Improved identification of pay zones in complex environments through wavelet analysis on nuclear magnetic resonance log data

Abstract

Identification of pay zones is as a challenging topic in the reservoir characterization. Various methods have been employed to tackle this prominent task. Well logs are one of the most useful means for identifying pay zones, chiefly resistivity logs. However, resistivity measurements in some complex environments such as mixed lithology reservoirs, low resistivity and low contrast pay zones may fail to unveil productive zones. As an essentially lithology-independent tool, nuclear magnetic resonance (NMR) log may be the only approach to deal with detecting such subtleties. Using wavelet analysis technique and the NMR log data, this paper was aimed at identifying hydrocarbon bearing zones in two carbonate reservoirs. Discrete wavelet transform (DWT) was applied to the spin echo train at each depth to extract transverse relaxation time (T2). By comparing the generated T2 log and resistivity log, a striking similarity was found between them. The T2 log manifested strong correlation with porosity log. Therefore, the DWT was repeatedly applied to the T2 log so as to remove the correlation. Various wavelets were adopted to remove the effect of porosity from the T2 log, leading to achieve pore fluid transverse relaxation time, referred to as T2f. Scrutinizing the T2f log demonstrated that this log is not only highly functional in detecting productive zones but also highly capable of highlighting the subtle changes of pay zones. Consequently, the T2f log revealed the latent pay zones unseen by resistivity log.