The high costs associated with drilling operations, especially in offshore environments, make it necessary to optimize each step of the job. The drilling mud plays a critical role in the drilling operation, as it is responsible for several functions (solids transport, wellbore mechanical stability, signal sensor transmission, etc). Besides that, the attention given to the control of drilling mud properties has not changed much from the last 50 years. Downhole and surface online measurements for several drilling parameters are available for anticipated diagnostics of operational problems. Drilling automation is already a reality (automated pipe handling, controlled tripping operations, etc), but the drilling mud properties continues to be measured manually. The drilling mud sample has to be collected, transported, treated and analyzed to only then the measured property is reported. This work aims to present the development and result of experiments performed at a large scale drilling fluid loop, aiming the evaluation of commercial and built in property sensors. The following properties were determined: rheological parameters, mud weight, water-oil content, emulsion electrical stability (for oil based muds), fluid conductivity (for water based muds) and particle size distribution. The use of neural networks (Multi-Layer Perceptron type) allows the connection of the on-line equipment results to increase the reliability of mud properties determination. Comparisons with the results obtained from laboratory equipment were performed to train the neural networks as well validate the developed techniques.
Number of Pages
Looking for more?
Some of the OnePetro partner societies have developed subject- specific wikis that may help.