Prediction of Reservoir Properties from Seismic Data by Multivariate Geostatistics Analysis

A new technique has been developed and is being improved, which allows, on the basis of probabilistic and statistical analysis of seismic data, to predict and evaluate the most important parameters of rock properties (including the reservoir properties such as porosity and permeability), that is, oil saturation, effective thicknesses of reservoirs, their sand content, clay content of seals, and others; it is designed to predict the reservoir properties with sufficient accuracy and detail, for subsequent consideration of these estimates when evaluating hydrocarbon reserves and justifying projects for the deposits development. Quantitative reservoir-property prediction is carried out in the following stages: –Optimization of the graph ("scenario") of seismic data processing to solve not only the traditional structural problem of seismic exploration, but also the parametric one that is, the quantitative estimation of rock properties.–Computation of seismic attributes, including exclusive ones, not provided for in existing interpretation software packages.–Estimation of reservoir properties from well logs as the base data.–Multivariate correlation and regression analysis (MCRA) includes the following two stages: Establishing correlations of seismic attributes with estimates of rock properties obtained from well logs.Construction of multidimensional (multiple) regression equations with an assessment of the "information value" of seismic attributes and the reliability of the resulting predictive equations. (By the "informative value" we mean the informativeness quality of the attribute.)–Computation and construction of the forecast map variants, their analysis and producing the resultant map (as the most optimal map version) for each predicted parameter.–Obtaining the resultant forecast maps with their zoning according to the degree of the forecast reliability. The MCRA technique is tested by production and prospecting trusts during exploration and reserves’ estimation of several dozen fields in Western Siberia: Kulginskoye, Shirotnoye, Yuzhno-Tambaevskoye, etc. (Tomsk Geophysical Trust, 1997-2002); Dvurechenskoe, Zapadno-Moiseevskoe, Talovoe, Krapivinskoe, Ontonigayskoe, etc. (TomskNIPIneft, 2002–2013).

A step by step procedure for multivariate modeling

<p><span><span>In environmental sciences, it is very common to observe spatio-temporal multiple data concerning several correlated variables which are measured in time over a monitored spatial domain. In multivariate Geostatistics, the analysis of these correlated variables requires the estimation and modelling of the spatio-temporal multivariate covariance structure.<br>In the literature, the linear coregionalization model (LCM) has been widely used, in order to describe the spatio-temporal dependence which characterizes two or more variables. In particular, the LCM model requires the identification of the basic independent components underlying the analyzed phenomenon, and this represents a tough task. In order to overcome the aforementioned problem, this contribution provides a complete procedure where all the necessary steps to be followed for properly detect the basic space-time components for the phenomenon under study, together with some computational advances which support the selection of an ST-LCM.<br>The implemented procedure and the related algorithms are applied on a space-time air quality dataset.<br>Note that the proposed procedure can help practitioners to reproduce all the modeling stages and to replicate the analysis for different multivariate spatio-temporal data.</span></span></p>

GIS-Based Analysis and Mapping of the Winter Chilling Hours in Mainland Spain. Application to Some Sweet Cherry Cultivars

The knowledge of the chilling requirements for breaking rest and flowering of fruit trees is necessary to properly select cultivars and to avoid losses due to an inappropriate cultivar selection in a particular geographical location. With the aim of providing an analysis using three models (Chilling Hours, Utah Model, and Positive Utah Model) to estimate the accumulation of winter chilling, quantifying its spatial variability and representing the spatial pattern throughout mainland Spain, temperature data from 72 meteorological stations, considering the 1975–2015 period, were utilized. The statistical properties of values corresponding to each winter chilling model were assessed and, later, they were mapped by means of an integrated geographic information system (GIS) and a multivariate geostatistics (regression-kriging) and algebra map approach. The results show that measures obtained with the three chilling models were highly related, which were used to visualize the spatial variability of the accumulated winter chilling considering each model. Moreover, the fact that elevation and latitude are related to the chilling hours enables their use as auxiliary variables to better estimate at unsampled locations and generate more accurate maps. Knowledge of the spatial patterns of chill accumulation in different areas of mainland Spain is of great importance when appropriate fruit trees and cultivars have to be selected. Moreover, if a high probability of satisfying the chilling requirements in any area is considered, quantile maps can be used instead of maps based on mean values. Finally, the potential spatial distributions of three sweet cherry cultivars were delineated using the obtained maps.

Multivariate Geostatistical Modeling of Lower Calorific Value in Multi-Seam Coal Deposits

The estimation of fuel characteristics and spatial variability in multi-seam coal deposits is of great significance for the optimal mine planning and exploitation, as well as for the optimization of the corresponding power plants operation. It is mainly based on the quality properties of the coal (i.e., Lower Calorific Value (LCV), ash content, CO2, and moisture). Even though critical, these properties are not always measured in practice for all available borehole samples, or, they are generally estimated by using non-parametric statistics. Therefore, spatial modeling of LCV can become problematic due to the limited number of data. Thus, the use of other available correlated attributes might be helpful. In this research, techniques of multivariate geostatistics were used to estimate and evaluate the spatial distribution of quality properties in a multi-seam coal deposit, with special reference to the LCV. More specifically, kriging, cokriging, and Principal Component Analysis (PCA) techniques were tested in a case study as estimators of the LCV, using an extensive set of borehole data from the South Field lignite mine in Ptolemais, Greece. The research outcomes show that the application of kriging with two PCA factors and the use of inverse transform result in the best LCV estimates. Moreover, cokriging with two auxiliary variables gives more accurate values for a LCV estimate, in relation to the kriging technique. The research outcomes could be considered significant for the coal mining industry, since the use of correlated quality attributes for the estimation of LCV may contribute to a reduction of the estimation uncertainty at no additional drilling cost.