The accurate characterization of drilling fluid rheological properties at the in-situ downhole temperature and pressure is essential for designing the hydraulic program as well as for managing potential challenges during the drilling operation. A rheometer capable of operating at high temperature and high pressure (HTHP) is the common experimental method used to measure drilling fluid rheological properties at in-situ conditions. This method is costly and time consuming and requires skilled fluids laboratory personnel to carry out the tests. At the rigsite these devices are usually not available.
This paper/study presents an analytics based method to estimate drilling fluid rheological properties at in-situ downhole temperature and pressure and provide the required inputs for hydraulic modeling during well planning, for real-time monitoring and automation. The performance of the developed method is evaluated by utilizing data obtained from both, Mud Check tests and HTHP Viscosity measurements of synthetic and oil based drilling fluid samples.
Drilling fluid samples were collected from the rigs in operations around the world. An algorithm is developed to retrieve data out of unstructured field service laboratory databases and to uncover hidden patterns by utilizing the text pattern matching, text analytics, table-based approach, data visualization and classification techniques. Several computational intelligence techniques and statistical methods are applied to interpret the data by correlation of variables measured in the Mud Check tests and HTHP rheometer data. Of the available data 70 percent have been randomly extracted to train the data driven-model to predict drilling fluids rheological behavior at the in-situ downhole temperature and pressure.
The model is further tested with the remaining 30 percent of the available data to confirm that the model can not only fit the data used for training but also characterize the drilling fluid rheology at the in-situ downhole temperature and pressure with the required accuracy. This study indicates that in addition to temperature and pressure, several parameters including density, salt concentration, low and high gravity solid volume percentage, oil-water ratio, and oil phase volume percentage should be considered to accurately predict the rheological properties of drilling fluids at downhole conditions. Since the method presented here depends on several variables measured in the Mud Check tests, it can effectively be employed independent of location or formation.
The developed method can be utilized at the rigsite by field personnel to estimate downhole rheological properties of drilling fluids with a small error margin, without using expensive test equipment and time consuming procedures.