# Study on automatic lithology identification method while drilling based on acoustic pressure-rock physics parameters mapping

## TL;DR

This study introduces a new method for automatically identifying rock types during drilling using the relationship between acoustic pressure and rock properties, improving coal mine exploration.

## Contribution

A novel automatic lithology identification method based on acoustic pressure-rock physics mapping is proposed.

## Key findings

- The method achieved recognition accuracies of 47% to 71% for various rock types.
- Acoustic pressure is an effective feature for identifying drilling lithology.
- The algorithm remains effective even in the presence of a perforated transition zone.

## Abstract

The lithology identification while drilling is a critical component of intelligent coal mine exploration. Investigating automatic lithology identification methods is of great significance for enhancing reservoir prediction accuracy and the automation level of drilling exploration. This study proposes a novel method for automatic lithology identification while drilling based on the mapping relationship between acoustic pressure and rock physics parameters. First, core samples were collected from an operational mine borehole to prepare homogeneous (single lithology) and layered (composite lithology) rock specimens, providing reliable materials for drilling experiments. Second, a full-scale laboratory drilling system was designed and constructed, providing a robust dataset for time-frequency analysis with strong engineering applicability. Furthermore, a quantitative fitting model between acoustic pressure and rock physics parameters was constructed, and the physical mechanism between acoustic pressure and rock physics parameters was revealed. Finally, The mapping relationship between acoustic pressure and physics parameters was established, an automatic lithology identification algorithm was developed based on this mapping relationship. The results demonstrated that the acoustic pressure can be used as an effective response feature for identification of drilling lithology. The proposed method achieved recognition accuracies of 47%, 58%, 53%, 48%, 66%, and 71% for sandy mudstone, coal, mudstone, shale, limestone, and granite. The existence of the perforated transition zone does not affect the identification of lithology by the automatic identification algorithm. This research introduces a novel approach for lithology identification while drilling, which is pivotal for advancing the intelligent development of coal mine exploration.