Reservoir fluid characterization by gas chromatography (GC) has an impressive capability of detection and quantification of a wide range of single carbon number (SCN) groups in oil analyses. However, some researchers prefer to report analyses to C 20+ only, with estimation of the C n+ fraction distribution obtained using various correlations. Conversely, other researchers prefer to extend GC analysis to the highest possible SCN group using high-temperature gas chromatography (HTGC), with programming to ca. 370À430 °C. However, the reliability of extended GC analyses to high carbon number fractions is questioned because of a possible overestimation of light and intermediate fractions in the original oils caused by thermal decomposition products. The thermal stability of heavy hydrocarbons at the above HTGC conditions has been a major concern for some authors based on the results of thermogravimetric analysis (TGA) published by Schwartz et al. [Schwartz, H. E.; Brownlee, R. G.; Boduszynski, M. M.; Su, F. Anal. Chem. 1987, 59 (10), 1393À1401], who highlighted thermal instability of heavy oils from around 370 °C. To that end, in this study, a pyrolysis model spanning the n-alkanes (nC 14 H 30 ÀnC 80 H 162 ) at low conversion has been developed and applied to mixtures at the GC column pressure and oven temperatures up to 450 °C. On the basis of this model, the minimum SCN, which could possibly be at risk of thermal cracking at some commonly used HTGC temperature programs, has been obtained by comparing the retention time of n-alkane standard mixtures (nC 10 H 22 ÀnC 75 H 152 ) and the minimal pyrolysis time at the same SCN range of equimolar, heavy, and light mixtures at different dilutions in He and some low isoconversion pyrolysis curves. Finally, this study gives the first insight into the limitation in the practice of GC and introduces a new approach for calculating the minimum SCN not suffering pyrolysis inside a particular GC column.

International audienc

Asphaltenes are a very important component of reservoir fluids. They have a huge impact on crude oil viscosity and are a Flow Assurance concern. They can undergo a phase transition, giving rise to tar mats that seal aquifers precluding aquifer sweep. Local tar deposits can act as a drilling hazard. Upstructure tar (or bitumen) deposition can occur which can flow with produced light hydrocarbons greatly reducing the productivity index. In EOR, miscible gas injection can also give rise to asphaltene deposition. Characterizing these disparate observations is now performed within a simple overarching framework. Here, we combine asphaltene nanoscience, thermodynamics, and fluid mechanics to model asphaltene-rich fluid flow and asphaltene deposition that occur in reservoirs in geologic (or even production) time. This analysis successfully accounts for extensive measurements in several reservoirs in different stages of similar processes. Reservoir black oils with a late, light hydrocarbon charge experience asphaltene instability. This instability does not necessarily cause precipitation; instead, weak instability can cause a change in the nanocolloidal character of asphaltenes without precipitation. Consequently, this less stable asphaltene remains in the crude oil and is thus mobile. This process can result in fluid density inversions and gravity currents that pump asphaltene 'clusters' in oil over reservoir length scales relatively quickly in geologic time. These asphaltene clusters then establish very large asphaltene and viscosity gradients at the base of the reservoir. If the light hydrocarbon instability event continues, a regional tar mat can form. In contrast, if the light hydrocarbon charge is sufficiently rapid, the displacement of the contact between the original and new reservoir fluids overtakes and precipitates asphaltenes locally producing deposition upstructure often near the crest of the field. In this paper, several reservoirs are examined. Two reservoirs have massive, current gas charge and have bitumen deposition upstructure. Another reservoir is shown to be midway through a slower gas charge, with the asphaltene instability causing migration of asphaltenes from the top to the base of the oil column in the form of clusters creating large asphaltene gravity gradients. Another reservoir is shown to have this process completed yielding a 50 meter column of heavy oil at the base of the oil column underlain by a 10 meter regional tar mat. This integrated analysis enables a much simpler understanding of many production issues associated with asphaltenes and provides a way forward for treating disparate asphaltene problems within a single framework.

The purpose of this research work is to determine the maximum single carbon number (SCN) which can be reliably quantified using High Temperature Gas Chromatography (HTGC) analysis of heavy oil hydrocarbons, accounting for (i) thermal cracking risk and (ii) the non/incomplete elution. To that end, an in-house coupled numerical Pyrolysis-GC model has been developed, capable of calculating the degree of elution and of simulating the migration, partitioning, and pyrolysis conversion of a mixture of 11 heavy n-alkanes spanning the range from nC 14 H 30 to nC 80 H 16 throughout the GC column. On the basis of this model and using a commonly used column configuration and temperature program, two conclusions have been made: (i) half of the mass injected of nC 80 thermally decomposed before nC 70 has eluted, suggesting a possible coelution of both nC 70 and the pyrolysis products of nC 80 and therefore making the HTGC analysis of nC 70 and heavier n-alkanes no longer reliable, and (ii) alkanes heavier than nC 70 take progressively longer to elute completely from the column, compromising the resolution of the peaks, i.e., nC 70 takes 2.5 min and nC 80 takes 8.5 min. Moreover, nC 80 remained 12.9 min in the isothermal plateau before complete elution, implying that the nC 80 peak will be overlooked and masked by the FID plateau signal, in combination with column bleed products. Therefore, in the case study the maximum reliable SCN which can be quantitatively analyzed with HTGC will be the lighter components than nC 70 .

No abstract