Augmented Artificial Intelligence Improves Data Analytics in Heavy-Oil Reservoirs

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

Recently, many heavy-oil fields have seen exponentially higher volumes of data made available as a result of omnipresent connectivity. Existing data platforms have focused traditionally on solving the problem of data storage and access. The more-complex problem of true knowledge discovery and systematic value creation from the massive amount of data is less frequently addressed. The authors of this paper propose a novel work flow for the problem of building intelligent data analytics in heavy-oil fields.


Optimal reservoir management for heavy-oil reservoirs requires systematic solutions that combine both engineering ability and advanced analytics. The authors believe that this requirement is addressed by what they call augmented artificial intelligence (AAI), a process inspired by the intelligence-amplification concept in which machine learning and human expertise are combined to improve solutions derived by systems that learn without any type of input from engineers or geoscientists. Practical deployment of AAI will involve automated work flows that use solid technical expertise and proven processes to transform field data into more-effective reservoir-management solutions.

Even with rapid data-preprocessing solutions in place, developing an optimal reservoir-management framework for heavy-oil assets is inherently complex. Identifying key recovery obstacles (KROs) and field-development plans (FDPs) typically takes many months, involving a large team of experts and the construction of sophisticated full-field simulation models. The recommendation is that automated work flows and AAI solutions are combined to identify those KROs rapidly and prepare robust FDPs that increase production and optimize current operations.

Perhaps the less-intuitive step in developing systematic solutions for heavy-oil fields is the process of developing a quantitative reservoir diagnostic framework. This process must build from big-data analytics platforms and an array of analytical, numerical, and empirical models combined to deliver a catalog of KROs affecting field performance. To this end, the entire historical set of well, field, and reservoir data must be processed and input into this diagnostics platform. Once the KROs are understood, the next step is to translate the diagnostics into detailed action plans in the field that can generate production, reserves, or capital-efficiency improvements.

This paper aims to offer an alternative approach to traditional work flows that identify recovery obstacles and development opportunities in heavy-oil fields by labor-intensive solutions. In contrast, the authors propose a systematic framework that provides three key advantages:

  • Execution time is fast, and an initial opportunity inventory can be generated.
  • The user can choose from multiple algorithms and methods to customize the technology to unique field/reservoir complexities.
  • The core algorithms are data-driven, integrate multidisciplinary data sets, and leave little room for the biases of the user, which allows for a consistent and repeatable analysis.


The proposed work flow moves from field data to smart analytics and then from analytics to diagnostics and opportunity identification. The first step in this work flow is to generate a valid process for fast, multidisciplinary data extraction. Once the field data are properly stored and formatted, a systematic analytical process can be implemented across all disciplines. From that point, KROs are obtained that illuminate immediately the suboptimal points in the reservoir-management approach currently implemented in the field. Building from these KROs, a comprehensive list of field-development opportunities is proposed, which must be vetted by subject-matter experts before final delivery and deployment into an FDP.

A critical step in this work flow is the proper identification of the true KROs in the field, which needs to be performed with an unbiased, quantitative, ­engineering-centered diagnostics process. In fact, any mistake at this stage will lead unavoidably to suboptimal opportunities. As an example, a general diagnostics process for reservoir-performance diagnostics is summarized in Fig. 1. Diagnosis of heavy-oil fields is a subset of this general process. The analytics and diagnostics of each section covered by the figure are discussed in detail in the complete paper.

Fig. 1—General diagnostics process for reservoir performance.


The production-performance section in Fig. 1 captures performance at the field and well level. For the well-level analysis, this must include both producers and injectors and active and inactive wells. The well-level analysis lays the foundation for the diagnostics of the subsequent sections, which begin to explain the reservoir-based causes of observed trends. Many of the analytics contained in this category use well-by-well decline-curve analysis (DCA). The accuracy of these decline curves is strengthened greatly by use of an event-detection algorithm and quantile regression techniques.

The reservoir-contact section integrates completion, geologic, and petrophysical data to analyze all aspects of reservoir contact for all wells over the entire production history of the field. Statistics are generated for well-completion contact, measured in meters, for each zone and field area and analyzed by well type, completion type, artificial-lift type, and production/injection commingling among flow units.

The pressure and voidage section is intended to describe quantitatively the relationship between reservoir energy and production and injection quantities.

The sweep efficiency and fractional-flow section completes the analysis of reservoir fundamentals by integrating the findings of the previous sections into an analysis of fluid movement through the reservoir.

Building from all the analytics derived in this diagnostic process, systematic identification of field-development opportunities also can be addressed. The ability to identify optimum well opportunities (recompletions and new drills) is improved greatly if saturation maps can be estimated correctly. In the complete paper, the authors provide a ­machine-learning solution to address this problem.


The authors demonstrate the application of systematic and data-driven work flows to a large heavy-oil field in Latin America designated Field X (the true field name is not provided for confidentiality reasons). This field has been onstream for more than 30 years, with a total well count of more than 1,500. The field is operated currently with approximately 600 producers and 20 injectors (for water disposal). Well locations are shown in Fig. 2. The original oil in place attributed to this field is more than 4 billion STB, and the field sees a current recovery factor of approximately 10%. The field occupies an extension of more than 600 km2, with a reservoir depth varying between 600 and 700 m. Reservoir temperature is approximately 57.8°C. Oil viscosity ranges from 150 to 700 cp, and field porosity ranges from 27 to 33%. Rock permeability values of 1 to 20 darcies are observed. Initial reservoir pressure was approximately 1,200 psi, but current reservoir pressure is less than 1,000 psi for this field.

Fig. 2—Field X well locations with well trajectories. Disposal wells are represented as injectors.


From a reservoir-management perspective, this is a complex field. The very unfavorable oil-mobility ratio has led to historically high water-production rates, which makes the field marginally economical. Average oil production has declined to 192,000 STB/D. Production rate is being mitigated by drilling new wells, which is capital-intensive. Application of the authors’ automated DCA methodology for this field indicates a baseline decline rate of 20%, with an average well estimated ultimate recovery of 570,000 STB. From this analysis, the associated ultimate recovery in Field X will be 795 million STB by year 2035. Fundamentally, this ultimate recovery is negatively affected by excessive water production.

The problem of water production in this field is illustrated clearly in the analysis of shut-in wells caused by high water cut. On average, 90 wells have gone offline each year after 2012. Water/oil ratio (WOR) mapping identifies a strong east/west trend. The performance of new wells in both the low- and high-WOR area might be improved if different completion designs, coupled with proper production mapping, are used. Before that, the authors’ analysis shows the potential to save more than 1,680 B/D of water production (while deferring only 14 STB/D of oil production) with the existing well configurations.

The complete paper includes some of the analytics used to diagnose this heavy-oil field. The KROs are summarized as

  • High water production and rapid water breakthrough attributed to the aquifer; suboptimal water production management
  • Excessive reservoir and fluid complexity, complicating optimal reservoir-management policies
  • Uncertainty in reservoir characterization caused by limited available data

The main FDPs for this field were ­identified as

  • Identification of sweet-spot locations to drill horizontal wells for increasing recovery
  • Deployment of specific, state-of-the-art smart completions to reduce or delay water production
  • Reduction of water production by proposing cyclic production or shutting in high water producers
  • Alternative recovery methods such as polymer flooding

The total amount of time used for this study was a fraction of the amount needed for similar studies of heavy-oil fields. This savings was possible only after using the systematic and automated work flow proposed in the paper.

For a limited time, the complete paper SPE 193650 is free to SPE members.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 193650, “Augmented-Artificial-Intelligence Solutions for Heavy-Oil Reservoirs: Innovative Work Flows That Build From Smart Analytics, Machine Learning, and Expert-Based Systems,” by David Castineira, Xiang Zhai, and Hamed Darabi, Quantum Reservoir Impact Group, prepared for the 2018 SPE International Heavy Oil Conference and Exhibition, Kuwait City, Kuwait, 10–12 December. The paper has not been peer reviewed.

Augmented Artificial Intelligence Improves Data Analytics in Heavy-Oil Reservoirs

01 May 2019

Volume: 71 | Issue: 5



Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.