Training

Data Analytics Workflows for Artificial Lift, Production and Facility Engineers

• 2 days / 16 hours (classroom or in person)
• Four virtual half-day sessions
• 1 day or two half-day virtual sessions (shortened option)
• Level: Intermediate

Synopsis

Business impact: the main aim is to provide insight and understanding of data analytics and machine learning principles through applications. Field data is used to explain data-analysis workflows. Using easy-to-follow solution scripts, participants assess and extract value from the data sets. The hands-on solution approach gives them confidence to try out applicable techniques on data from their field assets.

Data analysis means cleaning, inspecting, transforming and modeling data with the goal of discovering new, useful information and supporting decision-making. In this hands-on course, participants learn data analysis and data science techniques and workflows applied to petroleum production (specifically artificial lift) while reviewing code and practicing. The focus is on developing data-driven models while keeping close to the underlying oil and gas production principles.

Specifically, the following use cases are discussed, covering their business impact, code walkthroughs and solutions:

• Gas-lift optimization: single-point gas-lift injection for gas wells in tight formation using simulated data
• Choke flow rate estimation for high-volume wells using an offshore dataset
• Rod pump diagnosis (card classification) using onshore field data
• Flow pattern prediction and slug catcher design using a laboratory dataset
• Liquid loading in gas wells using a laboratory dataset
• Virtual flow meter using an ESP dataset
• Multiphase flow meter prediction using a three-phase measured dataset
• Gas-lift multi-well optimization: discussion only

Learning outcomes

• After completing the course, participants will have a set of tools and some pathways to model and analyze their data in the cloud, find trends, and develop data-driven models.

Who should attend

This Intermediate level course is primarily intended for artificial lift, production and facilities engineers and students to enhance their knowledge base, increase technology awareness, and improve facility with different data analysis techniques applied to large data sets.

Prerequisites

• Understanding of petroleum production concepts
• Knowledge of Python is not a must but is preferred to get the full benefit
• The Google Colaboratory environment, available in Google Cloud, is used for hands-on exercises
• Trainees need to bring a computer with a Google Chrome browser and a Google email account (available for free)

Course outline

1. Digital Oil Field Data Explorations/Workflows
   1. Digital Transformation and Oilfields
   2. Key technologies for digital oilfields
   3. Oilfield System Data Verification and Management
2. A Brief/Incomplete Primer on ML/AI
   1. Data Science versus Data Analytics
   2. AI, ML and Deep Learning
   3. Data Analytics Lifecycle
   4. Bias-Variance-Complexity Tradeoff
   5. Data Preparation
   6. Model Types
   7. Role of Domain Knowledge
   8. Training and Evaluating Model
   9. Toolsets
3. System Setup & Checks
   1. Google CoLab: why do we need it?
   2. Pull datasets and codebase from the GitHub repository
4. Data Workflows & Best Practices in Exploratory Data Analysis
   1. Data types in the production domain: streaming (real-time or time-series) vs. static (non-streaming)
   2. Data processing challenges
   3. Data basics: cleaning, filtration and regulation
   4. Best practices in exploratory data analysis
5. Rod Pump Dynamometer Card Classification (brief description of the data set/problem use case and expected outcome)
   1. The problem, input and output variables definition (SPE paper)
   2. Data set
   3. Hands-on exercise: model development and testing
6. Flow Pattern Prediction (brief description of the data set/problem use case and expected outcome)
   1. Problem, input and output variables definition
   2. Data set
   3. Hands-on exercise: model development and testing
7. Gas Lift Slugging (brief description of the data set/problem use case and expected outcome)
   1. Problem, input and output variables
   2. Hands-on exercise: regression solution
8. Choke Flow Rate Study (brief description of the data set/problem use case and expected outcome)
   1. Problem, input and output variables
   2. Hands-on exercise: multiple ML models and comparison
9. Multiphase Flow Meter (brief description of the data set/problem use case and expected outcome)
   1. Problem, input and output variables (SPE paper)
   2. Hands-on exercise: multiple ML models and comparison
10. Virtual Flow Meter with ESP Data Set (time permitting)
    1. The problem, dataset: inputs/outputs
    2. Two or three ML solutions

Formats and customization

• Level: Intermediate. Two days (16 hours), in person or virtual.
• The course content is for two days in the classroom or four virtual half-day sessions. The training can also be presented as a 1-day or two half-day virtual sessions.
• Training method: the course can be taught virtually or in the classroom. It discusses several business use cases that are amenable to data-driven workflows. For each use case, the instructor shows the solution using a data analysis technique with Python code deployed in Google Cloud. Trainees solve a problem and tweak their solution.
• The course can be customized to focus on specific artificial lift methods, e.g., gas-lift or ESP only.
• Client dataset-based examples are optionally incorporated in the class discussions. This option requires discussions with the client about the problem, two days of consulting effort, and access to the client dataset at least 4 weeks before the class.
• Software: Python code in Google Colaboratory (Google Cloud); datasets and code pulled from a GitHub repository.

Request this course in-house