Production Evaluation
Determination of structure in a field under development is critical in the assessment of where in-fill wells should be drilled. Communication among reservoir zones is also critical in the assessment of development, production allocation, and eventually enhanced recovery approaches. Figure 1 is a diagrammatic illustration of an oil field. How can you answer the following questions?
Figure 1. Assessing reservoir continuity and production allocation is critical in development of a field. Reservoir continuity and production allocation can be determined from inexpensive Reservoir Oil Fingerprinting (ROF) techniques without pressure testing and loss of production or possible failure of a well to come back on-line.
Development of an oil field requires careful assessment of field structure and connectivity of reservoirs. Reservoir continuity assessments must be continually evaluated as step-out wells are drilled. Likewise, completion of different horizons, i.e., commingled production, requires assessment of allocation and the ongoing effectiveness of well plumbing.
Standard Production Evaluation Practices
Reservoir continuity can be determined quite reliably by pressure testing and mapping of reservoirs. Of course, these tests are expensive and include considerable risk:
The Perception
It is generally perceived that physical testing is the only means to evaluate these production questions. In fact pressure testing is an excellent means of evaluation of reservoir connectivity, but (1) it is not the only means, and (2) it may not necessarily be the best means particularly when low oil prices reduce profitability and when high oil prices means lost production, lost wells, and inability to evaluate commingled production and well plumbing on an ongoing basis.
An Alternate and Well Tested Method: Reservoir Oil Fingerprinting (ROF)
Organic geochemistry is generally thought of as strictly an exploration tool to evaluate source rocks or oils in order to high-grade prospects. In fact it is a complete petroleum systems evaluation tool, which includes assessment of reservoired oil. While oil physical properties are the common approach in engineering and development, oil chemistry determines those properties, of course, depending on PVT conditions. What better way to evaluate an oil, than to evaluate its chemistry?
One of the techniques utilized in organic geochemistry to characterize rock extracts and oils is gas chromatography. Gas chromatography is a technique that will separate and detect various compounds present in a simple or complex mixture. It does this by vaporization of the volatile components in the sample, e.g., oil, and then forcing these vapors through a narrow bore fused silica tube (0.10 to 0.55 mm i.d.) of a certain length coated with chemicals that interact with the vapors. These factors are tuned to the compounds of interest in order to obtain the maximum separation and resolution. As the compounds elute from the opposite end of the column, they are detected usually by a flame ionization detector (FID) or in the case of biomarkers, the ionized particles are detected by differences in their mass and charge (mass spectrometer or MS). When oil is injected, the hydrocarbons present in the oil are separated and detected resulting in a "fingerprint" of the oil. A fingerprint is basically the yield and distribution of hydrocarbons present in a sample. The predominant compounds in normal crude oils are paraffins (generally the highest peaks in the fingerprint) and aromatic hydrocarbons.
However, in biodegraded oils the resin and asphaltene fractions of the oil predominate the fingerprint and typically yield few resolved compounds and a high unresolved hump of complex composition (not high quality oil). Figure 2 illustrates the gas chromatographic fingerprinting process.
Figure 2. Gas chromatographic analysis of an oil provides a "fingerprint" of the oil, which is the relative yield and distribution of resolvable compounds present in the oil under the analytical conditions employed.
Interestingly enough, the earth itself also employs a bit of chromatography, for example, compounds are separated during expulsion and migration of oil through various chemical and physical interactions. These natural geochromatography processes result in a distinctive crude oil composition in a given reservoir.
For example, in the laboratory we use gravity column liquid chromatography (LC) to separate crude oils by physico-chemical processes into compositionally distinct chemical families: saturate, aromatic, resin, and asphaltene fractions. Other processes can be used to separate the saturate fraction into normal, branched, and cyclic hydrocarbon fractions. These processes utilize beds of alumina oxide and silica gel to physically adsorb the molecules until organic solvents are passed through the system. Utilizing the simple chemical process that likes dissolves likes, solvents are used in which only certain hydrocarbon fractions are soluble, e.g., pentane solubilizes the saturate fraction hydrocarbons from the alumina and silica beds. Not surprisingly, the results are dependent upon a variety of physical and chemical properties utilized in the system, including pore size, purity, water content, solvent, temperature, and pressure.
In the natural system similar processes occur as crude oil is generated and expelled from the source rock some alteration of the originally generated oil occurs. This alteration is dependent upon the chemistry of the oil and gas and the chemical and physical properties of the rocks encountered during expulsion and migration to the trap. The variation between source rock extracts and oils are quite obvious when their chemical composition is compared.
The differences in oils generated by the same source rock feeding two different reservoirs are quite subtle. Rather than focusing on the primary components of the oil, minor constituents of the oil are carefully compared. These are compounds that are most directly affected by physico-chemical processes that result in minor variations in crude oils from the same source. The ability to identify distinct reservoir compartments based on oil composition is key to addressing production problems (Slentz, 1981).
The Technique
Reservoir oil fingerprinting (ROF) is a combination of analytical and interpretive techniques that utilizes the "fingerprint" of oils to determine whether they are part of the same reservoir compartment. The differences seen in ROF fingerprints are not usually related to source rock differences, but are a result of differences in migration pathways to the reservoir compartment. This results in geochromatography of the oil causing very slight differences in oil chemistry that can be discerned through careful analysis of oil fingerprints.
Since a GC fingerprint is a representation of the relative concentrations of compounds present in an oil as analyzed in the laboratory setting. It is, in fact, a histogram of the relative yield and distribution of these compounds. A fingerprint is similar to a well log: it is a molecular log of the composition of a specific oil. While fingerprints are directly related to the source rock of the oil, reservoirs in continuous communication have nearly identical fingerprints even in trace components of the fingerprint, the so-called "grass between the trees", where the normal paraffins are the "trees" that dominate most crude oils. For example, oils 1-3 are all very similar and from the same source rock (Figure 3). However, oils 1 and 2 have nearly identical distributions of the minor hydrocarbons eluting between major paraffin peaks as shown by the enlargement of the area between two adjacent normal paraffins (Figure 3, lower fingerprints). Oil 3 has quite a different distribution of these compounds as shown by the comparison of peak height ratios C-G among these 3 oils (Figure 3, peak ratios). For example, ratio "F" is reversed in oil #3 compared to oils #1 and #2.
A polar plot of various peak ratios illustrates that oils #1 and #2 are nearly identical, whereas oil #3 is quite different (Figure 4). The implication is that oils #1 and #2 are in a continuous reservoir, while oil #3 is not in the same continuous reservoir. This technique has been widely used by major oil companies for evaluating reservoir continuity, production allocation, and well plumbing problems (e.g., see Amoco's Ross and Ames 1988 paper, Chevron's et al. 1987 and 1991 papers, Saudi Aramco's 1995 paper, and Chevron's and Lenz 1995 paper).
While the technique utilizes simple and inexpensive gas chromatographic analysis of oils, interpretation of results is dependent on the quality and reproducibility of these chromatographic results, as well as the ability of the interpreter to locate and assess differences among oil samples. Since a relative comparison of unknown peak heights is utilized, oils need to be analyzed at the same time under the same conditions. Absolute yields may be used when quantitative GC analysis is completed and specific amounts of a given compound can then be compared independent of the time of analysis (Ganz et al., 1997).
Figure 3. Oils from the same source rock have very similar gas chromatographic fingerprints. However, these similar looking oils may have minor compositional differences as shown by selected peak ratios of minor constituents of the oils, which are indicative of different reservoir compartments.
|
|
Figure 4. Radar or star plot of selected peak ratios from gas chromatographic analysis of oil samples. Oils 1 and 2 are similar and likely in the same reservoir compartment, whereas Oil 3 is in a different compartment. |
A polar plot or "star" diagram is a simple way to illustrate differences in GC fingerprints particularly selected peak ratios used to distinguish oils from distinct reservoirs. These selected peak ratios may be evaluated by statistical means. Hierarchial cluster analysis (HCA) is a common way to demonstrate statistical similarities or differences among oils.
The null hypothesis in the ROF technique is that oils are identical and that no differences can be found. If differences are found, the oils are not alike, and, therefore, not part of the same reservoir compartment. However, if oils are found to be alike, there are situations where there could be discontinuous reservoir compartments.
Two distinct oil fields approximately 2 miles apart located offshore Cabinda are separated by a structural low. The two oil pools are horizontally and hydraulically connected via a common aquifer and have a common oil-water contact. Standard geochemical analysis and interpretation revealed that the oils have a common source rock of comparable maturity. Do the oils form a common, continuous pool?
Using selected peak ratios that are indicative of distinct compositional differences among the oils from the two fields, the polar plot resulting from the selected peak ratios from the ROF technique shows two distinctly different reservoirs despite the hydraulic connection between these two fields (Figure 5). This plot illustrates that the oils are very similar, but also that differences in oils between the fields are greater than the intra-field variation. Likewise, in the Takula field, there are vertically distinct reservoirs. These are distinguished by variations in peak ratios of unknown compounds found between n-C9 and n-C18 paraffins. These variations are easily recognizable in a polar plot (Figure 6).
Figures 5-6. Star plots of selected peak ratios from the Takula and Numbi oil fields, offshore Cabinda. Figure 5 (left) illustrates different peak ratios from the gas chromatographic analysis for the oils from the 2 fields. Figure 6 (right) illustrates differences within the Takula field indicative of vertically separated reservoirs. (After Kaufman et al., 1990)
Thus, this case study demonstrates:
It also illustrates a relative insensitivity to the quality of the oil sample – how it was gathered, shipped, and stored. While the loss of light hydrocarbons occurs regardless of handling procedures (except for pressurized samples collected under reservoir conditions), the compositional differences are still apparent in peak ratios utilizing C9+ hydrocarbons.
In the course of development of the Bay Marchand Field, Chevron used the ROF technique to distinguish whether oils from various wells were in the same fault block. Prior to the geochemical analysis, prevailing information suggested two distinct fault blocks with wells 1-2 in Block "N" and wells 3-6 in Block "S" (Figure 7A). Gas chromatographic analysis showed distinct compositional differences among various oil samples as illustrated by the polar plot of six oils (Figure 8). This polar plot shows differences in various gas chromatographic peak ratios (A-K).
Based on these results combined with available production data and reconsideration of the structural geology, a new structure map for the eastern portion of this field was constructed (Figure 7B). Development well 7 was later drilled in the revised Block "N" structure and proved productive. Subsequently, its fingerprint and selected peak ratios were shown to match the other "N" block oils.
Figures 7A-B. Structure maps before (A) and after reservoir continuity assessments (B). (after Kaufman et al., 1990).
Figure 8. Star plot illustration of selected peak ratios. Oils 1, 2, 3, and 7 are similar, whereas 5 and 6 and 4 are different. (After Kaufman et al., 1990).
This case study illustrates the use of Reservoir Oil Fingerprinting (ROF) technique to:
The value of having oil samples collected from wells through time was illustrated in this example from the Gulf of Mexico. The No. A-2 well was completed in two separate zones via a short and long stringer. Oils were collected and archived over a period of years (1967-1986) from both horizons in this well. The original 1967 oils from each horizon had distinct compositions revealed through an assessment of selected peak ratios to demonstrate these differences (Figure 10).
|
|
Figure 10. Star plot of selected peak ratios from reservoir oil fingerprinting. The 7000 ft. oil and 7800 ft. are different indicative of separate reservoir compartments. (After Kaufman et al., 1990) |
As the number of oils and peak ratios used in this study was large, statistical analysis of the selected peak ratios was used to evaluate any differences among the oil samples. A cluster analysis of oils from the short and long stringers by year illustrates the differences between the two horizons from the 1967 and 1972 oils (Figure 11). However, the 1981 long stringer oil is intermediate between the short and long stringer oils and thereafter is indistinguishable per se from short stringer oils. Subsequent reworking of the well in 1987 found that the long stringer had lost seal near perforations for the short stringer production. This case study demonstrates:
Figure 11. Hierarchial cluster analysis of selected peak ratios of oils through time.
The ability to monitor production changes through time in commingled production wells is another valuable application of ROF analysis. These data can be used to evaluate the extent of commingling and also the effect of secondary completion efforts.
In this study Chevron collected oil samples quarterly from 12 wells. Production from these wells is commingled from 2 or more zones that are under either steam drive or water flood. Peak ratios were measured and recorded throughout the year. From May 1986 to August 1986 production was noted to increase in the upper horizons based on measured changes in peak ratios (Figure 12). These results show the ability of the ROF technique to monitor enhanced recovery efforts.
Figure 12. Change in production yields from 2 zones through time after secondary recovery efforts.
Laboratory mixes of two end member oils were used to construct a calibration curve for oil mixing. Three distinct peak ratios were measured and recorded using precise mixes from each end member oil. An unknown oil "X" had peak ratios that identified it as a 40/60 mix of oils "A" and "B" (Figure 13).
This example illustrates the possibility of evaluating oil mixes to monitor production. This includes the capability to monitor commingled production for well plumbing, depletion, and effect of enhanced recovery efforts when present. This and the previous case study illustrate the importance of having a good, historical oil collection.
Figure 13. Laboratory calibration curves for evaluating oil mixes such as from commingled production. Evaluation of the linearity of the calibration curves is an important consideration in evaluation of oil mixes. (After Kaufman et al., 1987)
A development geologist for Chevron samples oils leaking from shut-in wells in Eddy County, New Mexico. These wells were originally completed in lower horizons for gas and some oil production was later obtained. Because of low oil contents, these wells were scheduled for sale or abandonment.
The oil samples recovered from the casing backside as well as end member oils from nearby oil productive horizons in the Delaware and Bone Springs formations were analyzed by ROF. The "backside" oil was found to be a mix of Delaware and Bone Springs oils but with a more "Delaware-like" fingerprint (Figure 14). While porous zones were noted from well logs, it was not attractive to perforate all zones due to costs and likely high water contents. Based on the geochemical assessment, the well was recompleted in the Delaware and flowed 268 BOD of 41° API oil.
Figure 14. Hierarchial cluster analysis (HCA) of backside produced oils as well as archived oil samples representative of specific productive intervals. The backside and produced oils were found to be more similar to the Delaware-type oil than the Bone Springs-type oil. (After Burgess and Lenz, 1994)
The first year economics of this analysis and recompletion costs are summarized as follows:
First year revenue: $577,000.00
Analytical Costs: $ 5,000.00
Recompletion costs: $ 74,700.00
Gross Income: $502,300.00
This case study illustrates:
Oils recovered from the Rus Formation above the Ghawar field in Saudi Arabia were fingerprinted to assess their origin. Were they migrated oils or due to leaking casing? Since many oils in this area and elsewhere in the world are light condensates, essentially devoid of the heavier hydrocarbons used to fingerprint and assess connectivity in other examples presented in this application note, light hydrocarbon ratios were utilized by Halpern. Ratios were constructed from unaltered or primary oils and biodegraded oils from the same oil type as determined by other geochemical means. Comparison of oils from the Rus Formation were found to be similar to the Arab-D oil and quite different from the deeper Khuff condensate using light hydrocarbon ratios. Since the oil was unaltered and in a shallow reservoir, it was concluded that the oil was likely the result of leaking casing.
|
|
Figure 15. Star plot of light hydrocarbon peak ratios of increasing variability from Tr1 to Tr8 as determined by Halpern (1995). The Rus oils were found to be very similar to the Arab D oil, and quite different from the deeper Khuff condensate. This illustrates the use of light hydrocarbon ratios to assess oil similarities or differences, which is especially important when dealing with condensates. (After Halpern, 1995) |
A 700+ well oil field was completed as open hole production. In this case it was difficult to tell from which zones the oil is being derived. Geologically, two producing horizons (Zones A and B) had different porosity and permeability. If production was found to be largely from low porosity strata, it would not have been useful to inject water, because it would have traveled the less resistant route through the more porous horizon. The goal of this study was to determine the relative contribution of each horizon using gas chromatographic methods. Well 696 showed an increase in production after shutting-in and pressurizing the reservoir system (Figure 16) and the gas chromatographic fingerprints showed differences in peaks eluting between normal paraffins compared to the fingerprint before pressurization.
|
|
Figure 16. Well #696 showed an increase in production yield after stimulation and the gas chromatographic fingerprint of the produced oil changed. |
Based on an experience set of fingerprints of oils from both horizons A and B, it can be determined whether the stimulation increased production from A, B, was a change in the mix of A and B, or was a different horizon. A principal component scores plot using the Pirouette™ chemometrics software (Infometrix, Woodinville, Washington) was used to assess the new oil production. This analysis suggests that the oil now being produced from the 696 well is from a completely different horizon and is different from other oils in the field (Figure 17). Oils either plot along a mix line of Zone A to Zone B or toward a third axis, the new zone C oil.
|
|
Figure 17. Principal scores plots showing oils known to be Zone A or B, mixes, and the new oil from Zone C. |
The 700+ well field is shown (Figure 18) with water injection wells (blue dots), Zone A oils in yellow, Zone B oils in green, and the new Zone C oils in red. The large red circle is well 696, which is producing 13% from Zone A, 17% from Zone B, and 70% from Zone C.
Figure 18. Field map showing the main producing zone by color coding: yellow square – Zone A oil predominates; green square – Zone B oil predominates; red square – Zone C oil predominates. The large red circle is well 696. The blue points are injection well locations. (After Ramos et al., 1997)
To complete Reservoir Oil Fingerprinting (ROF) analysis, the following is needed:
Sampling
Samples need to be carefully and consistently collected and precisely documented as to operator, well name, horizon, sampling location (e.g., well head), and date. It is best to maintain a historical collection to evaluate changes in production resulting from pump down or plumbing problems. Sampling details are available upon request as well as details on archiving oil samples. Humble offers oil storage capabilities in addition to ROF analytical services. We currently have over 15,000 oils in storage.
Oil samples from DST, RFT, swab runs, well head, or single reservoir storage tanks may be used for ROF analysis. Well head samples are preferred. Pressurized samples need to be carefully subsampled to avoid sample fractionation due to gas exsolution.
Samples should be archived through time from the well head to provide a historical record of oil composition particularly to evaluate commingled production and well plumbing problems. These archived oil samples can be used when new oils are produced or new questions arise about field structure or continuity.
For detailed sampling instructions and sample bottles, please contact us via telephone, fax, or email as shown below.
Quotations on field and well studies are available upon request.
Analytical and Interpretive Costs
The cost for reservoir continuity analysis or production evaluation is remarkably low. Each oil is analyzed by very high resolution light hydrocarbon whole oil gas chromatography providing detailed light hydrocarbon and C8-C10 hydrocarbon peak ratios. Duplicate analyses are included to assess reproducibility of the method and any inherent variability as a result of changing analytical conditions. Each GC fingerprint is carefully compared peak-to-peak and peak ratios showing the greatest differences are selected for comparison among oils. These selected peak ratios are plotted using a radar or star plot (or line chart) and processed by cluster analysis. A report is constructed based on these data and fully explains the analytical results and their significance.
For a partial field assessment program of 10 oils, the cost is about $3,500.00 for a complete, detailed interpretive report. Larger, more complete field assessments are comparably priced on a per project basis. The detailed interpretive report includes star plots and hierarchial cluster analysis (HCA), peak ratios, and all gas chromatographic data for reference.
Software for looking at GC traces is available to read HP binary GC ChemStation files or AIA files. This software also allows GC fingerprints to be pasted into internal reports. In addition peak integrations, ratios, and the chromatograms themselves may be archived in a data base.
Analytical WorkStations for Laboratories
Humble Instruments & Services, Inc. offers ROF WorkStations for in-house completion of ROF analysis. Likewise, chromatographic fingerprints and peak ratios can be statistically evaluated using our ROF software that incorporates the ability to read chromatograms (fingerprints) directly or entry of selected or reported peak ratios. Descriptive statistical analysis can be completed on oils and models constructed to assess unknown oils.
This technique requires highly reproducible gas chromatography analysis. Using the ROF technique with Retention Time Locking (RTL) provides results that are directly comparable GC-to-GC, project-to-project, and year-to-year.
Reservoir oil fingerprinting for development and monitoring of oil fields
The Geochemical Advantage!
References
and R. Lenz, 1994, Gas chromatography identified "backside" oil leading to a profitable oil well completion in New Mexico, 207th Annual Am. Chem. Soc. Nat. Meeting, San Diego, CA, oral presentation.
Ganz, H. H., X. Varlet, F. V.D. Veen, M. A. Keen, and G. v.d. Bos, 1997, Reservoir Geochemistry of Oil Fields in the Netherlands, 18th International Meeting on Organic Geochemistry, September 22-26, 1997, Maastricht, The Netherlands, oral presentation.
, 1995, Development and Applications of Light-Hydrocarbon-Based Star Diagrams, AAPG Bull., Vol. 79, No. 6, pp. 801-815.
Kaufman, R. L., A. S. Ahmed, and, , 1990, Gas Chromatography as a Development and Production Tool for Fingerprinting Oils from Individual Reservoirs: Applications in the Gulf of Mexico, in GCSSEPM Foundation Ninth Annual Research Conference Proceedings, p. 263-282.
Kaufman, R. L., A. S. Ahmed, and W. B. Hempkins, 1987, A new technique for the analysis of commingled oils and its application to production allocation calculations, 16th Annual Indonesian Petroleum Assoc., paper IPA 87-23/21.
Ramos, L. S., B. G. Rohrback, and , 1997, Pattern Recognition Techniques in Petroleum Geochemistry, 1997 ACS Meeting, Cancun, Mexico, oral presentation.
Ross, L. M. and R. L. Ames, 1988, Stratification of oils in Columbus basin off Trinidad, Oil & Gas Journal, Sept. 26, 1988, pp. 72-76.
Slentz, L. W., 1981, Geochemistry of Reservoir Fluids as a Unique Approach to Optimum Reservoir Management, Middle East Oil Technical Conference of the SPE, paper SPE 9582, Manama, Bahrain, oral presentation.
Copyright © 1999-2003 Humble Instruments &
Services, Inc.. All rights reserved.
Revised:
