Enhanced recovery

Gas Tracers: A Decade of Learning and Experience

This paper discusses a crestal gas-injection project that was carried out in a supergiant heterogeneous-carbonate oil field.

Fig. 1—Reservoir map with sector distribution.

In interwell tracer studies, reservoir complexities often make tracer-breakthrough time difficult to assess. Although reservoir-simulation studies serve as an effective tool in predicting a breakthrough, the tracer behavior in the real porous media sometimes presents intriguing surprises. This paper discusses a crestal gas-injection project that was carried out in a supergiant heterogeneous-carbonate oil field.


A tracer-survey study can provide a variety of information about the heterogeneity of a formation. While transient tests can provide information about reservoir continuity, thief zones are difficult, if not impossible, to detect. This is because pressure-transient tests provide an arithmetic average for reservoir total transmissibility over the tested formation thickness, while a tracer survey provides a direct evaluation of the flow field between the injection and production wells.

Apart from heterogeneity, tracer studies also serve as an informative tool to yield fluid-flow paths inside the reservoir, popularly known as reservoir streamlines. A tracer survey not only can precisely identify these preferential paths that are detrimental to the sweep of the reservoir in the enhanced-oil-recovery stage but also can provide information about the time that fluid takes to move from one point (injection well) to another point in the reservoir. This time is termed the mean residence time. In addition, any flow barrier or directional thief zones such as faults can be identified by delayed tracer recovery. The mass-balance technique can be used to calculate the amount of tracer recovered to distinguish between the existence of a fault, a flow barrier, or a low-permeability zone.

Tracer-Project Design

Interwell tracer testing consists of injecting chemical tracers into injection wells at the beginning of a flood project or after the reservoir has reached its fill-up condition, depending on the objective of the project, and subsequent sampling of production wells for a prescribed period of time, which also depends on the objective of the project. Samples are analyzed for tracer content, which will delineate communication between the injection and production wells. The time during which samples are collected and analyzed depends greatly on the objective of the project. If the objective is to identify thief zones, deduce mean residence time, and determine other factors that could lead to full-field reservoir characterization that could be fed directly into a simulation model, sample collection and analysis must continue for a considerable period of time after tracer is first detected in order to establish a more-defined elution curve.

Gas-Tracer Application

Field Background. The subject field is an offshore oil field located in the Arabian Gulf that has been on production since the 1960s. After the initial natural-­depletion phase, during the 1970s, crestal dump-flood water injection was carried out to maintain the reservoir pressure. During the 1980s, peripheral water injection was introduced to serve as the principal source of energy to arrest reservoir-pressure decline. In 2006, a crestal gas-injection project was initiated to support production from the wells lying away from the periphery.

Geologically, the field belongs to the Lower Cretaceous and is divided into six distinct anticline layers; for purposes of monitoring reservoir performance, each layer is further divided into six sectors, as shown in Fig. 1 (above). The division of sectors arose out of operational convenience and was not defined by any special characteristics of the layers in the reservoir.

Project Objective. As part of the crestal gas-injection project, six gas injectors were drilled in the crest of the reservoirs; hence, each layer was completed with one dedicated gas injector. As part of the reservoir-management conformance reporting, for the sake of simplicity, sector voidage-replacement-ratio calculations assumed equal gas-injection split among all six sectors corresponding to each reservoir layer. However, the imprecision of this assumption was well-understood. Accordingly, a tracer study was initiated to track the injected-gas movement and hence to find out the effective distribution pattern of the injected gas in different sectors within each reservoir layer.

Tracer-Job Design. A qualified gas tracer is one that approximately follows the fluid front with minimal slip velocity and whose chemical composition is compatible with the formation salts. Building on the project objectives, six different tracers were selected for injection through the crestal gas injectors. The target area of investigation was estimated at a couple of hundred square kilometers on the basis of the shallowest oil-producer locations and the fluid-movement predictions.

To calculate the required tracer quantities, simulation sensitivity results were integrated to analyze the fluid-front movement. On the basis of the outcome, volumetric calculation was carried out while considering an appropriate tracer-dilution factor.

Sampling Program and Candidate Selection. A simulation study on gas saturation was conducted to track the gas movement over the period of time in which hydrogen sulfide was introduced into the compositional model as a tracing marker to monitor the gas movement in each reservoir. Time-lapse analysis showed that gas moved along the preferential path in the southwest direction (i.e., toward the pressure-sink area).

Overall, 54 sampling candidates were selected in six reservoir layers to cover the entire area of investigation. With the review of the simulation results together with field-production history and well spacing, it was anticipated to have tracer breakthrough into the targeted producers after 3 years; accordingly, the sampling campaign started in 2010.

Project Execution and Data Collection. To extract the maximum value of the tracer data, each sample was analyzed for all six tracers; the additional analyses were intended to ensure injection-well integrity and to negate the possibility of any interlayer communication.

During the course of the project, revised reservoir-management guidelines were introduced in which the oil producers having gas breakthrough had to be shut in for the purposes of conserving energy and transmitting the pressure support to the needy areas of the field. With the implementation of this revised strategy, several wells with crestal-gas-injection breakthrough were forced to be shut in even before the probable gas-­tracer arrival, which probably affected the potential deliverables of the tracer study.

Until 2015, several samples were collected and analyzed per the set plan; out of 54 wells, tracer breakthrough was observed only in three. Interestingly, all three wells belonged to the deepest reservoir (i.e., Layer 6).

Tracer-Data Analysis and Results. Tracer detection in a given sample is an indication of direct communication between the injection well and the corresponding monitoring wells. However, the extent of communication depends on the breakthrough times and tracer-concentration values.

Data Integration. While the crestal-drive gas injection commenced in 2004, gas tracers were introduced in the six crestal gas injectors across six different layers in August 2006 after a steady front was established. Until 2016, only three wells in the deepest layer had experienced breakthroughs lying on the northern part of the field. The simulation study did predict the gas movement toward the pressure sink. However, no other shallower layers had experienced breakthroughs.

Gas tracers tend to partition between different hydrocarbon phases. While Layers 3 and 6 had no gas caps, the movement of the tracers was expected to be fastest in these layers because no possible partitioning caused by the presence of the dual phase was anticipated. This was indeed observed in the case of Layer 6.

Reservoir complexity could be another reason for the delayed breakthrough. From the logs, it could be seen clearly that top Layers 1, 2, and 3 have a higher degree of homogeneity when compared with the bottom three layers. Accordingly, Layers 4, 5, and 6 were expected to see a faster breakthrough when compared with the top layers. While this is evident in the case of Layer 6 (the only layer that saw a breakthrough), Layers 4 and 5 did not see any breakthrough at all.

It was found that the closure of the high-gas/oil-ratio wells in Layers 4 and 5 resulting from revised reservoir-management guidelines led to an apparent disruption of the pressure streamlines within the reservoir. The pressure-differential paths that the tracers were initially following were modified on account of the closure of a few wells; this led the chemical entities to reroute their path toward a new pressure sink within the closed system.


The tracer studies for this reservoir in the subject case study had to be terminated (owing to long incubation time of tracers) before substantial amounts of tracer could be seen. If the tracer elution could be completed with a finite amount of data, considerable information about the reservoir, both qualitatively and quantitatively, could be generated.

It is well-understood that tracer studies take a considerable amount of time, and a constant production behavior is almost impossible to achieve for most practical purposes. Strategic planning of wells serves as a key solution in such cases.

Simulation serves as a key to the modeling of a tracer study. However, as is the case here, the concept of “dynamic simulation” should be used in monitoring a tracer project. Any change that is taking place in the field should be updated periodically in the model to gauge the actual tracer movement in the reservoir.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 183413, “A to Z for Gas Tracers: A Decade of Learning and Experience,” by M.N. Khan, H. Iwama, A. Al-Neaimi, and O. Al-Shehhi, Abu Dhabi Marine Operating Company, and M. Chatterjee, Tracerco, prepared for the 2016 Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, 7–10 November. The paper has not been peer reviewed.