|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
The Abu Dhabi National Oil Company (ADNOC) announced that it has completed the first phase of its large-scale multiyear predictive maintenance project, which aims to maximize asset efficiency and integrity across its upstream and downstream operations. ADNOC says its predictive maintenance platform uses artificial intelligence (AI) technologies such as machine learning and digital twins, ADNOC’s to help predict equipment stoppages, reduce unplanned equipment maintenance and downtime, and increase reliability and safety. The company said it expects use of the platform to result in maintenance savings of up to 20%. The predictive maintenance project, which was announced in November 2019, is being implemented over four phases. ADNOC’s predictive maintenance project is part of the company’s digital acceleration program, which focuses on embedding advanced digital technologies across the company’s operations.
The Abu Dhabi National Oil Company (ADNOC) completed the first phase of its large-scale multiyear predictive maintenance project to improve asset efficiency and integrity across its upstream and downstream operations. Announced in November 2019, the project is being implemented over four phases as part of the company’s digital acceleration program to embed advanced digital technologies across its operations. Phase 1 covers the modeling and monitoring of 160 turbines, motors, centrifugal pumps, and compressors across six ADNOC Group companies. All phases of the project are expected to be completed by 2022 and will enable monitoring of up to 2,500 critical machines. Using artificial intelligence (AI) technologies including machine learning and digital twins, the company’s predictive maintenance platform helps with equipment stoppages, reduces unplanned equipment maintenance and downtime, increases reliability and safety, and is expected to deliver maintenance savings up to 20%.
The intent of this paper is to recognize and address the challenges that downstream companies face across their value chain including a history of having siloed businesses processes that have created ‘value leaks’. We propose an unified operations management approach that orchestrates all activities across the value chain as part of an enterprise digital transformation strategy. Business process can be deeply transformed when operations, supply chain and process optimization are connected in a collaborative environment.
Refineries are at different stages in their digital transformation journey and have built their business workflows over many years across multiple point solutions.
We propose a digital transformation strategy for operations across five key characteristics, that is:
Across a refinery's value chain there are significant benefits that can be realized by approaching digital transformation across their operations enabling end-to-end value chain optimization including:
Supply Chain ($20-150M/year):
Process Performance ($10-50M/year): Longer equipment life, increased availability, increased quality, increased yields
Blending and Oil Movements ($5-50M/year): Reduced giveaway, no rework, minimum inventory, minimum downgrades, higher fuels agility
Energy Management ($5-25M/year): Reduced energy conversion cost, reduced energy consumption cost, reduced cost of crude for energy
Production Management ($2-5M/year): Reduced accounting losses, reduced inventory, reduced hydrogen and steam consumption, increased throughput
Operations Management ($2-5M/year): Reduced unplanned shutdowns, increased yield, reduced hydrogen and steam consumption, increased throughput
This paper is novel as it takes a wholistic look at a refinery operations value chain and eliminating existing value leaks end-to-end. It recognizes how existing systems have been built-up on outdated technology and siloed business process and proposes a path forward bringing operations, supply chain and process optimization together as a key element of a digital transformation strategy. The paper also explores how AI & prescriptive models pave the future for optimization.
The initial hydrocarbon saturation has a major impact on field development planning and resource estimation. However, it is derived from indirect measurements from spatially distributed wells applying saturation height modelling based on uncertain parameters. The methodology presented here is deriving posterior parameter distributions by using Machine Learning in a Bayesian Framework honouring the petrophysical uncertainty in the field. The results are used for initialization and will be applied for forecasting under uncertainty.
To determine the initial hydrocarbon saturation, the Saturation Height Model (SHM) needs to be conditioned to the interpreted logs. 2500 geological realizations were generated to cover the interpreted ranges of porosity, permeability and saturations for 15 wells. For the SHM, twelve parameters and their ranges were introduced. Latin Hypercube Sampling was utilized to generate a training set for Machine Learning models (ML). The trained ML models were conditioned to the log derived saturation data. To ensure a field wide consistency of the models only parameter combinations honouring the interpreted saturation range for all wells were selected.
The presented method allows for consistent initialization and for rejection of parameters that do not fit to observed data. In our case study, the most significant observation concerns the posterior parameter distribution of the Free Water Level range which is narrowed down. Furthermore, the SHM parameters are proven independent; thus the resulting posterior parameter ranges for the SHM can be used for conditioning production data to models and subsequent hydrocarbon production forecasting. Additional observations can be made from the ML results such as the correlation between wells, this allows interpreting groups of wells that have a similar behaviour, favour the same combinations and potentially belong to the same compartment.
A workflow is introduced honouring the interpreted error of log derived porosity-permeability and saturation data. The workflow is based on Saturation Height Modelling (SHM) to generate a training set for Machine Learning. The SHM is conditioned to all wells in a Bayesian Framework to ensure that the posterior parameter distributions of the SHM are consistent and can be used for conditioning dynamic models to dynamic data and production forecasting.
A workflow is introduced honouring the interpreted error of log derived porosity-permeability and saturation data.
The workflow is based on Saturation Height Modelling (SHM) to generate a training set for Machine Learning.
The SHM is conditioned to all wells in a Bayesian Framework to ensure that the posterior parameter distributions of the SHM are consistent and can be used for conditioning dynamic models to dynamic data and production forecasting.
Technology gaps to harness source rock or shale, unconventional reservoirs in the Middle East and North Africa pose unique challenges. Carbonate reservoirs, supplying Ca2+ ions resulting in passivation of traditional magnesium-based water reactive alloy alloys and abundance of H2S/CO2 rich production fluids at high pressure high temperature downhole conditions cause unpredictable degradation of plugs. This leads to the technology gap, an economical prediction mechanism to assess the degree of dissolution for deployed tools, which has prevented rapid uptake of water reactive alloys in these markets. Here we present, materials and methods to design and manufacture water reactive (fully dissolvable) alloy plugs for multi-stage stimulation (MSS) AKA hydraulic fracturing for the Middle East and North Africa (MENA). The backbone of our water reactive plug is a self-articulating amorphous and bulk nanocrystalline alloy and/or partially or fully vitrified dissolvable material with artificial intelligence (AI), sensing in its DNA, with an ability to learn from the environment it is deployed in. Additionally, these materials will deploy articles that relay information from downhole to surface without conventional telemetry. Other embodiments of the technology building blocks encompass novel, multi-layered smart water reactive nano-materials, further enhancing tailored and timed dissolution.
Resilient supply chain management plays an essential role in the success of organizations in terms of operational performance, efficiency, and sustainability that in turn ensures business continuity. While effective supply chain is directly dependent on logistics, inventory control, and materials handling, it is also strongly interlinked with the utilization of new technologies that will enhance the complete supply chain cycle. This paper develops a framework that introduces three-dimensional (3D) printing in the oil and gas industry, taking into consideration the challenges that this technology is facing to penetrate the conservative oil and gas industry and realizing its potential benefits to the business.
This research work utilizes different purchasing models to cover a significant category of expenditures in an oil and gas company, which is providing an alternative cost effective option for sourcing high value and long lead old/obsolete spare parts from original equipment manufaturers (OEM)s overseas. Several frameworks used worldwide were employed, such as the Kraljic matrix & PESTLE (Political, Economic, Sociological, Technological, Legal and Environmental) analysis, to employ different manufacturing/supply technologies for different stock item classes. These tools provided a systematic pathway to perform market analysis of the existing local 3D printing market and potential for growth and competitiveness in upcoming years. This paper also presents a trial attempt to 3D scan and digitize 215 materials as a first step in the prepared framework guideline. This study works in line with recent ADNOC initiatives in promoting local workshops and In-Country Value (ICV), andUAE precautions taken to mitigate the impact of the COVID-19 pandemic in oil and gas plants.
The adaptive framework provided a roadmap to debottleneck the constraints that conventional supply chains faces with the critical supplies. The developed decision support matrices, employing semi-quantitave approaches, serves in achieving the business need of every supply in terms of cost, schedule, or combination of both. The framework provides a guideline on the utilization of 3D printing, while maintaining the highest levels of quality. A qualitative analysis showed that the introduction of 3D printing would increase the diversity and market competition leading to potential price reduction.
The utilization of 3D printing technology as part of a strategic supply chain process is not well developed. This study presents an effective framework that will increase the confidence level of oil and gas operators in utilizing this technology and benefit from its values.
Petrophysical Groups (PG) from cores and log facies (electrofacies) are two of the main outputs generated in the petrophysical domain. Those are particularly contributors to the highest degree of uncertainty during the reservoir modeling, with the subsequent high-impact in field development decisions. Detailed core analysis is the preferred main source of information to condition the PG, however, since there are generally more un-cored than cored wells, logs are the most frequently applied source of information to extrapolate petrophysical groups by the use of log facies. The approach of this investigation is to apply machine learning to move from the core domain to the log domain and to determine relationships to generate properties for the three-dimensional reservoir model.
All wells have a full set of logs (Gamma Ray, Resistivity, Density, PEF and Neutron) and few have routine core analysis (Permeability, Porosity and MICP). On a first pass, logs from selected wells are classified into Self Organizing Maps (SOM) without analytical supervision. The idea of this step is to observe the capability of the log data to self-classify according to the log response. In a second stage, core data is used to define petrophysical groups (PG) by clustering the data in into similar pore throats from the MICP and permeability vs. porosity relationships, also thin section were used to introduce geological meaning to the to the PG. The PG are linked to the NMR pore-size distribution (Invasion correction were applied in order to observe only the pore geometry content on the T2 distribution) by correlating into a pre-determined standard NMR pore geometry distribution, the result is that each PG group has a particular T2 range and proportion. In the next step a pseudo NMR curve is create by the use of a Neural Network constrained by the range and proportions obtained and observed in cored wells. This pseudo NMR can be generated in all wells (even the ones Un-cored or without NMR) and generate PG groups at log scale hence the PG logs are controlled by the relationship between core (Pore throat, permeability, porosity and Geological Facies) and pore-size distribution coming from the NMR Log.
The result is a robust model capable to capture the core relationships and able to predict properties preserving the geological features of the reservoir. The application of this method makes possible to determine the minimum and most relevant source of data to address the issues caused by the heterogeneity of the rock. The applied workflows also show how to break the autocorrelation of variables and maximize the usage of logs.
This work demonstrates that the introduced data-driven methods are useful for rock typing determination and address several of the challenges related to core to log properties derivation.
Mohan, Richard (ADNOC HQ) | Hussein, Ahmad (Schlumberger) | Mawlod, Arwa (ADNOC Onshore) | Al Jaberi, Bashaer (ADNOC Onshore) | Vesselinov, Velizar (Schlumberger) | Abdul Salam, Fouad (ADNOC Onshore) | Al Hadidy, Khaled (ADNOC Onshore) | Pal, Anik (Schlumberger) | Al Yazeedi, Haifa (ADNOC HQ) | Al Daghar, Khadija (ADNOC Onshore) | Mustapha, Hussein (Schlumberger) | Razouki, Ali (Schlumberger) | Al Hamlawi, Imad (ADNOC HQ) | El Yossef, Bassem (ADNOC HQ)
ADNOC is continuously enhancing its capabilities to manage its oil and fields efficiently by better planning, execution and operations that drives field development decisions, well performance, and safe operations. In this regard, ADNOC envisages to leverage the evolving Oil and Gas 4.0 technologies to enhance the well planning decisions of the sub-surface and drilling team through data-driven and AI methods.
Effective well planning and operations require collaboration between different subsurface teams and drilling team leveraging multidisciplinary data, historical events and risks and constructing integrated drilling and sub-surface model for collaborative planning and keeping the model live. This requires having a live sub-surface model that is kept close to the field reality while reducing uncertainties. However, extracting key learnings, knowledge and experience from a variety of sources and reports is intense and requires lot of manual processing of data.
An AI-based solution leveraging data analytics, natural language processing and machine learning algorithms is developed to automatically extract knowledge from a variety of data sources and unstructured data in building a live intelligent model that enables effective well planning, predicting operational hazards and plan mitigation. The solution systematically extracts, collects, validates, integrates, and processes a variety of data in different formats such as well trajectory, completion, historical events, risk offset well information, petrophysical data, geo-mechanical data, and technical reports. Newly acquired data comprising drilling events, geological and reservoir properties are integrated continuously to keep the model live and digital representation.
Almadhoun, Wael (ADNOC Offshore) | Al-Shimmari, Nasser (ADNOC Offshore) | AL-Qamzi, Abdulla (ADNOC Offshore) | Al Owaid, Hatim (ADNOC Offshore) | Al-Shamsi, Ali (ADNOC Offshore) | Al-Naqbi, Maha (ADNOC Offshore) | Ateya, Ali (ADNOC Offshore)
In ADNOC Oil and Gas 4.0 mission, we are committed to empower people with the needed capabilities and Artificial Intelligence (AI) technologies to fuel innovation, efficiency and more importantly achieve and sustain a 100% HSE, by transforming the way of handling HSE events by moving from reactive to proactive approach. The ultimate objective is to save lives, empower the vessel Captains to immediately identify and respond to violators, improve the HSE culture of the crew, and automatically generate live data analytics and statistics with the aim of improving safety in operations. The implemented AI use cases are; deviation for not wearing Protective Safety equipment in designated areas, violation of not utilizing safety passages, alert when no watchman in muster station, alarm when man overboard incident, alarm when man fell on stairs, and live Personnel on board each weather-deck. When introduced the Artificial Intelligence cameras, our marine vessels will adopt a smarter automated response and reporting culture, which will in turn, lead to increased safety oversight of our critical offshore operations. Therefore, with the advent of the AI technology, many common business processes have been automated thus enabling personnel to increase their focus on more important tasks while technologies like the AI System can handle many of the time consuming tasks.
The solution components consists of Artificial Intelligence platform, high definition cameras, local server, wide-range WiFi access point, network infrastructure and a tablet. On the tablet device, the captain have full coverage of the vessel weather decks, working areas and restricted zones with a feature to generate alerts when detecting an emergency situation. This was provided to empower the vessel Captain to acknowledge and respond to violations as well as take a proactive action to prevent incidents from happening. The Machine Learning algorithm has been trained on actual scenarios and will be continuously improved by adding more recorded event to retrain the initial model. Currently, the prediction model is performing on the vessel operation mode and recording events with high rate of accuracy. In case of automatically detecting an alerting or non-compliance event, the captain would be notified, beacon lights and sound, and log recorded in the local and central system with a photo and a short video clip of the incident. The process of identifying HSE deviations are becoming digitally transformed by deploying AI capabilities on real-time video streams. The AI-based camera system leverages Computer Vision features that enables machines to get and analyze visual information and take action. The whole process of identifying HSE violation events has been digitally transformed by deploying an artificial intelligence solution to perform real time video analytics.
Unique within the ADNOC Group, Real Time Optimization (RTO) has transformed the operation of NGL trains by optimising all 12 NGL trains of one of ADNOC Gas Processing gas facilities. This enables flexibility such as selectively diverting feed to high yield plants.
The objective of the RTO is to achieve optimal distribution of feed gas and condensate RVP across various processing units of the facility in order to maximize the overall C2 and C2+ recovery, minimize energy consumption, reduce emissions and reduce overall operating cost.
Successfully proven, this is a good case for rolling out to similar complexes. Initiatives are now being taken up for other ADNOC Gas Processing sites and a ONE ADNOC RTO.