Automated characterization of drilling fluid properties

Accurate measurement of drilling fluid properties is essential in order to optimize wellbore construction, and in particular to properly manage hydraulics. It becomes even more crucial during deepwater drilling when a narrow mud window is present which may require the use of more advanced drilling technologies such as Managed Pressure Drilling (MPD) and Dual Gradient Drilling (DGD). Operating these technologies properly requires the use of sophisticated hydraulic models that require accurate rheological information as input. However, a full mud check with determination of all relevant rheological parameters is usually only carried out once per day, and augmented with one or two partial checks in the 24-hour period. Such intermittent and unreliable measurements are unfortunately not sufficient to provide the required inputs for ‘real-time' hydraulic modeling and control. A more practical approach for a continuous, automated monitoring of the drilling fluid properties is therefore called for. The method used here is based on the pipe viscometer approach rather than the traditional rotational viscometer method. In addition to the fluid rheology, important inputs for hydraulic models, such as mud density, transition to turbulent flow (critical Reynolds number), and real-time friction factor for non-Newtonian drilling and completion fluids are also obtained using the pipe viscometer. A prototype of this equipment was constructed, tested, and fully automated at The University of Texas at Austin. The flow loop was tested with several weighted and unweighted mud systems. During the measurement process, the driving pump was ramped up and held intermittently at various flow rates to measure the laminar frictional pressure loss in the pipe section. The data thus obtained was analyzed by software that generated a flow curve and from it derived relevant mud rheological parameters using a suitable rheological model. It also proved possible to extend the test to the turbulent flow regime and obtain the ‘true' friction factor in real-time for each particular fluid, rather than relying on a limited number of correlations that quite often exhibit inaccurate results, particularly for the Yield Power Law (YPL) fluids. Several successful tests with different mud systems indicate the reliability and robustness of the proposed technique.