Manuscript Number : SHISRRJ236913
Big Data-Driven Framework for Predicting Crude Quality Variations Across Distributed Offshore Production Lines
Authors(4) :-Andrew Tochukwu Ofoedu, Joshua Emeka Ozor, Oludayo Sofoluwe, Dazok Donald Jambol Crude oil quality variations present a significant challenge in offshore production environments, especially across distributed production lines operating under diverse reservoir conditions and flow regimes. Inconsistent crude properties such as API gravity, sulfur content, water cut, and viscosity directly impact downstream processing efficiency, transportation logistics, and product valuation. Traditional monitoring methods rely heavily on manual sampling and periodic laboratory analysis, which are inadequate for real-time quality assurance and proactive decision-making. To address this challenge, this study proposes a big data-driven framework for predicting crude quality variations across distributed offshore production lines using advanced analytics and machine learning techniques. The proposed framework integrates heterogeneous data sources, including wellhead sensors, flow meters, PVT (pressure-volume-temperature) data, and historical quality measurements, into a centralized data lake architecture. Data preprocessing modules manage noise reduction, normalization, and imputation of missing values. Predictive models built using supervised machine learning algorithms such as gradient boosting machines, long short-term memory (LSTM) networks, and ensemble learning are trained to forecast key crude quality parameters in near real time. Feature engineering incorporates geological, operational, and environmental factors to enhance model accuracy. A key component of the framework is its scalability and adaptability to multiple production units operating under varied offshore conditions. Model performance is validated using real and synthetic datasets derived from offshore assets, demonstrating high accuracy in predicting short- and medium-term quality trends. Additionally, the system offers visual dashboards and alert mechanisms for production engineers, enabling timely adjustments to blending strategies, flow allocation, and equipment settings. This big data-driven predictive approach not only enhances operational awareness but also supports economic optimization and environmental compliance. It lays the groundwork for intelligent, quality-centric production planning in offshore oil and gas operations, where real-time adaptability is increasingly critical to competitive and sustainable performance.
Andrew Tochukwu Ofoedu Big data-driven, Framework, Predicting, Crude quality, Variations, Offshore production lines Publication Details Published in : Volume 6 | Issue 3 | May-June 2023 Article Preview
Shell Nigeria Exploration and Production Company, Nigeria
Joshua Emeka Ozor
First Hydrocarbon, Nigeria
Oludayo Sofoluwe
TotalEnergies Nigeria
Dazok Donald Jambol
D-Well Engineering Nig. Ltd. (Shell Petroleum Development Company of Nigeria), Nigeria
Date of Publication : 2023-06-30
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Page(s) : 136-155
Manuscript Number : SHISRRJ236913
Publisher : Shauryam Research Institute
URL : https://shisrrj.com/SHISRRJ236913