Autonomous Drilling Fluid Management System - Development of Fluid Advisory System and First Lab Trial
-
Published:2023-10-09
Issue:
Volume:
Page:
-
ISSN:
-
Container-title:Day 2 Tue, October 17, 2023
-
language:
-
Short-container-title:
Author:
Ettehadi Reza1, Onegova Elizaveta1, Fevang Finn Øivind2, Knizhnik Alexey3, Postovalov Sergey3, Brevik Jan Ove4, Thompson Jr Charles1, Egorenkova Tatyana1, Kaageson-Loe Nils1
Affiliation:
1. Baker Hughes, Houston, TX, USA 2. Baker Hughes, Tananger, Norway 3. Baker Hughes, Celle, Germany 4. Equinor, Bergen, Norway
Abstract
Abstract
This paper details a study to design a Fluid Advisory System which uses various algorithms and methods to calculate an optimal quantity of chemical additives which preserve the desired characteristics of the drilling fluid based on manual and real-time measurements while taking inventory into account. The algorithms and results of the first lab trial are presented for water-based mud.
The Fluid Advisory System methodology consists of a data pipeline for loading historical data into a master table, a training pipeline for creating a fluid property model (digital twin) using Machine Learning (ML); an automated predictive analytics tool for model selection; and incorporating the predictive models into a prescriptive model that determines quantity of additives to use to achieve a desired set of drilling fluid properties. Given the initial product composition, a set of desired and current fluid properties, the prescriptive model suggests necessary modifications in additive concentrations. This is achieved by setting and solving an optimization problem.
First, the Autonomous Fluid Management algorithms have been validated in small-scale lab experiments. Following these successful experiments, a yard-scale trial was conducted at the Equinor Sandsli Automatic Drilling Fluid Laboratory. The yard trial demonstrated how remote control of drilling fluid treatment equipment and the Fluid Advisory System can be employed. The results of the tests reveal that it is now feasible to remotely regulate drilling fluid mixing and apply the ML-based Fluid Advisory System to maintain desired drilling fluid properties for weight, rheology, API fluid loss, pH, hardness, and chlorides.
The Fluid Advisory System was developed to aid in the optimization of the real-time drilling fluids treatment process for usage in an Integrated Operations Level 3 environment (IO level 3) and higher (IO level). This takes a step towards autonomous drilling fluid management that will allow for reduction of personnel needed to monitor and maintain a drilling fluid system; more consistent fluid performance and properties; more efficient rig supply chain; lower NPT; safer operations; lower operational cost; and lower carbon footprint.
To our knowledge, this is the first time that a Fluid Advisory System for maintaining multiple properties of a drilling fluid was developed and validated experimentally in a remote-controlled mixing facility.
Reference11 articles.
1. Ettehadi, Reza, May, Roland, Dahl, Thomas, Clapper, Dennis, and RosaSwartwout. "In-Situ Fluid Rheological Behavior Characterization Using Data Analytics Techniques." Paper presented at theIADC/SPE Drilling Conference and Exhibition, Fort Worth, Texas, USA, March2018. doi: https://doi.org/10.2118/189648-MS 2. Artificial intelligence techniques and their applications in drilling fluid engineering: A review;Okorie;Journal of Petroleum Science and Engineering,2018 3. Ivan
Feng
, SercanGul, DongmeiChen, and Ericvan Oort. 2022. Model Predictive Control for Automated Drilling Fluid Maintenance. Presented atIADC/SPE International Drilling Conference and Exhibition, Galveston, Texas, 8-10 March 2022. IADC/SPE-208769-MS. https://doi.org/10.2118/208769-MS. 4. Mehrdad G.
Shirangi
, RezaEttehadi, CharlesThompsonJr, EdwardFurlong, BakerHughes. 2022. Digital Twins for Automated Treatment of Drilling Fluids. Presented atAADE Fluids Technical Conference and Exhibition, Houston, Texas, April 19-20, 2022. AADE-22-FTCE-025. 5. Application Of Artificial Intelligence Methods In Drilling System Design And Operations: A Review Of The State Of The Art;Bello;Journal of Artificial Intelligence and Soft Computing Research,2015
|
|