Inline Setup Provides Automatic Measurement of PVT Behavior of Drilling Fluids

Abstract

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 204084, “Automatic Measurement of the Dependence on Pressure and Temperature of the Mass Density of Drilling Fluids,” by Eric Cayeux, SPE, NORCE, prepared for the 2021 SPE/IADC International Drilling Conference and Exhibition, originally scheduled to be held in Stavanger, 9–11 March. The paper has not been peer reviewed. The mass density of drilling fluids usually is measured manually with a mud balance. The pressure and temperature dependence of the mass density of the fluid [i.e. its pressure/volume/temperature (PVT) behavior] then is estimated. Variations in the composition of the fluid mix and uncertainties regarding the PVT behavior of each component, however, may lead to inaccuracies. An apparatus that measures the PVT behavior of the drilling fluid contained in a pit directly and automatically has been designed. Inline PVT Measurement The pressure and temperature dependence of drilling fluids can be described by a biquadratic function. However, API Recommended Practice 13D recommends using a linear function of temperature combined with a quadratic form of pressure. Because this process involves six parameters, at least six measurements must be made under different conditions of pressure and temperature. A starting point is to measure the mass density of the fluid under six different pairs of pressures and temperatures. To keep the design of the apparatus as simple as possible, it ideally would not operate under high-pressure and -temperature conditions. Therefore, knowing the range of pressures and temperatures sufficient for taking sample measurements is useful in order to be able to extrapolate the model at higher pressure and temperature conditions with acceptable accuracy. The densitometer’s measurement precision of 0.05 kg/m3 and repeatability of 0.01 kg/m3 is known, so stochastic simulations of the possible measurement error for various spans of investigated pressures and temperatures can be performed. In this study, the authors con-sider that the calibrated PVT model shall be used for a range of pressure of 1000 bar and a range of temperature of 200°C. It is possible to calculate the root mean square of the proportion error between the predicted density value and the “true” value when varying stochastically the systematic bias on the density measurement when the calibration samples are spanning small ranges of pressure and temperature. A possible design for an inline apparatus could be to pump the drilling fluid past a controllable heating element and having a controllable choke downstream of the densitometer apply a pressure while measuring the mass density. The setpoints for the heating element and the choke would be changed six times in order to collect the necessary mass densities to calibrate the PVT model. Changing the temperature of the heating element, however, can require several minutes, and gathering a complete set of calibration measurements may easily take 15 to 30 minutes. An alternative could be to perform six measurements simultaneously. The densitometers can be mounted in series. The configuration could be with six parallel branches or any combinations between series and parallel branches. With two parallel branches, in one branch the temperature of the fluid is not modified, while it is modified in the second branch. For each of the two branches, back pressure is applied at two intermediate positions. This configuration has the advantage of using fewer chokes and pressure sensors (four instead of five).