The five spot set-up (Fig. 117) has been designed using a large rectangular vessel of 160 cm x 95 cm x 80 cm. It has been filled manually with sand with a density of 1.8 g/cm3 and an average granulometry of 0.75 mm (d10 = 0.5 mm, d50 = 0.75 mm, d90 = 0.90 mm). The size of the five spot is only 40 cm square to avoid boundary effects with 5 wells of 40 mm diameter.

It should be noted that the studied five spot used to analyse this effect is not symmetrical. The bed of sand is initially saturated with water.

Four experiments have been conducted, three experiments with stagnant water inside the bed and flow rates of 107 cm3/min (exp. 1), 107 cm3/min (exp. 2), 184 cm3/min (exp. 3) at the injected wells. The fourth experiment has been conducted with a linear velocity of the water table of 2 m/d. The experimental data treatment has been revised; curves have been time and area

FIG 117. Photo and scheme of the five spot set-up.

normalized in order to overcome the apparent discrepancy due to the non­homogenous radial injection. It should be noted that interpretation of tracer experiments in oilfields should take into account the near impossibility of having uniform radial distribution of the tracer injection. In some extreme cases, this may explain the low tracer recovery observed.

Figure 118 shows the superposition of the tracer responses at well 1 for the three static experiments and one of the curves obtained for well 4.