Prediction Models for Oilfield Production
In the 2015 Report on the Work of the Government, Chinese Premier Li Keqiang proposed the strategic direction for “Internet+” and the development of the Internet of Things (IoT) for all industries be led by the oil, gas, and petrochemical sector.
Data Mining Becomes Critical
Oil production has experienced tremendous benefits with smart digitization for extraction, pipeline, and station operations. Centralized management and control now stretches end-to-end with IoT-generated production data from wells, storage, pumping, transport stations, and processing plants, as well as equipment status information through the production command and control centers. IoT also creates additional information from the data captured between interconnected equipment.
Rapid IoT development in the oil and gas industry is propelling oilfields from “digital” to “intelligent.” According to Le Deren, Chinese Academy of Sciences and the Chinese Academy of Engineering member and worldwide expert in remote sensing, “While digitization can provide offsite personnel full visibility into an oilfield, intelligent oilfields can be managed remotely and controlled by engineers in real time from thousands of miles away.” The key to digital oilfields is data acquisition, in which data is collected quickly, summarized, and analyzed during oilfield exploration and development. Intelligent oilfields step up the game with predictability from data mining analysis.
Case 1: Use-Modeling
Oilfield production activities generate multiple streams of reference data every few seconds. Typical data payloads include: pump stroke length, pump speed, task duration, power consumption, fluid production, dynamic liquid level, and submerged depth.
The accumulation of historical data requires modeling and analysis to extract useful information. The use of IoT platforms for predictive optimization allows dynamic adjustments by examining real-time variations in wellhead productivity.
Case 2: Analytics Reduce Costs
The operating efficiency of wellheads directly affects oilfield yields. Maintenance and management of wells are important in production planning to detect, analyze, and resolve mechanical issues in a timely manner. Often, faults that are easily overlooked or hard to locate will have a direct impact on recovered volumes. Failing pumps must be stopped and repaired as quickly as possible to minimize downtime.
The resolution to many problems can be fast-tracked with the deployment of oilfield sensor networks. Common instrumentation includes load displacement, flow metering, differential pressure, and temperature measurements. Digital transducers enable IoT platforms to collect real-time data to render pumping unit diagram displays, wellhead productivity detail, water content, and other key production information. The pattern of changes in production data are recorded to establish a range of models to depict the normal and anomalous ranges for pump unit operations, sucker-rod vibration effects, pump liquid supply deficiencies, sand production, and heavy oil viscosity. These models provide a reference for well maintenance and management personnel to identify individual well problems, assess related events, and take immediate measures when the expected norms are exceeded. Data monitoring procedures are designed to trend for proactive and predictive maintenance, optimal pumping unit efficiency, long service lives, and, ultimately, maximum wellhead production.
Case 3: VR Training
Virtual Reality (VR) technology advances now include simulations of offshore drilling platforms. By means of three-dimensional (3D) displays of drilling platforms and interactive virtual peripherals, the simulation platform supports remote 3D inspection and emergency drilling operations, as well as staff training on complex equipment.
- 3D inspection: True-to-life inspection experiences use the real-time operating status of remote drilling platforms to graphically display live data for inspection personnel.
- Staff training: Staff members interact with simulated drilling platforms using visualization devices for improving the training experience, such as VR helmets, gloves, and physical touch screens.
- Emergency drills: By simulating accidents, such as leaks, fires, and explosions, operating staff can train and plan to handle actual emergencies and prevent further escalation.
Overview of an intelligent oilfield
Evolution to Intelligent Oilfields
Huawei’s IoT architecture is a four-layer [1+2+1] construction that fits well with the energy industry’s transition to intelligent oilfields.
- The first “1” refers to a central host platform with open connectivity for applications.
- The “2” refers to open network access, including wired and wireless.
- The final “1” represents LiteOS, the open IoT operating system.
The Huawei IoT solution equips oilfields to evolve using intelligent optimization from Big Data analytics.