United States Department of Transportation
ROSA P serves as an archival repository of USDOT-published products including scientific findings, journal articles, guidelines, recommendations, or other information authored or co-authored by USDOT or funded partners.
As a repository, ROSA P retains documents in their original published format to ensure public access to scientific information.
i
Organization and Analysis of Measurement While Drilling (MWD) Data
-
2025-02-01
[PDF-10.11 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
DOI:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report (August 2022 – March 2025)
-
Corporate Publisher:
-
Abstract:The scope of the MDT funded research project highlighted collection and organization of data onto a portal, data review and quality control and analysis of relationships between MWD drilling parameters and rock properties. Our initial approach was investigation of traditional linear correlations between individual MWD drilling parameters and rock properties such as SPT blow count for hollow stem auger data and UCS or unit weight for rock core data. In addition to individual MWD data types (depth, rotation rate, rotation torque, down pressure and advance rate) we also included the calculated compound parameter specific energy. Based on weak, single parameter, linear correlation results using MWD data from multiple boreholes, we extended our correlation analysis to exponential fitting with no improvement in correlations. To further investigate correlations, we implemented a multiple linear regression (MLR) approach using all possible combinations of the six inputs. Correlation results improved for a number of combinations of inputs but still resulted in weak predictive models. Finally, because of poor linear correlation model predictive results, we turned to a nonlinear approach by implementing a feedforward neural network. The neural network (NN) approach investigated all combinations of MWD drilling parameters as inputs, used one hidden layer with varying numbers of neurons, and a single neuron output layer for predicting either SPT blow count, UCS or unit weight. Using a nonlinear approach greatly improved the predictive power of the MWD inputs for rock properties. Based on our data investigation and analysis results, we suggest that a viable MWD program adhere to a set of guidelines developed from experience and other researchers input that insure a consistent, repeatable drilling methodology with close attention to real-time data quality control.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
