Evaluation of matrix stimulation operations

As in hydraulic fractures, it is important to evaluate the matrix treatments. The methodology for evaluating a fracture through a pre- and post-treatment analysis is fully accepted, but for matrix treatments this process is not economically justified in most cases.

Efficiency is frequently measured only by an apparent increase in the productivity index, without doing a post-treatment test. Looking only at the increase in productivity (or injection) rate will not necessarily give a correct picture of the optimal conditions for the well.

The skin damage can be determined with a post-operation well test, but its mechanical components may be an important part, and a correct analysis should determine them. Matrix treatment can only remove the skin portion due to formation damage.

These analyzes are post-mortem and allow us to improve the design of future work in neighboring wells, but not optimize the current treatment. So it is advisable to do a real-time evaluation in the same location.

This allows you to determine if the skin was completely removed, and thus maximize your investment. Several authors designed methods to approximately evaluate the efficiency of a treatment by monitoring the evolution of the skin on location.

These monitoring indicate whether it is necessary to modify the program, increasing or reducing mixture volumes, and help improve designs for future wells in the same reservoir.

The real-time evaluation is done in the same location, recording and interpreting the pressure and flow parameters with some software. The objective is to observe the decrease in the skin to adjust the volume of each stage.

Having the well data, by simply recording pressure and flow, the skin value can be estimated at each moment. Although evaluation techniques have existed for quite some time, they are not frequently applied in the field because:

• Lack of training for supervisors in service companies and oil companies.
• Lack of awareness of the importance of matrix treatments (rather it would be said: It's just a pump!).
• Lack of computer on location (not all companies give their supervisors a PC).
• Generally the fluids are premixed and therefore there is no flexibility.

In most applications the skin should decrease as the fluid penetrates the matrix, indicating that the treatment is improving well productivity.

Due to the convergence of fluids produced in the matrix, damage near the wellbore has a much greater effect on the skin value than deep damage. Therefore, skin decrease will normally be very rapid once the first fluid enters the matrix.

As the near-wellbore damage is removed, the reduction in skin becomes slower, and eventually stabilizes. At this point, the operator should consider transitioning to the next stage or employing divergence (if it is deemed that only a portion of the targeted zone accepted the fluid).

The goal of treatment should be to reduce damage to zero, or to an irreducible minimum. In the case of carbonates where permeability near the wellbore may be increased, the skin value may be slightly negative, which would indicate an increase in the effective diameter of the wellbore.

A significant increase in the skin will serve as an alarm to the supervisor warning him that the treatment is harmful to the training (unless it is due to the effect of a divergent). If this happens, the treatment must be stopped and the design reviewed.

If we consider the three examples in the graph, a real-time evaluation would have allowed decisions to be made in the well regarding the volumes and types of fluids to be pumped:

Well A: Fluid #1 shows little effect on the damage, the volume could be reduced to move to fluid #2 (a minimum volume to guarantee a spacer effect). Fluid #2 is very effective in reducing damage, and the volume used is what is necessary.

Well B: Both fluids are quite effective, but it would have been advisable to increase the volume of fluid #1 until stabilization was seen.

Well C: Fluid #1 is very effective for damage reduction. Since fluid #1 has reduced the skin to zero, it is not justified to continue the treatment by pumping fluid #2.

Several authors proposed techniques to evaluate in real time, some of the techniques are just improvements of previous techniques. Today they are used by commercial software. We will quote some:

• The Paccaloni technique uses the instantaneous pressure and flow rate during pumping to calculate the skin. This method is based on Darcy's Law for steady state, single phase, horizontal flow.
• The Prouvost and Economides technique takes into consideration that the fluids are injected under a transient flow regime (transient).
• Behena's technique is an improvement on the previous one.
• Hill and Zhu's technique is a combination of the Paccaloni and Prouvost techniques.

The Paccaloni technique uses point values ​​of instantaneous pressure and flow at different times during pumping to calculate the skin. This method is based on the Darcy equation for steady state, single phase, horizontal flow:

Considering that kh is known, the bottom-hole pressure can be estimated for each flow value and for any Skin value. Paccaloni uses the concept of damage ratio:

This technique is easy to use in the field but has two important limitations:

1. A possible misinterpretation of some flow changes as skin changes.
2. In early time, the pressure response is more transient than “steady test”.

A series of P curves are plotted for a range of injection flow rate and mouth pressure. As the fluid is injected, pressure and flow values ​​are manually placed on the graph indicating the progress of the treatment.

This technique was a very good approach when it was devised, as back then, nobody brought a computer to the location. Nowadays, other techniques are more recommended.

The basic concept of the Prouvost and Economides technique relies on the difference between the pressure response of a well with damage and that of a simulated well without damage (with skin = 0).

The response is a function of time, as the damage changes as a reactive fluid passes through the damage zone. As it is assumed that the other parameters remain the same, any difference between the pressures is due to a difference in the skin.

Where:

• Psim = Simulated pressure
• Pres = Reservoir pressure
• Pmeasure(t) = Pressure measured at time t
• qsf = Rate

The treatment is completed when the two pressure curves overlap. The application of this technique requires the use of a PC on location, with the appropriate software. The efficiency of the divergence is measured by verifying that:

• The entry of the divergent corresponds to an increase in the skin.
• The entry of the acid after the divergent corresponds to a decrease in the skin.
• The skin never reaches zero (0).

If the different stages of divergent have covered the entire area to be treated, the last acid should not show skin reduction effect.

Unfortunately, in most treatments the effects of the divergent are not seen as well. Today it is recommended not to divide the treatment into more than three acid stages (that is, two divergent stages).

The use of tracers in the different stages can provide good information about the efficiency of the treatment. In the example the well was stimulated twice. The reservoir is a 13 t/m (4 spf) perforating limestone at a depth of 2104/29 m (6902/82 ft).

The first treatment was 3 stages of 2000 gal each with two stages of salt as divergent. Scandium-46 was put in the acid, iridium-192 in the first divergent slug and antimony-124 in the second.

The profile shows that the diverger was very inefficient and that almost all of the fluid entered the upper zone.

The second treatment was with three acid stages (1000/1000/1500 gal), with two salt slugs as divergent. The upper zone was isolated with a packer at 2099 m (6886 ft) and treated alone.

This stage was labeled with iridium-192, the other two layers were treated together in two stages. The first acid step was labeled with candium-46, and the second with antimony-124.

In this case a correct divergence was achieved. But this example also shows that mechanical divergence is much more efficient than chemical divergence.

The well is an injector, the treatment was pumped through a 1-¼” CT. The formation is a sandstone with low solubility in HCl. Registered:

• The pressure at the CT inlet. The values ​​are high due to the friction in this small diameter tube.
• The pressure in the annulus between CT and tubing, with the BOP closed. Thus, the real bottom pressure in front of the mandrel can be estimated at every moment: annular pressure plus hydrostatic pressure.
• Pumping flow.

The preflush HCl reacted slowly with the formation, but the volume used did not allow us to see a stabilization of the pressure (and therefore the skin).

The pressure behavior suggests that not all of the soluble HCl material has been dissolved. There may be carbonate left. The HF reacted very quickly, and therefore removed damage very close to the well.

For this, the volume used was greater than necessary, and the treatment could have been completed sooner.