A Single-Well Tracer Test With In-Situ-Generated CO2 as the Oil Tracer
- S.L. Wellington (Shell Development Co.) | E.A. Richardson (Shell Development Co.)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- May 1994
- Document Type
- Journal Paper
- 85 - 91
- 1994. Society of Petroleum Engineers
- 4.1.9 Tanks and storage systems, 4.1.2 Separation and Treating, 5.6.5 Tracers, 3.1.6 Gas Lift, 2.4.3 Sand/Solids Control, 5.2.1 Phase Behavior and PVT Measurements, 4.2.3 Materials and Corrosion, 5.3.4 Reduction of Residual Oil Saturation, 5.3.2 Multiphase Flow, 5.4.10 Microbial Methods, 1.2.3 Rock properties, 4.1.5 Processing Equipment
- 1 in the last 30 days
- 329 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
A single-well tracer test (SWTT) that uses soluble amounts of in-situ-generated CO2 as the oil tracer is described. Water-soluble-only injectants simplify test interpretation and improve residual-oil-saturation (ROS) determination. The CO2-generator chemistry allows the test to be tailored over a wide range of injectivities and reservoir temperatures.
This paper documents a new single-well tracer chemistry that uses hydrolysis of halogen-organic acid salts as a means to generate CO2 and to trace ROS.1 The availability of generators with different, well-behaved hydrolysis rates allows the test to be tailored to each individual well. Suitable water tracers include methanol, tritium, bicarbonate, or the spent CO2 generator. Field results demonstrate test application and interpretation.
Comparison of Ester and CO2 Technology
Figs. 1a and 1b schematically present the generalized positions of the tracers used in the ester and CO2 systems at the end of the displacement step relative to the wellbore. Note in Fig. 1a that the ester is "overflushed" during injection and then "back-overflushed" by the displacement brine during production. In the CO2 system (Fig. 1b), the generator chemicals are water-soluble-only when injected. Thus, the reactive and nonreactive tracers travel together, remain in front of the displacement brine during injection, and have locations that always coincide beyond the injection-fluid-cooled near-wellbore region at pump shutdown.
Ester chemistry calls for injection of a reduced-volume "minitest." If the mini test shows some return of injected nonreacting tracer and the presence of generated tracer, a full volume "main test" is run.2 With the CO2-tracer chemistry, however, a minitest is unnecessary because the reaction rates ofthe acid generators are more predictable downhole than are ester reaction rates. The Appendix provides more information on CO2-generator kinetics. Little difference exists in field operations with either tracer system. Similar procedures and equipment (Le., injection pumps, storage vessels, and test separators) are used and other chemically distinct nonreacting water-soluble-only tracers may be added at various stages of the injection sequence to aid in interpretation and to address well- and reservoir-specific questions. Although CO2 generators are more expensive than esters, chemical cost is a small fraction of SWTT expense in either case.
Partition Coefficient Properties and Determination
The partitioning nature of CO2 is nearly ideal for ROS determination. Ko is ˜3 and is relatively insensitive to salinity and temperature. The oil-tracer profile is well-defined and adequately separated from the water-tracer profile for sensitive ROS measurements. Figs. 2 and 3 show that the differential change in Ko for CO2 is 0.01 and 0.008 nondimensional units per 1,000 ppm NaCl and 1°F change, respectively. Fig. 4 indicates that there is essentially no differential change in Ko with CO2 concentration in the range studied. Frontal analysis was used to obtain the CO2 Ko values between methane reconstituted crude and synthetic reservoir brine.3 Consolidated Berea and Bentheim cores and sandpacks made from Berea cores, Clemtex No. 5 sand (a well-sorted and rounded fine sand), or ground silicon carbide were used. Silicon carbide was the preferred medium for measurement of Ko values because it was essentially inert and constant results were obtained on repeated tests. The long time required to precondition and equilibrate live brine and oil with consolidated and disaggregated cores made those materials too costly to work with on a regular basis. Eight Ko measurements on reconstituted crude and brine at reservoir conditions yielded a value of 3.36± 0.15. A similar value was determined for a California crude at 200°F. The Ko values did not vary significantly when stock-tank oil was substituted for live oil. Separate in-situ CO2-generation tests were run in preconditioned Berea cores and packs. These studies confirmed measured hydrolysis rates, Ko values, and use of CO2 as an oil-saturation tracer. Multiple flow tests run in cores containing carbonate materials with no oil present showed that the CO2 propagated at the same rate as tritiated brine.
Interpretation of Tracer Response
During production, the generated CO2 is retarded by the oil and separates from the water-soluble-only tracers that return directly through the water. Assuming that (1) the flow pattern is reversible, (2) the hydrolysis rate of the reacting tracer is constant during the process, and (3) the main part of the reaction occurs during the soak period, the ROS can be calculated from the differential response of the two tracers and Ko. The governing equation is
where Voi and Vwi are the associated tracer-determined arrival volumes. Voi is always greater than Vwi (unless there is no oil in the reservoir; then Voi=Vwi) because it takes more production volume to transport the partitioning tracer back to the wellbore than for the nonpartitioning tracer. When the Voi/Vwi ratio in Eq. 1 is constant throughout tracer production, the required quasistatic equilibrium condition for application of chromatographic principles is satisfied and Sor is uniquely determined. In the ideal case (Gaussian-shaped response curves), Sor is simply calculated from the peak-to-peak tracer separation volumes and Ko. Various studies,4-9 including this one, have shown that Ko can be measured with sufficient accuracy to prevent this parameter from being considered a problem in oil-saturation determination. Mathematical modeling is required for the interpretation of complex nonideal response patterns.
Injection of water-soluble-only tracers in the CO2 method allows an improved oil-saturation calculation procedure compared with the iterative and possibly nonunique procedures that are required when oil-soluble esters are used. All parameters specific to the well and reservoir are determined from the water-tracer response and held constant. Next, various Voi/Vwi ratios are applied to the water-tracer simulation results to calculate a series of possible oil-saturation responses. The measured and calculated oil-tracer responses are then overplotted. If an overplot match is found, Eq. 1 is the governing equation and the oil saturation is uniquely determined. Note that dispersion is the same for both oil and water tracers because they return from the same position in the reservoir through the single flowing brine phase.
If the measured oil-tracer curve does not overlay one of the calculated curves (i.e., measured Voi/Vwi is not constant), the test is flawed. This interpretation procedure, unique to water-soluble-only injectant tracers, provides a check on the validity of the test physics, generator chemistry, engineering design, tracer and volume determinations, well integrity, and operating procedures.
|File Size||927 KB||Number of Pages||7|