This paper describes a trial project to evaluate autonomous inflow control device (AICD) technology to better manage water production in a large heavy oil field in Colombia. The Cajua block is part of the Rubiales field is in the Llanos basin of Eastern Colombia, and has reserves estimated at 7.5 billion barrels. One of the main production challenges is the high water cut, or BSW, driven by strong aquifer flow in the underlying sands of the Carbonera formation. Many wells experience early water break-through and must be produced above 95% BSW for long periods of time. Horizontal wells typically produce up to 8000 barrels per day of total fluid with electric submersible pump (ESP) on cold production, and do not utilize any thermal recovery methods. The loosely-consolidated sandstone reservoir has variable water saturation and permeability, which has continuously frustrated operators'’ attempts to manage water production ever since the Rubiales field was first brought online in the 1980's.
In late 2018, a three well pilot project was initiated to evaluate the ability of inflow control technology to manage water influx at the sandface of the horizontal completions. Three wells in the Cajua block were equipped with AICD screens and swellable packers to evaluate oil production and water cut. The AICD technology works by limiting water inflow based on fluid viscosity. Each segment, or compartment, of the horizontal wellbore is isolated by swellable packers, and the AICD creates a higher or lower drawdown on the reservoir depending on the fluid properties, favoring the inflow of high-viscosity heavy oil over the low-viscosity water.
The early production results show that AICD completions can effectively manage water production by delaying water break-through and restricting water inflow from the reservoir. Each of the three trial wells responded positively to the autonomous ICDs, allowing engineers to produce heavy oil wells more effectively with lower cumulative water volumes.
This project marks the first implementation in South America of AICD technology with rate-controlled production (RCP) valves to manage water production in a heavy oil field. It is also the second application worldwide, after Canada, to show that AICDs can effectively to manage water cut in a heavy oil, cold-production scenario.
Africa (Sub-Sahara) Marathon Oil has produced first gas from its Alba B3 compression platform offshore Equatorial Guinea. The startup enables the company to convert approximately 130 million BOE of proved undeveloped reserves, which more than doubles its remaining proved developed reserve base in the country. Marathon holds an operating interest of about 65% in the field, with Noble Energy holding the remaining stake. Aminex said that gas production from the Kiliwani North-1 well in Tanzania has reached 30 MMcf/D (about 5,000 BOE/D). The project's commissioning process is expected to conclude with a well test to determine the optimal production rate, which previous test data suggest will be approximately 30 MMcf/D, the company said. The operator of the Kiliwani North Development License, Aminex holds a 54.575% interest in the well.
This paper presents the development and implementation of a production and power optimization system supported by a nodal analysis simulation tool and an updated and calibrated multiphase flow hydraulic model.
The workflow implemented in Castilla field identifies operating conditions that may affect the performance of 577 production wells configured in a complex gathering system, managing multiphase flow for heavy oil with low gas oil ratio and high water cut. Initially, this involves the development of a tool for the systematic data transfer from official databases to the simulation model. Furthermore, the hydraulic model was calibrated to reproduce operating conditions. And once the main nodes and boundary conditions at reservoir, wells, network and processing facility were defined, various optimization scenarios were simulated.
With this workflow the field engineers are capable to propose operating changes to increase production taking into consideration network bottlenecks, power consumption, and facilities limitations. The wells with electrical submersible pumps were ranked into the following categories: pump speed variation, well redesign, surface facilities improvement and drawdown limitation. As a result, actions to be implemented in each well were defined. First, ‘pump speed variation’ classified wells quickly demonstrated how the production and power optimization procedure benefit the field. Specifically, it was possible to optimize the extraction process by increasing the oil production up to 10% using power efficiently. Second, several bottlenecks on the network were identified finding atypical pressure drops, using traditional nodal analysis for pressure, temperatures, velocities, and liquid rates calculation per node. Some of them are related to mechanical configuration and others gas accumulation in pipelines. Third, according to the simulation results, a considerable amount of production can be gained and power consumption can be reduced if these conditions are solved.
This workflow contributes to optimize power consumption and enables faster decision-making to efficiently meet the production targets by increasing oil production and reducing water/oil and power/oil ratios.
Mogollon, M. (Frontera Energy) | Arguelles, A. (Frontera Energy) | Rodriguez, A. (Frontera Energy) | Anaya, O. (Ecopetrol) | Miranda, S. (Schlumberger) | Velasquez, E. (Schlumberger) | Villalobos, J. (Schlumberger)
Electric submersible pump (ESP) applications with heavy oil pose substantial technical challenges: decreased pump head capacity, increased power requirements, and poor cooling capacity. This paper shares experiences, lessons learned, equipment standardization exercises, and improvements performed in a field with heavy oil (13º API to 15ºAPI) to extend mean time between failures (MTBF) in more than 3,000 days in a high ESP population with an average of 500 wells. To achieve production goals and extend ESP run life, the project tracks five elements of the ESP life cycle: design, optimization, failure analysis, monitoring, and equipment standardization. As the field evolved, ESPs faced the challenges such as: increased production with increasing water cuts to higher flow rates in horizontal completions with high dogleg severity. Chronological performance of pump mean time between failure is shown before and after improvements. The ESP lifecycle is used as the basis to analyze several factors that caused either total system failure or inability to meet production expectations. This paper explains the implications of each factor and how they affect ESP components, through case studies of representative or repetitive failures and examples of how they were remediating without incurring the expense of oversizing the ESP equipment or completion. This paper shares lessons learned in a five-year, dynamic heavy oil project and includes practical tools to improve ESP run life and optimize well production, which are applicable across the industry and around the world.
Welcome to the SPE Heavy and Extra Heavy Oil Conference--Latin America, in the beautiful city of Medellín, Colombia! This conference assembles the heavy and extra heavy oil community to document our collective knowledge and best practices in the field. You will have the opportunity to enrich your understanding and find ideas to improve productivities and recovery factors of these abundant but challenging reservoirs that encompass more than 60% of the world's oil resources. We have balanced technical-and business-oriented perspectives to accompany the increased attention to viscous oil reservoirs. The aim is to improve skills to develop, produce, transport, and upgrade economically and with energy efficiency the heavy oil and extra heavy oil resources while creating minimal social and environmental impacts.
Prediction of pore pressure in carbonate formations is a major challenge. When this challenge is combined with the possibility of abnormal pressure occurrence, accurate pore pressure prediction becomes a necessity. The over pore pressure intervals in those formations narrows the size of the drilling window to ~10 pcf (1.33 ppg) which requires high accuracy of pore pressure prediction. Inaccurate pore pressure data can lead to well-control incidents, poor casing designs, and inefficient reservoir modelling among other things. Therefore, a new model was developed to predict pore pressure in over pressured carbonate formation.
By employing acoustic logging data, a correlation between the effective vertical stress and Poisson’s ratio was developed to be used in combination with formation bulk density measurements in order to predict the pore pressure in 3D space relying on the Poro-elasticity theory. Comparing the results of the prediction process with the available Modular formation Dynamics Tester (MDT) yielded a significant fit. Through this correlation, the abnormal variations in the high pressure carbonate are detected to a significant level of accuracy.
Based on the produced correlation, the results of the pore pressure prediction process showed that the pressure gradient increases from 0.48 psi/ft to 0.52 psi/ft in adjacent carbonates. This gradient reached an abnormally high value at 0.95 psi/ft at the base of dolomitic formation, which then decreases at lower formation. Based on the true vertical depth of these high pressure sequences, this abnormal pressure gradient equates to around 140 pcf (18.7 ppg) mud weight. Having knowledge of this value is substantial for minimizing the number of kicks encountered in these zones.
Peregrino is a field offshore Brazil with a FPSO and 2 fixed platforms currently producing close to 100 000 stb/d of oil. Production wells are equipped with electric submersible pumps (ESP) and all produced water is reinjected back into the reservoir using 6 injectors for pressure support. The viscosity of the crude is high (163 cp at reservoir conditions). The present work explores the merits and provides the development details of a model-based production optimization scheme to advice on the best frequency settings of the ESPs in each well. This to ensure the maximum amount possible of oil is produced when water injection capacity is a bottleneck. Furthermore, it studies how the optimal operating conditions change with time.
The optimization formulation considers maximization of total oil production, the maximum allowable water produced given by the available injection capacity and the operational constraints of the ESPs. The optimization was formulated as a Mixed-Integer Linear Problem (MILP). The performance of the wells is represented with piecewise linear tables generated from a commercial simulator.
For the cases tested, the proposed optimization scheme works successfully: It has low running times suitable for real time production optimization, handles successfully multiple operational constraints, and guarantees global optimality. Optimization results are presented for future times.
By using piecewise linear tables to represent the well performances the fidelity of the original model is maintained, while ensuring a robust and fast optimization of the problem. Moreover, it is suitable for frequent model updates without requiring changing the optimization formulation. In summary, this work proposes a method to handle part of the operational complexities of the Peregrino field using a digital twin.
Thin interbed seismic reservoir predictions, attribute analysis and seismic inversion are the most useful techniques for reservoir characterization using seismic data. The seismic inversion and reservoir properties prediction with multi-attributes and neural network techniques was focused on a complex stratigraphic situation; where a strong lateral facies variation was perceived from well information. The combination of 7 seismic attributes including P wave and density, derived from inversion was used in the step-wise regression. The network was trained with “shale content” data available at well locations to generate a shale volume. 3D seismic data was used to propagate the relationship established with this training.
This paper focuses on the methodology of sand body identification for reservoir development with seismic inversion, multi-attributes and probabilistic neural network (PNN), to demonstrate that multi attributes techniques are useful in the identification of thin sand bodies.
Presentation Date: Wednesday, September 27, 2017
Start Time: 11:25 AM
Presentation Type: ORAL
ABSTRACT: This study provides an improved interpretation and delineation of pore pressure in the Chukchi shelf region. Between 1989-91, five exploration wells were drilled on the Chukchi shelf and an array of drilling performance and petrophysical data were acquired. Resistivity, conductivity, sonic travel time, sonic porosity, and drilling exponent data were used to evaluate formation pore pressure. Normal compaction trendlines were established using depth-dependent relationships then integrated with Eaton and adapted Eaton equations to estimate formation pressures. Four of the five wells were found to contain significant overpressure at onset depths ranging from 1098 to 2317 meters subsea. The overpressure is mainly associated with organic-carbon-rich source rocks that have experienced thermal exposures sufficient for hydrocarbon generation. The driving mechanism for the origin and maintenance of the overpressure appears to be linked to hydrocarbon generation at the wellsites and access to hydrocarbons migrating from the regional generation center beneath the Colville basin to the east.
Pore pressures in many deep sedimentary formations often exceed hydrostatic pressures, a condition termed “overpressure”. If not properly balanced by drilling fluids, an overpressured formation may cause a well to flow to the surface. Overpressure thus forms a hazard to drilling operations if not anticipated in well planning.
The purpose of this study was to develop an improved interpretation and delineation of pore pressure in the frontier Chukchi shelf region. The Chukchi shelf is located offshore northwestern Alaska in water depths typically 50 meters (165 ft) or less. Five exploration wells were drilled on the Chukchi shelf during open water seasons from 1989-91 and an array of drilling performance and petrophysical data were acquired. In this study, resistivity, conductivity, sonic travel time, sonic porosity, and drilling exponent data are used to evaluate formation pore pressure. Normal compaction trendlines are established using depth-dependent relationships then integrated with Eaton and adapted Eaton equations to calculate formation pressure. Four of the five wells penetrated overpressure and hydrocarbon generation is suggested as the driving mechanism behind the origin and maintenance of formation overpressure in the Chukchi shelf region.