Flat rheology drilling fluids are synthetic-based fluids designed to provide better drilling performance with flat rheological properties for deep water and/or cold environments. The detailed mud properties are mainly measured in laboratories and are often measured twice a day in the field. This prevents real-time mud performance optimization and negatively affects the decisions. If the real-time estimation of mud properties, which affects decision-making in time, is absent, the ROP may slow down, and serious drilling problems and severe economic losses may take place. Consequently, it is important to evaluate the mud properties while drilling to capture the dynamics of mudflow. Unlike other mud properties, mud density (MD) and Marsh funnel viscosity (MFV) are frequently measured every 15–20 minutes in the field. The objective of this work is to predict the rheological properties of the flat rheology drilling fluids in real-time using machine learning (ML). A proposed approach is followed to firstly predict the viscometer readings at 300 and 600 RPM (R600 and R300) and then calculate the other mud properties using the existing equations in the literature. For forecasting the viscometer readings, the created model using the decision trees (DT) demonstrated good accuracy. The results revealed a maximum average absolute percentage error (AAPE) below 4.5% and a correlation coefficient (R) of greater than 0.97. The estimated rheological properties showed a good matching with the actual values with low errors.
Drilling fluid or mud is a mixture of a base fluid and additional ingredients in certain proportions used while drilling. Several materials are added to adjust the mud properties such as, but are not limited to, the weighting agents for density, the fluid loss control materials, and viscosifiers for controlling the rheological properties (e.g., plastic viscosity (PV), yield point (YP), and gel strength). Despite mud represents 5% to 15% of total drilling costs, it may cause most of drilling problems. Drilling fluids are put to even greater strain by high-angle wells, high temperatures, and lengthy horizontal sections across pay zones (Bloys et al., 1994). Newtonian and non-Newtonian are the main two types of fluids. Newtonian fluid is characterized by a constant viscosity at a certain temperature and pressure. Non-Newtonian fluid such as most drilling fluids and cement slurries has viscosities that rely on shear rates for certain pressure and temperature (Rabia, 2002). Drilling fluids are mainly classified as water-based mud (WBM) or oil-based mud (OBM). OBM typically contains a base oil representing the external continuous phase; a saline aqueous solution representing the internal phase, emulsifiers at the interface, and other additives for suspension, weighting materials, oil-wetting, fluid loss, and rheology control additives. Oil based drilling fluids have two main categories which are invert-emulsion and all-oil drilling fluids (Alsabaa et al., 2020). An invert emulsion mud contains about 50:50 to about 95:5 by volume oil to water ratio. An all-oil mud contains 100% oil; that is, there is no aqueous internal phase. The invert emulsion mud is used to tackle some drilling problems like shale instability, minimize damage to water zones, and, and protect the casing and tubing against corrosion (Gray and Grioni, 1969; Growcock et al., 1994). The invert emulsion mud is characterized by its low toxicity and the brine is added to control the salinity to prevent water molecules from invading the formations (Hossain and Al-Majed, 2015). The invert emulsion drilling fluid is mainly used to drill the HPHT wells owing to its thermal stability which outperforms the WBM and can be used in drilling up to 400 ℉ (Lee et al., 2012).