Automated Real time Data Cleansing and Summarization; Case Study on Drilling Hook load Real time Data

Abstract

Abstract Real time data has become a vital component in drilling operations since it provides details that assist understanding of operational behavior. Real time data has played a big role in helping engineers maintain an optimized operational envelop to avoid drilling problems. It is well known that the amounts of data generated from real time tools (downhole and surface) are enormous; they bring other challenges such as managing and interpreting the high volume of data. The real time data is full of noise and it's difficult to process and analyze in real time. Cleansing and summarizing this huge volume of real time data is needed to improve the quality of the data for analysis and decision making in real time. In this paper, hook load data will be a case study for automated real time data cleansing and summarization using statistical analysis techniques. This paper will present a technique for converting the drilling hook load real time data from 2D representation to 3D representation; the third dimension will be the data concentration, and this dimension will play a major role in allowing the dataset to self clean and summarize, which will provide an automated manner to clean the drilling real time data and summarize it. This paper will show a set of real time data before and after applying this technique, and it will highlight how this technique was able to successfully clean and summarize a huge amount of real time data without losing trend descriptiveness. This technique can prepare the real time data to be processed by a new level of advanced analytics applications that provide the drilling operation with real time prediction, simulation, and decision making solutions.