Appropriate selection of rheological models is important for hydraulic calculations of pressure loss prediction and hole cleaning efficiency of drilling fluids. Power law, Bingham-Plastic Herschel-Bulkley models are the conventional fluid models used in the oilfield. However, there are other models that have been proposed in literature which are under / or not utilized in the petroleum industry. The primary objective of this study is to recommend a rheological model that best-fits the rheological behaviour of xanthan gum based biopolymer drill-in fluids for hydraulic evaluations. Ten rheological models were evaluated in this study. These rheological models have been posed deterministically. Obviously this is unrealistic so these deterministic models are replaced by statistical models by adding an error (disturbance) term and making suitable assumptions about them. Rheological model parameters were estimated by least-square regression method. Models like Sisko and modified Sisko which are not conventional models in oil industry gave a good fit. Modified Sisko model which is a four parameter rheological model was selected as the best-fit model since it produced the least residual mean square. There is 95% certainty that the true best-fit curve lies within the confidence band of this function of interest.
The use of rheological models to approximate the behaviour of non-Newtonian fluids is very paramount in the oil and gas industry especially during drilling, well completion, workover and acidizing. In drilling operations, mathematically designed rheological models are used to describe the viscous forces to develop frictional pressure loss equations. Accurate prediction of pressure losses help in the determination of bit optimization hydraulics, estimation of equivalent circulating density (ECD) and drilling fluid compressibility. The benefits of a more accurate estimation of ECD is adequate hole cleaning efficiency to enhance total drilling rate which in turn reduces total drilling cost. Prevention of circulation loss, maintenance of under-balanced drilling conditions and detection of potential kick are achieved if ECD is rightfully predicted (Bailey and Peden, 2000). Estimated model parameters help to perform other hydraulics calculations.
Power Law and Bingham Plastic models are widely used for hydraulics evaluation. They are assumed for standard API hydraulics calculations. Herschel-Bulkley, Roberston-Stiff and Casson models have been accepted to some extent in the petroleum industry. These models and the corresponding hydraulic calculations do provide a way for fair estimates of hydraulics for conventional wells using simple drilling fluids as asserted by Guo and Hong in 2010. Power Law model predicts shear stress well at low shear rate (in the annulus) and Bingham Plastic model describes the characteristics of drilling fluid at high shear rate (in the drill pipe).
Biopolymer drill-in fluid is a complex fluid formulated with several compositions to desired properties for optimum performance particularly in unconventional wells. It is a water soluble ‘rheology engineered’ drilling fluid designed to optimize the performance of rotary drilling. It is a complex high molecular weight (MW) polymer with a strong bond between the chains of its molecules which is efficiently used in unconventional wells like onshore and offshore horizontal wells, coiled tubing drilling and slim holes. The elastic structures of biopolymers make them have a higher carrying capacity than the other polymers applied in the petroleum industry during drilling. Due to the complex nature of this type of fluid and its unusual behaviour, it is very prudent to use a more precise rheological model to characterize its behaviour over a full range of shear rate to achieve a proper hydraulics evaluation.
Drill-in fluids are specially designed fluid system for drilling through the reservoir interval of a wellbore. They are basically formulated to drill the reservoir zone successfully, often a long, horizontal drainhole, to minimize damage and optimize the production of the exposed zones and to enhance the well completion needed. It contains additives that can principally control filtration loss and facilitate optimum carrying capacity. Its composition may be brine with right aggregate size (salt crystals or calcium carbonates) and polymers (www.oilfield.slb.com). Brian et. al (1997) asserted that polymers typically used as drill-in fluids are xanthan gum, starch, cellulose and scleroglucans. Hemphill et al. in 1993 proposed that Herschel-Bulkley model which is a three-parameter model is more likely to approximate the non-Newtonian behaviour of polymeric fluids.
This study focuses on ten rheological models proposed in various literatures and come out with a statistical criterion to select the most likely model to predict the rheological characteristics of xanthan gum base biopolymer drill-in fluids.
1.2 PROBLEM DEFINITION
The use of rheological models in the characterization of the behaviour of non-Newtonian fluid aids in the evaluation of drilling fluid hydraulics. The Power Law and / or Bingham Plastic models are more often used in evaluating hydraulics of drilling fluid in the oilfield during drilling operations. These are used because their resultant flow equations are simple and it is also easier to estimate parameters of the models by explicit solutions. However, none of these is able to predict the behaviour of the fluids over the wide ranges of deformation rate during the circulating of drilling fluid system throughout the wellbore. The advent of computers makes it realistic to estimate parameters of more complicated models and thereby deriving expression of pressure drop as function of flow rate. Rig site computers are now readily available making the requirement for simple parameter estimation and easily manipulated flow functions redundant and provide a platform conducive to more rigorous analysis.
The objective of this work is to be able to identify the fluid flow model that is more likely to predict the behaviour of xanthan gum base biopolymer drill-in fluids and estimate the confidence interval of the selected model function in order to quantify the degree of certainty. The Power law, Bingham Plastics, Herschel-Bulkley, Roberston-Stiff, Modified- Roberston-Stiff, Prandtl – Eyring, Modified Prandtl – Eyring, Sisko, Modified Sisko and Casson models were used for the purpose of this study……