The Practical Challenges of Managing and Processing Measuring-While-Drilling (MWD) Data in Real Life Situations

This article deals with processing of parameters during the execution of fissure grouting works for ground improvement. Measuring-while-drilling (MWD), whilst not a new technique, can currently benefit from the potential of data-driven methods and software that can support faster analysis of the MWD data. But challenges remain in: defining the rules for the initial cleaning of data (to remove non-relevant data points caused by the drilling process) as there are no clearly defined internation rules for this operation, the practicalities of combining different type of data, e.g drilling and grouting data, and the initial exploratory analysis of the data to choose the most relevant characteristics or parameters for that specific ground. Additionally, although the current machine learning techniques have the potential to turn MWD into planning tools, the early stages of its use in MWD means that it is unclear what techniques actually work and / or several have to be tested and reviewed for usefulness. This paper presents a site example of the challenges above, attempting to bring to the forefront the basic work that needs to be done but it´s not often reported or mentioned in passing in technical documents.