A combination of solid-state 13C NMR, X-ray photoelectron spectroscopy (XPS) and sulfur X-ray absorption near edge structure (S-XANES) techniques are used to characterize organic oxygen, nitrogen, and sulfur species and carbon chemical/structural features in kerogens. The kerogens studied represent a wide range of organic matter types and maturities. A van Krevelen plot based on elemental H/C data and XPS derived O/C data shows the well established pattern for type I, type II, and type III kerogens. The anticipated relationship between the Rock−Eval hydrogen index and H/C is independent of organic matter type. Carbon structural and lattice parameters are derived from solid-state 13C NMR analysis. As expected, the amount of aromatic carbon, measured by both 13C NMR and XPS, increases with decreasing H/C. The correlation between aromatic carbon and Rock−Eval T max, an indicator of maturity, is linear for types II and IIIC kerogens, but each organic matter type follows a different relationship. The average aliphatic carbon chain length (Cn‘) decreases with an increasing amount of aromatic carbon in a similar manner across all organic matter types. The fraction of aromatic carbons with attachments (FAA) decreases, while the average number of aromatic carbons per cluster (C) increases with an increasing amount of aromatic carbon. FAA values range from 0.2 to 0.4, and C values range from 12 to 20 indicating that kerogens possess on average 2- to 5-ring aromatic carbon units that are highly substituted. There is basic agreement between XPS and 13C NMR results for the amount and speciation of organic oxygen. XPS results show that the amount of carbon oxygen single bonded species increases and carbonyl−carboxyl species decrease with an increasing amount of aromatic carbon. Patterns for the relative abundances of nitrogen and sulfur species exist regardless of the large differences in the total amount of organic nitrogen and sulfur seen in the kerogens. XPS and S-XANES results indicate that the relative level of aromatic sulfur increases with an increasing amount of aromatic carbon for all kerogens. XPS show that the majority of nitrogen exists as pyrrolic forms in comparable relative abundances in all kerogens studied. The direct characterization results using X-ray and NMR methods for nitrogen, sulfur, oxygen, and carbon chemical structures provide a basis for developing both specific and general average chemical structural models for different organic matter type kerogens.

No abstract

While the broadness of the pyrolysis profile of most kerogens is described well by a parallel reaction model, the pyrolysis profile at a constant heating rate for certain well-preserved algal kerogens is narrower than can be described by a single first-order reaction. Further, these kerogens show an acceleratory period under isothermal conditions that is inconsistent with any parallel or nth-order reaction model. Three different models (serial, Bouster, and three-parameter) are tested against isothermal and nonisothermal pyrolysis data for a few samples, with the conclusion that the three-parameter model fits well and is the most stable and reliable. The three-parameter model reduces to a first-order model when the acceleration parameter is zero. The overall activation energy and frequency factor from this model are very close to those of the T max-shift method recommended earlier.

Many marketing problems require accurately predicting the outcome of a process or the future state of a system. In this paper, we investigate the ability of the support vector machine to predict outcomes in emerging environments in marketing, such as automated modeling, mass-produced models, intelligent software agents, and data mining. The support vector machine (SVM) is a semiparametric technique with origins in the machine-learning literature of computer science. Its approach to prediction differs markedly from that of standard parametric models. We explore these differences and benchmark the SVM's prediction hit-rates against those from the multinomial logit model. Because there are few applications of the SVM in marketing, we develop a framework to position it against current modeling techniques and to assess its weaknesses as well as its strengths.automated modeling, choice models, kernel transformations, multinomial logit model, predictive models, support vector machine

Abstract:Hydraulic connectivity of petroleum reservoirs represents one of the biggest uncertainties for both oil production and petroleum system studies. Here, a geochemical analysis involving bulk and detailed measures of crude oil composition is shown to constrain connectivity more tightly than is possible with conventional methods. Three crude oils collected from different depths in a single well exhibit large gradients in viscosity, density, and asphaltene content. Crude oil samples are collected with a wireline sampling tool providing samples from well-defined locations and relatively free of contamination by drilling fluids; the known provenance of these samples minimizes uncertainties in the subsequent analysis. The detailed chemical composition of almost the entire crude oil is determined by use of comprehensive two-dimensional gas chromatography (GC×GC) to interrogate the nonpolar fraction and negative ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS) to interrogate the polar fraction. The simultaneous presence of 25-norhopanes and mildly altered normal and isoprenoid alkanes is detected, suggesting that the reservoir has experienced multiple charges and contains a mixture of oils biodegraded to different extents. The gradient in asphaltene concentration is explained by an equilibrium model considering only gravitational segregation of asphaltene nanoaggregates; this grading can be responsible for the observed variation in viscosity. Combining these analyses yields a consistent picture of a connected reservoir in which the observed viscosity variation originates from gravitational segregation of asphaltene nanoaggregates in a crude oil with high asphaltene concentration resulting from multiple charges, including one charge that suffered severe biodegradation. Observation of these gradients having appropriate magnitudes suggests good reservoir connectivity with greater confidence than is possible with traditional techniques alone.3