Prediction of sonic velocities from other borehole data: An example from the Kevitsa mine site, northern Finland

Abstract

P‐wave and S‐wave velocities are vital parameters for the processing of seismic data and may be useful for geotechnical studies used in mine planning if such data were collected more often. Seismic velocity data from boreholes increase the robustness and accuracy of the images obtained by relatively costly seismic surface reflection surveys. However, sonic logs are rarely acquired in boreholes in‐and‐near base metal and precious metal mineral deposits until a seismic survey is planned, and only a few new holes are typically logged because the many hundreds of holes previously drilled are no longer accessible. If there are any pre‐existing petrophysical log data, then the data are likely to consist of density, magnetic susceptibility, resistivity and natural gamma logs. Thus, it would be of great benefit to be able to predict the velocities from other data that is more readily available. In this work, we utilize fuzzy c‐means clustering to build a “fuzzy” relationship between sonic velocities and other petrophysical borehole data to predict P‐wave and S‐wave velocity. If boreholes with sonic data intersect most of the important geological units in the area of interest, then the cluster model developed may be applied to other boreholes that do not have sonic data, but do have other petrophysical data to be used for predicting the sonic logs. These predicted sonic logs may then be used to create a three‐dimensional volume of velocity with greater detail than would otherwise be created by the interpolation of measured sonic data from sparsely located holes. Our methodology was tested on a dataset from the Kevitsa Ni‐Cu‐PGE deposit in northern Finland. The dataset includes five boreholes with wireline logs of P‐wave velocity, S‐wave velocity, density, natural gamma, magnetic susceptibility and resistivity that were used for cluster analysis. The best combination of input data for the training section was chosen by trial and error, but differences in the misfit between the various training datasets were not particularly significant. Our results show that the fuzzy c‐means method can predict sonic velocities from other borehole data very well, and the fuzzy c‐means method works better than using multiple linear‐regression fitting. The predicted P‐wave velocity data are of sufficient quality to robustly add low‐frequency information for seismic impedance inversion and should provide better velocity models for accurate depth conversion of seismic reflection data.

Predicting Missing Seismic Velocity Values Using Self-Organizing Maps to Aid the Interpretation of Seismic Reflection Data from the Kevitsa Ni-Cu-PGE Deposit in Northern Finland

“…We used SOM to predict missing seismic velocity values in order to see how SOM can handle the missing data that can be crucial for seismic interpretation. Kieu et al [8] have predicted seismic velocities (Vp and Vs) for the Kevitsa borehole data using fuzzy c-means clustering, and they compared their results with real data and predictions from multiple linear regression fitting. In their study, petrophysical data (P-and S-wave velocities, density, natural gamma, magnetic susceptibility, and resistivity) were used from five boreholes, and these boreholes all contained measurements for all the selected variables [8].…”

“…SOM is used for several functions, such as prediction, clustering, Minerals 2019, 9, 529 8 of 16 pattern recognition, and classification, operating on large volumes of data. SOM analysis has been widely used for different applications in different research fields [30][31][32], and recently it has been also applied to hard rock mineral exploration (e.g., [3][4][5][6][7][8][9][10]). Commonwealth Scientific and Industrial Research Organization (CSIRO)'s self-organizing map software (SiroSOM; [33]), which is based on a Matlab SOM Toolbox [34,35], is employed in this study.…”

“…Table 2. Different sets of parameters used for different SOM runs (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18) to predict missing seismic velocity values for boreholes KV171, KV173, and KV156. Correlation coefficients between actual measured and predicted SOM logs are shown for both Vp (CC Vp ) and Vs (CC Vs ).…”

“…For test boreholes, for which measured seismic velocity logs are also available, the correlation between actual measured and predicted velocities is strong to moderate, depending on the parameters included in the SOM analysis. Predicted reflectivity logs, based on measured densities and predicted velocities, show that some contacts between olivine pyroxenite/olivine websterite-dominant host rocks of the Kevitsa disseminated sulfide mineralization-and metaperidotite-earlier extensively used "lithology" label that essentially describes various degrees of alteration of different olivine pyroxenite variants-are reflective, and thus, alteration can potentially cause reflectivity within the Kevitsa intrusion.Recently, data mining approaches, such as the self-organizing map (SOM; [1,2]) analysis employed in this study, have been increasingly used for analyses of the disparate and multivariate geoscientific datasets that are typical for mining and exploration environments (e.g., [3][4][5][6][7][8][9][10]). SOM is an artificial neural network that works on a vector quantization methodology allowing unsupervised analysis, i.e., no prior information is needed on the type or number of groupings within the data, to determine the underlying linear and non-linear relationships amongst different data.…”

“…Recently, data mining approaches, such as the self-organizing map (SOM; [1,2]) analysis employed in this study, have been increasingly used for analyses of the disparate and multivariate geoscientific datasets that are typical for mining and exploration environments (e.g., [3][4][5][6][7][8][9][10]). SOM is an artificial neural network that works on a vector quantization methodology allowing unsupervised analysis, i.e., no prior information is needed on the type or number of groupings within the data, to determine the underlying linear and non-linear relationships amongst different data.…”