The different mechanisms for the recovery of hydrocarbons from reservoirs has usually been divided into three types: primary, secondary and tertiary recovery mechanisms. Although this nomenclature suggests a chronological order among these stages, their application is subject to the specific characteristics of the reservoir and to the field production strategy. However, primary recovery is generally considered the very first process to occur.

First, the primary recovery mechanism consists in the displacement of hydrocarbons due to the reservoir natural energy sources, such as solution gas drive, gas-cap drive, natural water drive, fluid and rock expansion, gravity drainage (Don W. Green, 1998). For further information on this topic, please refer to our previous article “Drive Mechanisms in Reservoirs”.

Typically, secondary recovery methods are applied when the primary mechanisms are no longer sufficient to maintain production in economically favourable levels. Hence, artificial methods are necessary in order to promote an increase in the reservoir energy. This happens when the reservoir natural energy has been drained to the extent at which it becomes impossible, or uneconomical, to lift hydrocarbons from the reservoir to the surface. Examples of secondary mechanisms include water flooding, gas lift injection, gas-cap expansion, immiscible gas injection. The latter, although being considered a secondary recovery process, is rarely used due to the great viscosity and density disparity between the injected phase and the reservoir fluids.

One could also refer to this viscosity difference as a key cause of unfavourable mobility ratio, which is one of the most important factors influencing macroscopic volumetric displacement efficiency of an EOR operation. This efficiency measurement consists, basically, in the reservoir pore volume fraction that has been contacted by the injected fluid (Don W. Green, 1998).

The mobility ratio, in turn, is the relation between the displacing fluid mobility and the displaced fluid mobility. In the context of a water flooding procedure, for example, the mobility ratio can be defined according to the equation below.

MobilityRatio

Where krw and kro are the relative permeability of water and oil, and μw and μo the viscosity of oil and water, respectively. This ratio controls the stability of the displacement front, in the sense that when its value is higher than 1, viscous fingering occurs (Guillen, Carvalho, & Alvarado, 2012). On the other hand, if this ratio is lower than 1, the likelihood of a more uniform displacement front shape is higher.

To circumvent the shortcomings derived from techniques prone to low mobility ratios, other methods have been developed targeting a more favourable interaction between the injected fluid and the rock-fluid system. These methods generally employ chemical mechanisms in combination with the physical ones to reach the lowest possible value for residual oil saturation.

Hence, tertiary processes use the injection of miscible gases, chemicals, and application of thermal energy, mainly, aiming to reduce the mobility ratio by: lowering interfacial tension, promoting oil swelling and wettability modification, reducing oil viscosity. As such, these practices benefit the displacement of additional hydrocarbons commonly after the application of secondary methods. It is worth noting that the term tertiary recovery has been preferably referred to as “Enhanced Oil Recovery”, or EOR, once the three stages of recovery may not occur in the specified order, as previously mentioned. For example, for heavy crude oil reservoirs where it is not possible to profitably produce from primary methods due to high oil viscosity, EOR techniques are applied directly. In these cases, it is a common practice to inject steam into the reservoir in order to reduce oil viscosity; interfacial tension between asphaltenes and paraffins to the rock; distill crude oil light components that in turn can condense in the oil bank, reducing its viscosity and helping to remove trapped oil.

Another type of tertiary method widely applied in the industry is polymer flooding, or polymer injection. Aiming to demonstrate such practice, the results of a simple reservoir simulation were analysed using the Kraken software. Results were originally generated by the UTCHEM reservoir simulator. The goal was to represent the conditions of a polymer flooding operation between two deviated wells, which happens in a reservoir limited by no-flow boundaries that could be interpreted as a sector of fault block. In the case considered, a 1500 ppm solution of xanthan gum is the chosen injection fluid to be pumped into the reservoir from the second year on.

The grid dimensions, well trajectories and initial oil saturation are presented in Figure 1.

Reservoir 3D model showing wells trajectories, oil-water contact (OWC), dimensions and initial oil saturation (Soi).

Figure 1: Reservoir 3D model showing wells trajectories, oil-water contact (OWC), dimensions and initial oil saturation (Soi).

In Figure 1, it is possible to observe the presence of an aquifer with OWC at approximately 7890 ft and that the reservoir initial oil saturation is 0.951. Also, it is important to note the wells location, where the injector is positioned lower than the producer, causing the fluid to flow preferably upward. This aquifer will provide a certain level of pressure support, mainly during the first year of production, after which water injection begins, increasing average reservoir pressure and improving production rates, as presented in Figure 2. From the second year onward, polymer injection occurs, contributing to oil displacement.

 

Figure 2: Average pressure, volume flow rate and polymer fluid concentration versus time, in years. The first chart displays the average pressure for the cells in the model corresponding to the reservoir (water level up), injector well, the entire model (main grid) and only the aquifer. The second chart shows the total injection flow rate (Hor-Inj), the total production flow rate (Hor-Pro) and the average total polymer concentration in the cells comprising the injector well.

Figure 2: Average pressure, volume flow rate and polymer fluid concentration versus time, in years. The first chart displays the average pressure for the cells in the model corresponding to the reservoir (water level up), injector well, the entire model (main grid) and only the aquifer. The second chart shows the total injection flow rate (Hor-Inj), the total production flow rate (Hor-Pro) and the average total polymer concentration in the cells comprising the injector well.

Just as the previously described tertiary process, the steam injection, polymer flooding also improves recovery by reducing the mobility ratio. However, it does that differently, by increasing the viscosity of the injected water, and, in many cases, reducing the water effective permeability. Aqueous polymer solutions tend to exhibit non-Newtonian viscosity properties, meaning that an increase in shear rate does not necessarily cause a proportional increase in viscosity. The polymer molecule size and extension are key for determining its viscosity-enhancing properties as well as to estimate the polymer retention by adsorption and entrapment, an important factor for the polymer propagation rate.

Taking that into account, by altering the mobility ratio, a more uniform displacement front is likely to be formed, thereby diminishing the chance of early water breakthrough. The histogram in Figure 3 shows that, at the end of the simulation, when the displacement front is likely to be at its most advanced stage and water fractional flow (fw) is likely to be the highest, approximately 59% of the cells above the aquifer level present fw values between 0.9 to 1.

 

Figure 3: Cumulative histogram showing distribution of water fractional flow values at the last simulation timestep.

Figure 3: Cumulative histogram showing distribution of water fractional flow values at the last simulation timestep.

Therefore, by selecting only those cells in the model that meet this criterion of 90% to 100% water fractional flow, it is possible to generate a good visualization of the displacement front. The animation below presents an estimated view of the polymer advancing front from the time injection starts to the end of the simulation. Also, it is possible to notice the aquifer water influx, which is promoted by the increase in aquifer average pressure due to injection.

Using only the filtered cells, it was possible to calculate the average displacement efficiency (Ed), which is the oil fraction that has been displaced by the injection fluid front. This value is presented in Figure 4 along with other measurements.

Ed

Figure 4: As polymer injection starts and its average concentration rises, average water viscosity increases, causing better displacement efficiency and lower average oil saturation. Also, it is interesting to notice that the higher the water fractional flow, the higher is the average displacement efficiency.

Figure 4: As polymer injection starts and its average concentration rises, average water viscosity increases, causing better displacement efficiency and lower average oil saturation. Also, it is interesting to notice that the higher the water fractional flow, the higher is the average displacement efficiency.

Finally, for our example, the correlation between polymer concentration increase and the rise in water viscosity becomes clear by analysing Figure 4. Also, It is possible to observe that the increase in average viscosity leads to a more favourable displacement efficiency, from approximately 17.5% in year two, to roughly 65.0% at the end of the simulation.

 

References

Don W. Green, G. P. (1998). Enhanced Oil Recovery. Richardson, Texas: Henry L. Doherty Memorial Fund of AIME, Society of Petroleum Engineers, 1998.

Guillen, V., Carvalho, M., & Alvarado, V. (2012, April 12). Pore Scale and Macroscopic Displacement Mechanisms in Emulsion Flooding. Transport in Porous Media, pp. 94-197.