures, with examples from Iceland and Mexico. Future targets include magma chambers and urban environment.
-
-
-
Learn to Invert: Surface Wave Inversion with Deep Neural Network
Authors S. Hou, S. Angio, A. Clowes, I. Mikhalev, H. Hoeber and S. HagedornSummaryWe propose a hybrid analytics and machine learning approach for large-scale surface wave inversion (SWI) for shear-wave velocities in the shallow overburden. A sparse grid of 1D velocity models are inverted using analytic optimization. Then, a deep neural network (DNN) with three hidden layers is trained using a spatially sparse subset of the data and non-linear inversion results. Finally, we use the DNN to predict the velocity model for the whole survey. This approach is demonstrated on a real high density land project. In comparison to the purely analytical approach, the hybrid analytic-ML method estimates a more reliable shear velocity model over the whole survey with significant reduction in computing time. We end with a discussion around the potential of this type of method for other geophysical inverse problems and seismic processing.
-
-
-
Neural Network Travel-Time Tomography
More LessSummaryTravel-time tomography is a non-linear inverse problem. Monte Carlo methods are increasingly used to provide probabilistic solutions to tomographic problems, but these methods are computationally expensive. Neural networks can be used to solve some non-linear problems at a much lower computational cost. We show for the first time that a form of neural network called a mixture density network can perform fully non-linear, rapid and probabilistic tomographic inversion using travel-time data. We compare two methods to estimate the Bayesian posterior probability density functions: first a vector of networks are trained such that each estimates the marginal posterior probability distribution of wave speed in one grid cell; second, a single network estimates the entire posterior probability density function across all cells. While both methods provide estimates of the true structure in the means of their distributions, their uncertainty estimates differ: when separate networks are trained to solve for wave speeds at each location in the model the standard deviations exhibit uncertainty loops, as expected, whilst a network trained to solve for speeds on the whole model at once does not. The former method is therefore likely to be more robust.
-
-
-
Including Physics in Deep Learning – An Example from 4D Seismic Pressure Saturation Inversion
Authors J.S. Dramsch, G. Corte, H. Amini, C. MacBeth and M. LüthjeSummaryGeoscience data often have to rely on strong priors in the face of uncertainty. Additionally, we often try to detect or model anomalous sparse data that can appear as an outlier in machine learning models. These are classic examples of imbalanced learning. Approaching these problems can benefit from including prior information from physics models or transforming data to a beneficial domain.
We show an example of including physical information in the architecture of a neural network as prior information. We go on to present noise injection at training time to successfully transfer the network from synthetic data to field data.
-
-
-
Stratigraphic Segmentation Using Convolutional Neural Networks
Authors D. Civitarese, D. Szwarcman and E. Vital BrazilSummaryThe Discovery of new possibles reserves is an critical activities for the oil and gas industry. The most used methods to understand the sub-surface are based on seismic surveys. however, the process of the interpretation of these surveys are very expensive and due to the volume of data it overload the human capabilities. On the other hand, deep learning techniques have been increasingly applied in several areas of science to help in tasks that were considered human-centered, such as image classification and language translation, among others. We propose a machine learning methodology to classify seismic data at the pixel level, producing an interpretation mask suggestion. Our methodology comprises three main parts: model selection, dataset preparation, and training. We also present Danet-FCN3, a deep neural network specifically designed to classify seismic images at pixel level resolution. We have recently demonstrated that our deep learning models can distinguish among different rock layers helping the expert to interpret new seismic images. The dataset preparation processes the raw post-stacked data and the interpretation labels to produce training, validation and testing sets.
-
-
-
Removing Elastic Effects in FWI Using Supervised Cycled Generative Adversarial Networks
More LessSummaryWe use a CycleGAN to map acoustic synthetic data to elastic data, and to map elastic field data to acoustic data, and use the resulting data to perform acoustic FWI on a 3D field dataset that shows strong elastic effects at top chalk. Using machine learning to change the effective physics of field data has many other potential applications.
-
-
-
Extracting and Classifying Graphic Information from Geoscience Unstructured Documents Using Deep Learning Based Computer Vision Approaches.
Authors H. Blondelle, J. Micaelli and P. KaurSummaryFormation Evaluation Logs (FEL) and composites integrate together a lot of information gathered together by the well-site geologist while drilling and logging. But, since they are frequently published as unstructured documents, they are not easy to use as a source of information in digital business processes.
We had the opportunity to support our customer Equinor to “read” lithological columns, O&G show symbols, and geological descriptions from FEL and composites using a state-of-the-art computer vision approach called YOLO and our indexing solution named iQC. A process based on YOLO and iQC transforms the graphical information into usable, numeric and text values that can be consumed by business databases. Computer vision and semantic analysis models were trained on composite logs, which were tagged by subject matter experts with expected labels. The developed models automatically detect and draw bounding boxes around target objects in test documents. This paper details this experiment, lessons learnt and provides some perspective to improve the accuracy of the first results obtained.
-
-
-
Stabilized Super Resolution Deep Generative Networks for Seismic Data
Authors R.S. Ferreira, E. Zabihi Naeini and E. Vital BrazilSummaryHigh-resolution seismic data enable us to characterize the reservoirs with higher accuracy and/or detect smaller targets. Enhancing the seismic bandwidth can be achieved with broadband acquisition, various processing algorithms or a combination of both. In contrast to classic spectral matching type processes, we propose to take a different approach by using deep Generative Adversarial Networks (GANs). In theory, they can reconstruct the seismic data both temporally and spatially. This is inherent by design given the convolutional architecture of the GANs. That means GANs allow recovering the frequency content or the missing traces in seismic data. We propose amplitude encoding and histogram equalization to stabilize the performance of GANs on seismic data and show promising preliminary results for typical seismic processing and interpretation applications.
-
-
-
Machine Learning For DAS Microseismic Event Detection
Authors S. Horne, A. Baird, A. Stork and G. NaldrettSummaryEmerging acquisition systems based on fiber optic technologies such as Distributed Acoustic Sensing are enabling dense spatial and temporal sampling of strain fields. This has resulted in a large increase in the volume in the volume of data and the rate at which data is generated. These increases can be interpreted as satisfying two of the 3 ‘V's of ‘Big Data’ i.e. Volume and Velocity (the third V refers to data Variety). In this presentation we show how we have used big data technologies such as Apache Hadoop and Apache Spark to tackle these data issues for the specific problem of microseismic event detection. Furthermore, traditional approaches were thought to be unlikely to efficiently scale to these new data so we turned to machine learning approaches based on computer vision. Rather than trying state of the art technologies such as Convolutional Neural Nets we decided to try a mature technology used for face detection known as a Haar Cascade. We have tested this approach on field data and found that this approach can work well and are motivated to try newer machine learning techniques with the expectation of moving beyond microseismic event detection.
-
-
-
EarthNET a native cloud web based solution for next generation subsurface workflows
Authors D. Oikonomou, E. Larsen, B. Alaei, G. Stefos and S. PurvesSummaryFaster and better data driven decision-making and shorter times to first oil and gas top the list of expected benefits that digital technologies can drive for upstream oil and gas companies. In the oil industry, Artificial Intelligence (AI) and Machine Learning (ML) tools have already moved from R&D projects into G&G tool boxes, slowly transforming the subsurface workflow.
We will discuss about cloud platforms and demonstrate how such an integrated platforms provides both the data access and applications required to apply ML at scale with examples that include integration of multi regional datasets.
We will show that such platforms are not only enhancing further creativity and enabling data driven decisions but in addition will shorten time to oil which seems to be the next challenge of the industry.
-
-
-
Can Machines Learn to Pick Horizons in Post Stack Data?
Authors L. Yalcinoglu and C. StotterSummaryThe presented method applies a supervised deep learning (DL) method to detect the horizons throughout a seismic dataset with high detection accuracy.
-