With the recent tremendous development in algorithms, computations power and availability of the enormous amount of data, the implementation of machine learning approach has spurred the interest in oil and gas industry and brings the data science and analytics into the forefront of our future energy. The idea of using automated algorithms to determine the rock facies is not new. However, the recent advancement in machine learning methods encourages to further research and revisit the supervised classification tasks, discuss the methodological limits and further improve machine learning approach and classification algorithms in rock facies classification from well-logging measurements. This paper demonstrates training different machine learning algorithms to classify and predict the geological facies using well logs data. Previous and recent research was done using supervised learning to predict the geological facies.
This paper compares the results from the supervised learning algorithms, unsupervised learning algorithms as well as a neural network machine learning algorithm. We further propose an integrated approach to dataset processing and feature selection. The well logs data used in this paper are for wells in the Anadarko Basin, Kansas. The dataset is divided into training, testing and evaluating wells used for testing the model. The objective is to evaluate the algorithms and limitations of each algorithm. We speculate that a simple supervised learning algorithm can yield score higher than neural network algorithm depending on the model parameter selected. Analysis for the parameter selection was done for all the models, and the optimum parameter was used for the corresponding classifier.
Our proposed neural network algorithm results score slightly higher than the supervised learning classifiers when evaluated with the cross-validation test data. It is concluded that it is important to calculate the accuracy within the adjacent layers as there are no definite boundaries between the layers. Our results indicate that calculating the accuracy of prediction with taking account the adjacent layers, yield higher accuracy than calculating accuracy within each point. The proposed feed-forward neural network classifier trains using backpropagation (gradient descent) provides accuracy within adjacent layers of 88%. Our integrated approach of data processing along with the neural network classifier provides more satisfactory results for the classification and prediction problem. Our finding indicates that utilizing simple supervised learning with an optimum model parameter yield comparable scores as a complex neural network classifier.
In this paper, a new type of sand-consolidation low-viscous binding material, based on a combination of inorganic and organic components, is presented. Researchers have developed a novel water-based-drilling-fluid system compatible with deepwater HP/HT wells in the Lingshui Block on the basis of a conventional drilling fluid and further optimization. The combination of ultrasonic pulse-echo and flexural-attenuation measurements was adopted in this project in the South China Sea for cement-integrity evaluation. Weather conditions aging offshore facilities presented challenges during a monsoon in the South China Sea. This paper presents the results of a performance-improvement effort in extended-reach drilling (ERD) in three campaigns in the Huizhou HZ 25-4 oil field in the South China Sea.
Africa (Sub-Sahara) Oil was discovered at the Ekales-1 wildcat well located in northern Kenya. The well has a potential net oil pay in the Auwerwer and Upper Lokone sandstone reservoirs of between 197 ft and 322 ft. Tullow (50%) is the operator in partnership with Africa Oil (50%). Drillstem tests on the Pweza-3 well offshore Tanzania flowed at a maximum rate of 67 MMscf/D of gas. The tests confirmed the excellent properties of the Tertiary-section reservoir. BG Group (60%) is the operator in partnership with Ophir Energy (40%). Asia Pacific China National Offshore Oil Corporation issued a tender to invite foreign firms to bid for oil and gas blocks in the east and south China Sea. Twenty-five offshore blocks will be offered, including 17 in the South China Sea, three in the East China Sea, and five in the Yellow and Bohai seas.
In this paper, we present for the first time, a classification system for naturally-occurring gas hydrate deposits existing in the permafrost and marine environment. This classification is relatively simple but highlights the salient features of a gas hydrate deposit which are important for their exploration and production such as location, porosity system, gas origin and migration path. We then show how this classification can be used to describe eight well-studied gas hydrate deposits in permafrost and marine environment. Potential implications of this classification are also discussed.
Liu, Jinpeng (Data Processing Company, Geophysical-COSL) | Zhong, Mingrui (Data Processing Company, Geophysical-COSL) | Fang, Zhongyu (Data Processing Company, Geophysical-COSL) | Dan, Zhiwei (Data Processing Company, Geophysical-COSL) | Sun, Leiming (Data Processing Company, Geophysical-COSL)
With the deepening of exploration and development, exploration in the south China sea is faced with increasingly complex geological targets, including complex fault blocks, lithological targets, middle and deep strata, small scales and more subtle and complex exploration targets. At present, the internationally recognized best seismic solution is "two wide and one high" acquisition and processing, namely wide azimuth, broadband, high-density field observation system and targeted processing. In the aspect of wide azimuth acquisition and processing, domestic land acquisition has also developed greatly, and in the past few years, there has also been a heated debate on the advantages and disadvantages of wide and narrow azimuth acquisition in complex areas. However, in the aspect of offshore acquisition, wide-azimuth acquisition is rarely carried out. The main reason is that the construction cost and difficulty are higher. With the primary 3d coverage of some mature areas on the sea, it is still unable to meet the exploration needs. In the past, the seismic observation system design was mostly based on the seismic acquisition and survey lines in accordance with the direction of vertical structure strike, so as to facilitate the accurate imaging of the main structure or the construction along the long axis of the work area according to the cost of offshore acquisition. However, in fact, the fault strike in the tectonic development area is complex and changeable, and there is no uniform rule. Some small faults that control the trap are completely perpendicular to the large structural strike, so the old 3d will lead to some poor fault imaging. According to the practical test data analysis, found that the different line direction observed data imaging effect is different, therefore some recently in the south China sea area for the secondary multi-dimensional three-dimensional attempt at sea to achieve multiple acquisition costs are relatively low, and the construction is convenient, but if you want to achieve benefit maximization, must consider the joint use of composite materials. Compared with the factors that need to be considered in collection and processing, it is much more complex, and the development of processing is relatively lagging behind. The main anisotropy is not fully considered, and relatively simple superposition is often difficult to reflect the effect of multi-direction.
One method of sustaining and optimizing a well through its lifetime is underbalance perforating. When hydrostatic pressure inside the wellbore at the zone of interest is kept at less than the expected reservoir pressure, the damaged and crushed zones across the critical matrix at the reservoir that cause low permeability in the perforation tunnels will be immediately cleaned up as soon as communication to the reservoir is established upon perforating. In an operation offshore Malaysia, underbalance perforating was performed in injection wells, rather than producing wells, to optimize injection rates. The operation employed a fiber-optic firing head deployed on a fiber-optic coiled tubing (CT) real-time telemetry system.
The most common and effective method to achieve underbalance is displacing the well to a lighter fluid, less than the water gradient, prior to perforating. Subhydrostatic wells with low bottomhole reservoir pressure pose challenges to achieving the underbalance state. For these wells, well fluids must be removed via nitrogen displacement and the completion perforated with a nitrogen cushion. After underbalance is reached, the well is ideally ready to be perforated as it is, without introduction of additional fluids.
In the offshore Malaysia field, water injector wells had been perforated overbalance because the objective of the wells was injection and not production. However, the injection rate of these water injectors started to decline below the optimum design rate only after a short period, thus affecting the production rate of the neighboring oil and gas producers. Two pilot wells were designed to be perforated underbalance, achieving immediate cleanup after firing. The challenge was to perform an underbalance perforation in a low-pressure, depleted reservoir, using nitrogen as a displacement fluid. After this condition was fulfilled with a 500-psi differential, the well was to be perforated without any liquid introduction to activate the guns, which restricted the use of pressure- and ball-activated firing heads.
The fiber-optic-enabled firing head deployed on CT with real-time telemetry system is considered the most efficient intervention approach to overcome the challenges set. The new firing head will allow the perforating command to be given through an optical signal instantaneously at depth with no disturbance to the well fluid dynamics. This technique will also optimize an online rig operation where displacement, perforation, and nitrogen lift contingency can be performed in one CT run, hence reducing operating costs. Since the initial startup of the two pilot wells, the injection rates of the wells are at optimum, and the performance gained from the two wells has increased overall production in the field. Real-time underbalance perforating is thus seen as the way forward not only to enhance producing wells, but also to boost injectors as well, prolonging the life of an offshore oilfield.
Li, Qinghui (Qianhai Harbour Energy Technology Development Shenzhen Co., LTD) | Zhu, Jinzhi (Petrochina Tarim Oilfield Company) | Li, Shaoxuan (Xidian University) | Zhang, Shaojun (Petrochina Tarim Oilfield Company) | Hisham, Nasr-El-Din (Texas A&M University) | Ren, Lingling (Petrochina Tarim Oilfield Company) | Li, Jiaxue (Petrochina Tarim Oilfield Company) | Al-Mujalhem, Manayer (Texas A&M University)
Global energy demand has driven the petroleum industry to develop hydrocarbon resources from extremely harsh formations which contain ultra-high pressure and temperature (HPHT) reservoirs. Ultra-high density drilling fluids are critical to successful drilling and completion practices in all of these wells. In this paper, potential weighting materials were systematically evaluated and screened to accomplish an ultra-high density oil-based drilling fluid system (19.62 to 22.12lb/gal) aimed to utilize in ultra HPHT conditions (>30000psi and >410°F).
Several potential high-density weighting materials were evaluated in the laboratory conditions. Basic properties (pure density, particle size/distribution, surface area etc.) were evaluated and compared. Special treatments were conducted to optimize the properties of weighting materials. HPHT filtration tests under static and dynamic conditions were conducted at higher than 410°F and 300 psi. Real cores with an average porosity of 19% and an average permeability of 50 mD were used in the filtration tests. Rheological properties, sag tendency, the volume of filtrate, and the filtrate cake characterization of oil-based drilling fluids were measured before and after heating at 410°F for 16 hours.
Results revealed that ultra-micro manganese and ilmenite complex after suitable surface treatment could act as an ideal weighting material than ultra-pure barite or other materials, which could fail in rheology and sag controlling measurement with such high temperature and density. The viscosity and filtration analysis confirmed the stability and reliability of this novel ultra-high density oil-based drilling fluid.
This study developed a challenged drilling fluid system under critical testing states, as well as established a systematical laboratory evaluation and screening procedure of weighting materials for ultra-deep wells and contributed recommendations on how to utilize it in the fields.
Deep geological repositories are a feasible and reliable way to dispose of spent nuclear fuel, and the mechanical and thermal properties of the host rock are key factors in their long-term safety. The mechanical properties of rock, including uniaxial compressive strength, tension strength and elastic modulus are thermally-dependent - for example, the uniaxial compressive strength of rock will decrease with increasing temperature, as a result of thermally-induced cracks between mineral particles. However, thermally-induced cracks and the micro-mechanisms of rock damage are difficult to observe by experiments. We therefore used uniaxial compressive strength test and Brazilian disk PFC models with heterogeneous thermal expansion properties to reveal the micro-mechanisms of rock damage induced by thermal heating.
Based on the mineral compositions of granite, heterogeneous models contained four particle types with different thermal expansion properties (quartz, potassium feldspar, plagioclase, and biotite). A series of heating simulations (temperature increment: ΔT = 80 – 300 °C) showed that thermally-induced cracks occurred when the temperature increment exceeded 250 °C, and that the main crack type was tension crack. In homogeneous PFC models no thermally-induced cracking occurred even at ∆T = 300 °C. In simulations of thermal-mechanical coupling, the heterogeneous PFC models have the advantage of revealing differences in radius expansion of various mineral particles with increasing temperature, resulting in the additional and uneven distribution of contact force, stress concentration between different particles and bond breaking, and it lead to decreases in uniaxial compressive strength, elastic modulus and initial crack stress.
In the construction and operation of deep geological repositories of nuclear waste, the physical and mechanical properties of host rock are influenced by excavations and decay heat induced by the installation of nuclear waste canisters. This may result in a variety of changes in rock porosity and seepage conditions in the rock mass, the formation of excavation damage zones and the occurrence of thermally-induced rock damage, etc. Thermally-induced damage results from heterogeneous thermal properties in rock - for example, the thermal expansion coefficient of quartz (24.3 × 10−6 /°C) is three times that of biotite (8.0 × 10–6 /°C), leading to the differential expansion of mineral volume and macro-mechanical properties of rock with increasing temperature. In order to understand and interpret the influences of thermally-induced damage on the mechanical behavior of rock, homogeneous and heterogeneous bonded particle models were adopted to simulate a uniaxial compressive test (UCS) and Brazilian indirect tensile test in this study.
Zhang, Guoqiang (China National Offshore Oil Corporation, Ltd. Tianjin Branch) | Tan, Zhongjian (China National Offshore Oil Corporation, Ltd. Tianjin Branch) | Hao, Zhongtian (China National Offshore Oil Corporation, Ltd. Tianjin Branch) | Fan, Zhao Ya (Schlumberger) | Chen, Ji Chao (Schlumberger) | Gao, Bei (Schlumberger) | Shelomikhina, Irina (Schlumberger) | Zhang, Tao (Schlumberger)
The new gas condensate has been discovered recently in Bohai Bay China, and due to the complex of the fluid behaviour in nature, appropriate characterization of the in-situ fluids and relevant flow testing can provide valuable insight into gas condensate reservoir forecasting. This paper discussed a comprehensive methodology for reliable productivity evaluation with the aid of in-situ fluid characterization and interval pressure testing, and its results as a key factor not only fed into DST design and equipment optimization for better production evaluation design but also can be optimal production prediction for reserve booking. Based on the efficient and reliable productivity evaluation, we can make the real-time decision on whether to carry on DST and which DST equipment should be chosen and where to test it. However, we has been dealing with heavy oil for very long time, and the common DST equipment in Bohai Bay is not designed for condensate reservoir, but heavy oil, plus there were no published case studies in China about DST modification from heavy oil to gas condensate, and what can be done for production forecasting in gas condensate reservoir. In this paper, a new solution was proposed base on the problem elaborated above. An integrated approach was conducted to overcome challenges by using all available data from Wireline Formation Tester (WFT) and conventional log data. An accurate permeability is always the foundation of a good prediction, a single well model was built by taking full advantage of all permeability information; Compositional model is the best way to simulate the condensate phase change, in this paper, a calibrated composition was deduced by combining with semi-quantitative DFA composition and PVT analysis software (PVTi) via calibrating downhole measured GOR and density; Black oil model is the most common and popular method to evaluated single well productivity, but the uncertainty of condensate viscosity usually leads huge error, based on the calibrated composition results, evaluated the downhole fluid viscosity, after downhole fluid viscosity correctly evaluated, black oil model can also be applied in condensate productivity evaluation. DST results were used to validate the results from this approach, the error percentage range is around 10% compare with DST results. This new solution has following values: an accurate gas condensate production forecast become possible even without PVT lab results.
Summary High-resolution 2-D single-channel seismic data were used to investigate the types and characteristics of the late Miocene to present seismic facies in the northern South China Sea (SCS) basin. Eight types of seismic facies were identified, which are parallel to subparallel, wavy, mounded divergent, divergent fill, progradational, stratified high-amplitude, chaotic high-amplitude, and transparent facies, respectively. Sedimentological interpretation of the seismic facies indicates that the deep-sea basin has not always been tranquil as imagined, large-scale sediment waves, overbank deposits, turbidity-current channel and levee deposits, and submarine fan lobes are developed. The gravity-flow deposits are mostly distributed temporally in the upper Miocene and Quaternary sequences and spatially in the northwestern, northern and eastern parts of the basin. Introduction The South China Sea basin is a large-volume sediment sink, in which lots of sediments from the surrounding continental margins and seamounts have been deposited.