About This Item

Share This Item

The AAPG/Datapages Combined Publications Database

GCAGS Transactions

Abstract


Gulf Coast Association of Geological Societies Transactions Vol. 58 (2008), Pages 747-752

EXTENDED ABSTRACT: New Geochemical Software Technology to Characterize Reservoir Systems in the Llanos Basin, Colombia: Cutting Costs and Learning More

Rick Schrynemeeckers1, Suhas Talukdar2, Javier Paez3, and Uriel Sanchez4

1InfoLogic, Inc., 143 Vision Park Blvd., Shenandoah, Texas 77384

2Baseline Resolution, Inc., 143 Vision Park Blvd., Shenandoah, Texas 77384

3Petrominerales Colombia, Ltd., Calle 114 No. 9-45 Torre B Oficina 1506, Bogotá D.C., Colombia

4Core Laboratories Colombia, S.A., Carrera 39 No. 168-52/56, Bogotá D.C., Colombia

EXTENDED ABSTRACT

Geochemical indicators of petroleum composition provide especially useful and cost-effective tools for evaluating reservoir continuity and allocating commingled production (Kaufman et al., 1990; Kaufman et al., 1997; McCaffrey et al., 2006; Hwang et al., 2000). These geochemical tools are highly complementary to engineering methods for deriving reservoir continuity information, and can effectively replace some engineering methods for allocating commingled production (Hwang et al., 1999). Using geochemistry to understand better the characteristics of reservoirs and reservoir fluids allows engineers to maximize the recovery of hydrocarbon fluids in a field and to determine how many wells should be drilled, where wells should be drilled, and how to maximize the production.

Results of a case study of a well in the Llanos Basin in Colombia are presented to demonstrate how the gas-chromatographic (GC) fingerprints of oils from the well were evaluated using InfoLogic’s ReserViewTM and OilUnmixerTM software to successfully delineate the producing reservoir intervals, identify vertical reservoir fluid compartments, as well as estimate and monitor the production allocation of a two-zone completion over a seven-month period. The results from the unbiased comparison of the oils using the programs identified the reason for a reduction in production, why the American Petroleum Institute (API) gravity of the commingled oil decreased, and why the gasoil ratio (GOR) increased over the seven-month period. The geochemical approach cost approximately 95 percent less than if three production logging tool (PLT) events had been employed.

In the initial discovery stage of the well, four drill stem test (DST) oils were produced in August 2007: DST-1 from the lower zone Upper Cretaceous (two sands), DST2 from the upper Eocene zone (one sand), DST-3 from the upper Eocene zone (three sands – including the one sand from DST-2), and DST-4 that was a combination of the upper zone three sands and the lower zone two sands.

The operator initially wanted to determine if there was vertical continuity between the upper and lower zones. A comparison of the gas chromatograms of the oils using InfoLogic’s ReserViewTM software was performed using the DST-3 oil (the upper zone three sands) and the DST-1 oil (the lower zone two sands).

Figure 1. For the analytical comparison of oil fingerprints it is critical to ensure that users are comparing the same peaks in each oil. This can be difficult due to the fact that retention times shift slightly from oil to oil. This figure shows how ReserViewTM uses Kovats indices to triangulate on each peak and align them for all oil samples. The software then color codes each peak to allow easy comparison for the user.

As seen in Figure 1, ReserViewTM uses Kovats indices to align all the peaks in the oil fingerprints and then color-codes them for easy comparison and visualization. This step is critical in the evaluation process because it compensates for retention time shift and ensures that all calculations are correctly performed on the same peaks across multiple samples. Once the GC range is selected, the user can quickly scan through the carbon ranges to select the peaks to be used for the oil comparison. ReserViewTM then determines the peak height ratios of adjacent peak pairs (Fig. 2). The relative standard deviation (RSD) is calculated for each set of peaks and the RSDs are ranked from largest to smallest. The RSDs indicate the mathematical differences between the oils that cannot readily be observed by just a visual comparison. The peak height ratios from the GC range are then plotted using a star plot, as seen in Figure 3, and on a dendogram, as seen in Figure 4, to provide easy visualization of the similarities or differences between the oils. Both Figures 3 and 4 indicate that the oils are chemically distinct; when these data are considered in light of the field geology and field engineering data, the geochemist can draw the conclusion that there are two vertical reservoir compartments with no fluid communication between the upper and lower zones. It should be noted that in the past, calculations such as these took much more time to perform and were often fraught with errors due to the fact the user could not accurately identify the same peak in multiple chromatograms. With InfoLogic’s ReserViewTM software, a comparison such as this can be performed in a few minutes.

Figure 2. ReserViewTM calculates the ratio of peaks within a carbon range (e.g. between nC8-nC-9) by calculating the ratio of the peak heights. ReserViewTM then calculates the Relative Standard Deviation (RSD) and ranks those. These mathematical comparisons allow users to detect differences between oils that cannot be detected visually.

Once it was determined that the upper and lower zones were not in vertical communication, geologists could use those same single-zone oils and the commingled oil to allocate quantitatively the contribution of each of the two zones to the commingled production. Geochemical determination of the contributions of multiple zones to a commingled oil stream is achieved by identifying natural chemical differences between "endmember" samples. An end-member is produced oil from a single zone, while commingled oil is oil produced from multiple zones into a single tubing string. Compositional differences in the end-member samples and the commingled samples are measured based on the GC data. The production allocation was at first determined by the classical method using mixing curves between the selected peak ratios and then was also determined using InfoLogic’s OilUnmixerTM software. OilUnmixerTM was originally developed by OilTracers L.L.C. and is now marketed by InfoLogic, Inc.

OilUnmixerTM utilizes the geochemistry of the oil for determinative purposes, but unlike ReserViewTM, OilUnmixerTM uses the peak heights of the end-member and commingled oils instead of peak ratios. OilUnmixerTM uses a linear algebra approach to mathematically define the relationship of the various oils. Since the solution commonly utilizes between 30 and 80 peaks to define a mathematical solution in n-dimensional space (where “n” is the number of GC peaks), the answer is highly over-constrained. In this case study peak ratios were used so the results could be compared to the classical method.

The production allocation estimation was performed using the commingled oil, DST-4, and the two end member oils DST-3 (from the upper formation) and DST-1 (from the lower formation). These samples were collected in August of 2007. Additional commingled samples were collected in October of 2007 and in February of 2008. The results of the production allocations are seen in Table 1.

As seen in Table 1, the data indicated a decrease in production from the upper zone and an increase in contribution from the lower zone during the seven-month period from August 2007 through February 2008. Based on the allocation results, the geochemists evaluated the hydrocarbon distribution of the end-member oils from the two formations. The oil chromatograms indicated that the lower formation oil contained higher concentrations of light hydrocarbons and lower concentrations of the higher molecular weight range hydrocarbons when compared to the oil from the upper formation.

Figure 3. If the oils are from the same sources, have the same migration history, and are in contact the RSD between oils will be minimal. The RSDs are plotted on a star chart to provide the user with a easy visualization tool. Areas on the chart where the patterns overlay indicate where the oils have comparability. However, if there are significant areas where the where the patterns to do not overlay, as seen in the blue and pink lines, the oils are not the same and, therefore, are not in contact.

Because the lower formation was increasing its percentage of contribution over time, the result was an increase in light hydrocarbons in the commingled oil at reservoir conditions. At production, the commingled oil lost more of the light hydrocarbon fraction and became enriched in heavier hydrocarbons resulting in lower API gravity. As a result, there was an increased proportion of separated gas resulting in a higher GOR.

The geochemical approach utilized two commercially available software programs, ReserViewTM and OilUnmixerTM, which helped to determine the changes in percent allocation from the two reservoir zones and to identify the reasons for the changes in the GOR and the API gravity of the commingled oil over the seven-month period. The results of this study, together with the production history and bottom-hole pressure data, helped the operator to detect formation damage and use acidification or fracturing to address the problem.

Pay-Per-View Purchase Options

The article is available through a document delivery service. Explain these Purchase Options.

Watermarked PDF Document: $14
Open PDF Document: $24