Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information
Planning and design of well to well tracer tests
A successful well to well tracer test is more than selecting the right tracer. It involves determining the appropriate timing, designing the field test, and collecting and analyzing the samples. A well designed sampling program will produce high quality tracer-response curves for further interpretation.
Timing of tracer programs
The timing for tracer injection depends on the information that is requested. Normally, it is desirable to inject the tracer early in the injection process to obtain information as soon as possible and be able to take the necessary actions to optimize the production strategy. However, it has been seen that water tracer is lost in the reservoir because of imbibition of water. Therefore, if there is large imbibition potential in the reservoir and water is trapped in volumes that do not contribute to the normal flow path, tracers can be trapped and only minor tracer concentrations leak out in the flowing part of the reservoir. This information can be valuable on its own, but if the problem is not considered, it may contribute to wrong interpretation of the results. In this situation, water may break through from an injector that is traced without observing any positive detection of tracer. In general, injecting the tracer in the very front of an injection program is not recommended.
Tracers should not be injected until the pattern area has been pressured up with the injected fluid. If the traced injected fluid is spent in collapsing a gas phase or otherwise pressuring the reservoir, the volume injected to tracer breakthrough, which indicates the volumetric sweep to breakthrough, may be significantly larger than that determined after the pattern area has been pressured.
Well-to-well tracer tests may be performed, in principle, either as a pulse injection or as a continuous injection. The most often applied method is pulse injection. The continuous injection may be useful especially where unsaturated water-wet rock may absorb short tracer pulses by water imbibition from the injected waterfront edge. To obtain a constant concentration in the injected water, a constant recording of the water-injection rate is necessary. In general, continuous injection is more complicated and a tracer engineer must follow the injection for a long period. For pulse injection, however, the whole program can be carried out within a few hours.
The design of a field tracer test has two components: a tracer part and an analytical part. The tracer component includes choosing the tracer for each well, estimating the required amount of each tracer, dealing with the regulations, and planning for acquisition and injection of the tracer into the ground. The analytical part includes selecting an analytical strategy, setting up a sampling program, and determining the background realistic detection limits in the samples collected from the particular field.
Traditionally, two methods have been applied to determine how much tracer must be added to obtain a tracer-production response significantly above background. In general, it is desirable to inject as small an amount as possible to reduce:
- Environmental problems
Both radioactive- and chemical-tracer compounds are potentially harmful to the environment; therefore, the amount applied must be kept to a minimum. In some cases, the maximum permissible concentration of the tracer that may be released to the environment is limited by requirement from the authorities.
In many situations, produced gas containing tracer is reinjected into the reservoir. This has, in some cases, contaminated the whole reservoir and destroyed the possibility of applying the same tracers in another segment of the reservoir. If not considered, flow conditions can be misinterpreted because of positive identification of tracer originated from reinjection of tracer in uncontrolled conditions.
The third reason for the use of low concentrations is to reduce costs. In small reservoirs, the tracer cost is a minor part of the total cost connected to the test but, in larger reservoirs, the tracer costs may be significant, especially if exotic tracers are required.
One method to estimate the tracer concentration is to assume that the injected tracer is diluted uniformly by the entire swept volume when it is produced. Sufficient tracer is added to ensure detection at this dilution volume. The peak tracer concentration is presumed to be well above the average.
The first part of the calculation is to estimate a dilution volume, which is obtained by calculating the water- or gas-filled PV between the injector and the production wells. The first approximation is to assume radial flow from injector, but it usually is modified by any known reservoir conditions, such as known flow channels or barriers, large permeability variations, or other constraints causing nonradial pressure gradients. It is important to know porosity, net pay zone, and distances between wells for the calculation. Optimal design can be obtained by performing full-field simulations. The problem with this method is that many of the parameters are unknown, which again is the main reason for performing the tracer test.
The smallest injection pulse required is usually the amount of tracer needed to obtain an average concentration of 10 times the minimum limit of detection in the dilution volume. This is expected to give a peak production on the order of 100 times the detection limit. To be able to follow production curves reflecting the contribution from different layers and zones, it is important to have at least this amount of tracer injected.
In the ideal situation, the dilution volume, Vd, can be calculated with a radial approximation.
where r = distance between the injector and the producer, h = height of the reservoir zone, ϕ = porosity, and Sw = water saturation. For a gas tracer, Sg should substitute for Sw. In addition, the expansion of the gas has to be taken into account. F is a correction factor that accounts for nonsymmetry caused by barriers, well location, and other restrictions causing changes in the drainage area.
An analytical method to determine tracer production curves in a layered system exists.  The method assumes that the tracer pulse moves radially from the injector to the producers through homogeneous, noncommunicating layers with longitudinal dispersion in the direction of flow. The number of layers, their thickness, and their permeability are used to represent the reservoir heterogeneity. The tracer pulse moving through each layer is diluted at the producers by untagged water from other streamlines in the same layer. This dilution effect is a consequence of pattern geometry. Production of the combined tracer responses from all these layers makes up the response curve of tracer concentration as a function of the cumulative volume of water injected.
The model predicts peak height, breakthrough time, and shape of the produced tracer-response curve from the amount of tracer injected and the pattern geometry. The fundamental equation (Eq. 2) used to calculate tracer-pulse flow in a streamtube is further applied to give analytical expressions for pattern-breakthrough curves. 
The concentration of a tracer at any location within the streamtube, ψ, is the difference between two terms, given by Eq. 2. L = the distance along a streamline, and s1 and s2 = the front and end location of the tracer pulse in the streamline, respectively. σ = the standard deviation that, in a radial system, can be found from
where α = dispersivity and r = the radius at the front defined at a location corresponding to the 50%-concentration point. On the basis of these predictions, tracer amounts can be injected to ensure a peak concentration in the production well that is well above detection limits. In practice, however, such calculations are often based on so many unknown or estimated parameters that it is advisable to inject more tracer than these calculations predict.
A reservoir simulator probably can provide the best estimate for tracer amounts; however, these simulations are based on detailed reservoir description. At the stage of tracer injection, the reservoir model is generally uncertain and, again, it is advisable to inject more than these calculations predict. Sufficient tracer must be injected to enable measurement of unexpected flow behavior.
Collection of samples for tracer analysis
Sampling frequencies will depend strongly on the field considered. Tracer tests are, in general, a method to verify proposed flow scenarios. To cover unexpected behavior, sampling should be started long before expected breakthrough. Sampling frequency should be highest at the start of the flood to avoid missing early breakthrough.
Water samples may be stored in bottles that normally are collected from the separator. Water sampling is cheap, and frequent sampling is advised. Initially, only some of the collected samples need be analyzed; intermediate samples can be discarded if no tracer is found. Once tracer is found, samples are analyzed backward until the tracer breakthrough is found. In certain situations, some tracers may biodegrade after sampling. To avoid this, a biocide may be added to the sample immediately after collection. Adding 0.1 ppm NaN3 to the stock solution can prevent bacterial growth.
Samples can be collected at the wells or from a test separator shared by a number of wells, provided the stream is sampled only near the end of the test when the water is representative of the currently sampled well. Well sampling usually involves a water-separation problem, which can be avoided by sampling at the separator. The optimal situation is to have a dedicated separator for the particular well. New technology involving multibranched wells, horizontal wells with perforation in several reservoir compartments, subsea manifolds, or even subsea separators cause additional problems in providing relevant samples.
Gas tracers normally have been collected on pressure cylinders. The gas may be collected directly on the flowline or from a separator. The tracer content in the gas will depend on where the sample is collected. The partition coefficient of the tracer depends strongly on the pressure. A sample collected on the flowline at 100 bars and one collected at a separator at 30 bars will give different results. Further estimation of the total amount of tracer produced also will depend on the gas/oil ratio (GOR) at the sampling point. To calculate the produced-tracer amount in one particular well, it is necessary to know the pressure, temperature, and GOR at the sampling point. In addition, it is necessary to establish the partition properties of the tracer at these particular conditions. This can be done by measurement or by applying pressure-volume-temperature (PVT) models; however, existing PVT models will not treat all tracer types with the same accuracy.
Collection of gas tracers is more expensive than the collection of water tracers. When gas is collected on pressure cylinders, the cost of the cylinders will add considerably to the analysis cost. A new technology has been developed in which the tracers are absorbed by an activated carbon trap. These capillary absorption tube samplers (CATS) are used for collection of perfluorocarbon (PFC) gas tracers. The tracer then is collected in these tubes in the field, and only small tubes, without any surplus pressure, are shipped to the laboratory. Special sampling equipment is available that ensures a reliable and reproducible sampling. The CATS method is applicable only for a limited number of tracers, and in many situations, it is still necessary to collect samples with pressure cylinders. The cylinder size and the amount of gas collected will depend on the tracers and the concentration expected in the produced gas. Normally, a 200-mL cylinder will be sufficient.
The laboratory applies a large variety of techniques to measure the concentrations of the tracers. The different techniques will have degrees of uncertainty and the differences between detection limit and quantification limit should be distinguished. At concentrations close to the detection limit, it may be very difficult to obtain accurate quantification of the tracer; therefore, some laboratories report only "detected" without quantification when the concentration is low.
The detection limit obtained by an analytical procedure will be influenced by the quality of the sample. In formation water, the detection limit may be different from the detection limit in production water that also contains emulsion breakers, scale inhibitors, corrosion inhibitors, and other additives. It is, therefore, important to have a good cooperation between the field operators and the laboratory to obtain the best quality on the field samples.
Because of the very sensitive analytical techniques needed, it is essential to avoid cross contamination. It is important to plan tracer operations carefully to avoid any possibilities of close contact between injection equipment and sampling equipment. For example, it might be a source of contamination if injection pumps or tracer containers are transported after injection in the same van or stored in the same building as sample bottles or sample equipment.
|CO||=||concentration of tracer in oil phase|
|h||=||height of reservoir zone, L|
|r||=||radius of tracer front in radial system, L|
|s1||=||front location of tracer pulse in the streamline, L|
|s2||=||end location of tracer pulse in the streamline, L|
|Vd||=||dilution volume, L3|
- Abbaszadeh-Dehghani, M. and Brigham, W.E. 1984. Analysis of Well-to-Well Tracer Flow To Determine Reservoir Layering. J Pet Technol 36 (10): 1753–1762. SPE-10760-PA. http://dx.doi.org/10.2118/10760-PA
- Abbazadeh-Dehghani, M. and Brigham, W.E. 1981. Computation of Tracer Production Curves from Various Flooding patterns. Proc., Annual Heavy Oil/EOR Contractor Reports, San Francisco, L-1,L-15
- Galdiga, C.U. et al. 2001. Experience With the Perfluorocarbon Tracer Analysis by GC/NICI-MS in Combination With Thermal Desorption. Tracing and Tracing Methods, Nancy, France, 2001, Recent Progress en Génie des Procèdes (May 2001) 15 (79).
Noteworthy papers in OnePetro
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read