The rapid development of machine learning algorithms and the massive accumulation of well data from continuous monitoring has enabled new applications in the oil and gas industries. Data gathered from well sensors are a foundation of the oilfield digitization and data-driven analysis. Here, we describe a deep learning approach to predict the long-term well performance based on a moderate duration of well monitoring data.
In this study, we first developed the data processing procedures for oilfield time series data and determined the proper selection of data sampling frequency, parameter combinations and data structures for deep learning models. Then we explored how Deep Learning (DL) models can be employed for well data analysis and how can we combine physics and DL models. Recurrent Neural Network (RNN) is a type of sequential DL model, which can be utilized for time series data analysis. This approach preserves preceding information and yields current response with memory of prior well behavior. Two candidate RNN models were tried to determine how well they were able to improve the accuracy and stability of well performance estimates. These two methods are Gated Recurrent Unit (GRU) and Long Short Term Memory (LSTM). In addition, a novel combination of RNN with Convolutional Neural Networks (CNNs), Long- and Short-term Time-series network (LSTNet), was also investigated.
These various models were tested and compared based on the public production datasets from Volve Field. Both GRU and LSTM achieved higher accuracy in performance prediction compared to the simple RNN. In the case of frequent well shut-in and opening, the failure in capturing fast pressure responses and the extreme fluctuations with the simple RNN ultimately leads to high error. In contrast, LSTNet is more stable to frequent or significant well variations. With advanced deep learning structures, engineers can interpret long-term reservoir performance information from responses estimated by deep learning models, instead of performing costly well tests or shut-ins.
Africa (Sub-Sahara) Aminex Petroleum Egypt (APE), a subsidiary of UK-based Aminex, discovered oil at its South Malak-2 (SM2) well on the West Esh el Mellaha-2 concession in Egypt. Tests showed flow rates of approximately 430 B/D of 40 API gravity crude oil. Based on the findings at SM2, a full field development program will be presented to the Egyptian authorities and the joint venture partners before commercial development. APE is the operator of the license with partner Groundstar Resources. Foxtrot International discovered oil and gas at its Marlin North-1 well in Block CI-27, offshore Cote d'Ivoire. A 22-m perforated section of a gas-bearing column in a Turonian interval flowed at a stabilized rate of 25 MMcf/D of gas and 150 B/D of condensate through a 46/64-in.
Copyright 2019 held jointly by the Society of Petrophysicists and Well Log Analysts (SPWLA) and the submitting authors. ABSTRACT Today, many machine learning techniques are regularly employed in petrophysical modelling such as cluster analysis, neural networks, fuzzy logic, self-organising maps, genetic algorithm, principal component analysis etc. While each of these methods has its strengths and weaknesses, one of the challenges to most of the existing techniques is how to best handle the variety of dynamic ranges present in petrophysical input data. Mixing input data with logarithmic variation (such as resistivity) and linear variation (such as gamma ray) while effectively balancing the weight of each variable can be particularly difficult to manage. DTA is conceived based on extensive research conducted in the field of CFD (Computational Fluid Dynamics). This paper is focused on the application of DTA to petrophysics and its fundamental distinction from various other statistical methods adopted in the industry. Case studies are shown, predicting porosity and permeability for a variety of scenarios using the DTA method and other techniques. The results from the various methods are compared, and the robustness of DTA is illustrated. The example datasets are drawn from public databases within the Norwegian and Dutch sectors of the North Sea, and Western Australia, some of which have a rich set of input data including logs, core, and reservoir characterisation from which to build a model, while others have relatively sparse data available allowing for an analysis of the effectiveness of the method when both rich and poor training data are available. The paper concludes with recommendations on the best way to use DTA in real-time to predict porosity and permeability. INTRODUCTION The seismic shift in the data analytics landscape after the Macondo disaster has produced intensive focus on the accuracy and precision of prediction of pore pressure and petrophysical parameters.
A high rate of penetration (ROP) is considered one of the most sought-after targets when drilling a well. While physics-based models determine the importance of drilling parameters, they fail to capture the extent or degree of influence of the interplay of the different dynamic drilling features. Parameters such as WOB, RPM, and flowrate, MSE, bit run distance, gamma ray for each rock formation in the Volve field in the North Sea were examined ensuring an adequate ROP while controlling the tool face orientation is quite challenging. Nevertheless, its helps follow the planned well trajectory and eliminates excessive doglegs that lead to wellbore deviations. Five different Machine Learning algorithms were preliminary implemented to optimize ROP and create a less tortuous borehole. The collected data was cleaned and preprocessed and used to structure and train Random Forest, Support Vector Regression, Ridge Regression, LASSO, and Gradient Boosting, XG boost among others and the appropriate hyperparameters were selected. A successful model was chosen based on maximized ROP, minimized deviation from planned trajectory, and lower CPF. An MAEP of 15% was achieved using GBM boost followed AdaBoost. The algorithms have demonstrated competence on the historical dataset, accordingly it will be further tested on blind data to serve as a real-time system for drilling optimization to enable a fully automated system.
Maintaining a stable borehole and optimizing drilling are still considered to be vital practice for the success of any hydrocarbon field development and planning. The present study deliberates a case study on the estimation of pore pressure and fracture gradient for the recently decommissioned Volve oil field at the North Sea. High resolution geophysical logs drilled through the reservoir formation of the studied field have been used to estimate the overburden, pore pressure, and fracture pressure. The well-known Eaton’s method and Matthews-Kelly’s tools were used for the estimation of pore pressure and fracture gradient, respectively. Estimated outputs were calibrated and validated with the available direct downhole measurements (formation pressure measurements, LOT/FIT). Further, shear failure gradient has been calculated using Mohr-Coulomb rock failure criterion to understand the wellbore stability issues in the studied field. Largely, the pore pressure in the reservoir formation is hydrostatic in nature, except the lower Cretaceous to upper Jurassic shales, which were found to be associated with mild overpressure regimes. This study is an attempt to assess the in-situ stress system of the Volve field if CO2 is injected for geological storage in near future.
With increasing focus on identifying cost effective solutions to well design with minimal impact on productivity, this paper will focus on an alternative to cesium formate as perforation fluid in the HPHT Gudrun field operated by Statoil. Cesium formate has been used with success for drilling and perforating many HPHT wells. However, given the significant cost of this fluid coupled with low oil prices, Statoil wanted to perform testing to assess the performance of an alternative low ECD oil based mud as a perforation fluid. The paper will describe the extensive qualification testing that has been performed. This includes coreflooding using representative plugs from Gudrun under downwhole temperature and pressure conditions. In addition, eight Section IV perforation tests have been performed to compare the performance of Cs formate and the low ECD oil based mud. These tests were undertaken using gas and oil saturated cores to reflect different production scenarios. The main aspects of the perforation operation that were reflected in the test design were as follows. Perforating at reservoir pressure and laboratory testing temperature of approximately 100°C Simulating an extended shut in period after perforation Undertaking a clean up sequence using scaled down flowrates
Perforating at reservoir pressure and laboratory testing temperature of approximately 100°C
Simulating an extended shut in period after perforation
Undertaking a clean up sequence using scaled down flowrates
Based on the results of the coreflooding combined with the section IV 19B testing, the low ECD OBM was selected as the perforating fluid for use on Gudrun. The perceived benefits of using the low ECD OBM were as follows. Simplification: use of the same fluid for drilling and perforating the reservoir section. Tangible cost savings in fluid cost and time savings of approximately 40M NOK ($5M). Potentially increased productivity compared to cesium formate. Improved standardization of the operational sequence.
Simplification: use of the same fluid for drilling and perforating the reservoir section.
Tangible cost savings in fluid cost and time savings of approximately 40M NOK ($5M).
Potentially increased productivity compared to cesium formate.
Improved standardization of the operational sequence.
Perforation modelling is described and comparison is made between this and the Section IV tests. Finally, the well start-up experiences and production data are presented demonstrating the effectiveness of the low ECD oil based mud as a perforation fluid.
Bluewater's Haewene Brim Floating Production Storage and Offloading (FPSO), moored by means of an internal turret mooring system in the central North Sea since 1998, ran out of its original 15 years design life. Good field performance and the tie-in of another field resulted in a targeted lifetime extension for the FPSO of 18-20 years. In the late 90s a number of turret mooring systems, with a typical design life of 15 - 20 years have been installed in the North Sea. A number of these, currently or in the near future, might also be subject to lifetime extension programs. The technical challenges experienced during the lifetime extension project of the Haewene Brim FPSO mooring system will be discussed in this paper. The main objective for the lifetime extension program of the mooring system was twofold: to design a mooring system which allows for continuous operations for the extended lifetime and to maximize the utilization of existing components.
Mode-converted PS-waves can provide valuable information for anisotropic parameter estimation that cannot be obtained from compressional waves alone. In a previous paper, we developed an efficient tomographic methodology for 2D joint velocity analysis of PP and PS data from VTI (transversely isotropic with a vertical symmetry axis) media. The algorithm is designed to flatten image gathers of PP-waves as well as of pure SS reflections computed using the PP+PS=SS method. An important additional constraint is provided by codepthing of the migrated PP and SS sections. The model is divided into square cells, and the parameters VP0, VS0, ε, and δ are defined at each grid point. Here, this methodology is applied to multicomponent data recorded on a 2D line from an OBS (ocean bottom seismic) survey acquired at Volve field in the North Sea. Although the parameter-updating procedure did not include any borehole information, the VTI model obtained by joint tomography of the recorded PP-waves and constructed SS-waves produced generally well-focused PP and PS depth images. The depth consistency between the migrated PP and PS sections also corroborates the accuracy of the velocity-analysis algorithm.
Nes, Olav-Magnar (Det Norske Oljeselskap ASA) | Boe, Reidar (SINTEF Petroleum Research) | Sonsteboe, Eyvind F. (SINTEF Petroleum Research) | Gran, Kjetil (Det Norske Oljeselskap ASA) | Wold, Sturla (Det Norske Oljeselskap ASA) | Saasen, Arild (Det Norske Oljeselskap ASA) | Fjogstad, Arild (Baker Hughes A/S)
Severe borehole-stability problems were encountered in a recent exploration well in the Norwegian North Sea. The problems occurred when drilling through Tertiary shale sections interbedded with permeable sand layers. Drilling was initially performed with water-based mud (WBM). However, because the section target was not able to be reached after more than 2 weeks of operation, the section was plugged back, and a sidetrack was drilled with an oil-based mud (OBM) without encountering major operational problems. On the basis of the post-drill analysis of drilling data, well logs, drill cuttings, and borehole cavings sampled from the well, this paper describes how the complex combination of drilling fluid salt concentration and geological constraints may be used to ensure successful future drilling operations in this part of the North Sea.Cuttings and preserved cavings collected during the drilling operation were selected from several depth intervals identified as potentially troublesome from drilling experience and log data. The determination of cuttings mineralogy enabled a better prediction of how the time dependency of the stable drilling-fluid-density window is influenced by an interaction between the shale and the drilling fluid. Mechanical strength is a key input parameter when predicting borehole stability. Dedicated rock-mechanical punch measurements on cavings were used to confirm the prediction of strength from log data alone. The examination of caving surfaces revealed the possible presence of in-situ-fractured rock. Such fractures would require special measures while drilling to maintain stability. Subsequently, a borehole-stability sensitivity analysis was performed that focused on time-dependent stability in the shale formations.The analysis used cuttings and cavings properties and logs as input. In particular, the modeling showed how the optimal potassium chloride (KCl) concentration in the drilling fluid changes with depth. The modeling further identified a relatively large sensitivity toward borehole inclination--even at fairly small inclinations. This paper thus illustrates the significance of properly accounting for rock-mechanical aspects when planning new wells.