This article takes an approach to creating a machine learning model for the oil and gas industry. This task is dedicated to the most up-to-date issues of machine learning and artificial intelligence. One of the goals of this research was to build a model to predict the possible risks arising in the process of drilling wells. Drilling of wells for oil and gas production is a highly complex and expensive part of reservoir development. Thus, together with injury prevention, there is a goal to save cost expenditures on downtime and repair of drilling equipment. Nowadays, companies have begun to look for ways to improve the efficiency of drilling and minimize non-production time with the help of new technologies. To support decisions in a narrow time frame, it is valuable to have an early warning system. Such a decision support system will help an engineer to intervene in the drilling process and prevent high expenses of unproductive time and equipment repair due to a problem. This work describes a comparison of machine learning algorithms for anomaly detection during well drilling. In particular, machine learning algorithms will make it possible to make decisions when determining the geometry of the grid of wells—the nature of the relative position of production and injection wells at the production facility. Development systems are most often subdivided into the following: placement of wells along a symmetric grid, and placement of wells along a non-symmetric grid (mainly in rows). The tested models classify drilling problems based on historical data from previously drilled wells. To validate anomaly detection algorithms, we used historical logs of drilling problems for 67 wells at a large brownfield in Siberia, Russia. Wells with problems were selected and analyzed. It should be noted that out of the 67 wells, 20 wells were drilled without expenses for unproductive time. The experiential results illustrate that a model based on gradient boosting can classify the complications in the drilling process better than other models.

With the depletion of traditional energy resources, the share of heavy-oil production has been increasing recently. According to some estimates, their reserves account for 80% of the world’s oil resources. Costs for extraction of heavy oil and natural bitumen are 3–4 times higher than the costs of extracting light oil, which is due not only to higher density and viscosity indicators but also to insufficient development of equipment and technologies for the extraction, transportation, and processing of such oils. Currently, a single pipeline system is used to pump both light and heavy oil. Therefore, it is necessary to take into account the features of the heavy-oil pumping mode. This paper presents mathematical models of heavy-oil flow in oil-field pipelines. The rheological properties of several heavy-oil samples were determined by experiments. The dependencies obtained were used as input data for a simulation model using computational fluid dynamics (CFD) methods. The modeling condition investigates the range of shear rates up to 300 s⁻¹. At the same time, results up to 30 s⁻¹ are considered in the developed computational models. The methodology of the research is, thus, based on a CFD approach with experimental confirmation of the results obtained. The proposed rheological flow model for heavy oil reflects the dynamics of the internal structural transformation during petroleum transportation. The validity of the model is confirmed by a comparison between the theoretical and the obtained experimental results. The results of the conducted research can be considered during the selection of heavy-oil treatment techniques for its efficient transportation.

This paper considers the issue of associated petroleum gas utilization during hydrocarbon production in remote petroleum fields. Due to the depletion of conventional oil and gas deposits around the globe, production shifts to hard-to-recover resources, such as heavy and high-viscosity oil that requires a greater amount of energy to be recovered. At the same time, large quantities of associated petroleum gas are extracted along with the oil. The gas can be utilized as a fuel for power generation. However, even the application of combined power modes (combined heat and power and combined cooling heat and power) cannot guarantee full utilization of the associated petroleum gas. Analysis of the electrical and heat loads’ graphs of several oil fields revealed that the generated thermal energy could not always be fully used. To improve the efficiency of the fuel’s energy potential conversion, an energy system with a binary power generation cycle was developed, consisting of two power installations—a main gas microturbine and an auxiliary steam turbine unit designed to power the technological objects in accordance with the enterprise’s power load charts. To provide for the most complete utilization of associated petroleum gas, a gas-to-liquid system is introduced, which converts the rest of the gas into synthetic liquid hydrocarbons that are used at the field. Processing of gas into various products also lowers the carbon footprint of the petroleum production. Application of an energy system with a binary power generation cycle makes it possible to achieve an electrical efficiency up to 55%, at the same time maintaining high efficiency of consumers’ energy supply during the year. The utilization of the associated petroleum gas in the developed system can reach 100%.

A steel riser has benefits over a flexible riser in terms of pressure rating and cost. A steel riser in lazy wave configuration (steel lazy wave riser (SLWR)) is often considered as a good alternative solution for harsh environments where large floater excursions take place. The SLWR configuration is achieved by introducing buoyancy modules into a steel catenary riser (SCR). The buoyancy modules act as a damper and isolate the floater motions from the critical touchdown area. Hence, the SLWR generally has better overall performance than an SCR configuration. This paper attempts to analyze the correlation between the geometric shapes of the SLWR configuration with its capability to absorb the dynamic loadings. For deepwater cases, the behavior of the bottom part of the riser is correlated with the velocity experiences at the riser's hang-off location. Hence, the riser's performance is analyzed by comparing the velocity at the riser's hang-off with the velocity at the sag, hog, and near touchdown. The geometric shape of an SLWR is represented by its arch height, which is the vertical distance between the lowest point at the sag and the highest point at the hog bend of a riser. The results show that there is a correlation between the arch height of an SLWR with the riser's strength and wave-induced fatigue performance. SLWR configurations with higher arch generally have greater capability to absorb the dynamic loadings, as indicated by the lower velocities along the riser, which leads to lower stress utilizations and lower fatigue damage.

The time required for rescue is a critical factor for surviving a marine incident. The regulatory framework, International Maritime Organization (IMO) Polar Code, utilizes a risk-based approach. It states that the vessel operators are to define the time required for rescue but never less than 5 days. Based on experience from the classification society DNV GL, utilization of the minimum requirement of five days is the current industry standard when conducting risk assessments. The dimensioning of search and rescue resources is a national issue. There are no international requirements defining the adequacy of the resources for different geographical areas. The remoteness and lack of resources present within the IMO Polar Code area imposes a significant challenge for mariners in distress. The time required for rescue is highly dependent on multiple variables. Based on this study, the number of persons to be rescued, the number and type of evacuation platforms and the distance each evacuation platform must travel significantly impacts the time required for rescue. In addition, the meteorological and oceanographical (metocean) conditions play a significant role when determining the efficiency of a search and rescue operation.

This paper presents the stochastic properties of orbital velocities of random water waves in intermediate water depth. Both the emergence effect and weak nonlinear effects are studied; the theoretical predictions are compared with measured kinematics and the deviations from linear theory are quantified. This study includes new ideas in fluid dynamics. An analytic formula for probability distribution for velocities modified by the emergence effect as well as by nonlinearities of the wave motion in intermediate water depth is developed. This probability function gives us the first statistical moment, the second statistical moment for modified velocities in an analytical form, and by numerical integration the third statistical moment for modified velocities. The theoretical formulae for the statistical moments for surface elevation and for velocities up to third order, with nonlinearities of the motion taken into account, for the case when the emergence effect can be neglected, i.e. below the surface layer, have been developed. This includes a generalized formula for free-surface elevation setdown and calculation of the asymmetry of the horizontal velocity, which is found to be negative in agreement with measurements of Anastasiou et al. (1982b). From the first statistical moment of the modified horizontal velocity, the mean flux between any two levels in the wave flume may be calculated. When the integration is carried out from the bottom up to + ∞, it leads in approximation to the formula for total mean flux found by Phillips (1960). This agreement with Phillips’ formula encourages one to interpret the positive mean value of horizontal velocities as a ‘real current’. This interpretation also provides a new understanding of the fluid dynamic implications of results presented by Tung (1975). Theoretical prediction of the measured kinematics has allowed a better estimation of the return flow in the wave flume, and in the vicinity of the mean water level currents in two different directions are noted. Firstly, the emergence effect gives rise to a current at the mean water level in the direction of the wave advance. Secondly, a flow in the opposite direction, interpreted as a return current in the wave flume, is noticed just below that level.