IV1.2.1. Case study 1

Figure 96 shows a well pattern of a reservoir sited in southern Argentina, where an interwell study by means of radiotracers was performed some years ago.

Since HTO was the selected tracer, the liquid scintillation technique was used for measurement. Because of operative limitations in the laboratory measurement, samples were not distilled before counting and, in addition, a short count time was used. For that reason the detection limit was much higher than usual. The detection parameters were:

• Background = 20 cpm;

• Efficiency = 0.28 (counts/disintegration);

• Count time = 10 min;

• Volume of the sample = 8 mL;

• Detection limit = 29.5 Bq/L.

From the detection limit, the mean output concentration was fixed at ten times this value (295 Bq/L), which leads to an activity of 167 GBq (4.5 Ci). In fact, 10 Ci of HTO was injected into well K-22 using the bypass device. Figure 97 shows an example of the tracer concentration and cumulative response curves for well K-329, whose output was followed during a full year and belongs to the K-22 pattern.

The first information obtained in production well K-329 from a quick analysis of the response curve using Anduril software was:

FIG. 97. Instantaneous (left hand scale) and cumulative (right hand scale) response curves (well K-329).

• Breakthrough = 86 d;

• Mean residence time = 193 d;

• Final time = 321 d;

• Tracer recovery = 9.2%.

The distance between wells K-22 and K-329 is 251 m, thus, the calculated minimum, medium and maximum water velocities are 0.78 m/d, 1.3 m/d and 2.9 m/d, respectively. Permeability was also evaluated by Anduril software using Darcy’s law. A value of 282 mD was obtained, which appeared reasonable to reservoir engineers.

Anduril software was used to model the experimental response curve obtained by sampling in production well K-329. A radial dispersion model was applied. Figure 98 indicates a good fit.

The model gives the following parameter values for the dynamics of tracer movement between injection well K-22 and production well K-329:

• Breakthrough = 75 d;

• Mean residence time = 210 d;

• Final time = 410 d;

• Tracer recovery = 9.1%.

Cumulative injected volume (m5)

The volumetric response curve for well K-329 appears in Fig. 99, given in terms of the cumulative injected volume. Reservoir information extracted from the tracer response curve in well K-329 is:

• Breakthrough = 86 d;

• Mean volume = 19.073 m3;

• Swept volume = 1.775 m3.

The last value is the pore volume swept from the injector to the production well (K-329) and equals the mean volume multiplied by the recuperation factor (0.092).

IV 1.2.2. Case study 2

Complex response curves can be obtained in some cases due to the mixed response from different layers. Anduril software can be used to decompose a complex response curve into several simple curves. These simple curves are supposed to represent tracer movement in a unique layer. Modelling each of the simple curves provides the mean residence times and the quantity of tracer recovered from each layer.

An example of this methodology is shown in Fig. 100 in which the complex response of a production well was approached using four simple functions (Fig. 101) based on the radial model. A possible explanation of the tracer behaviour could be that it reached the production well by following four paths of different permeability belonging to a unique layer.

The parameters of each function are given in Table 26.

TABLE 26. COMPLEX RESPONSE MODELLED BY FOUR FUNCTIONS Curve D1lvx ^mean (d) fi 1 0.020 87 0.292 2 0.004 120 0.105 3 0.020 190 0.384 4 0.008 300 0.219

Some 46% of the total quantity of tracer recovered in this well related to the injected activity. The parameter f is the contribution of each path expressed as a fraction of that percentage and was evaluated from the area under each curve. Dispersivities may be calculated by multiplying the ratio D1lvx by the distance between the injection and the production wells.