An Experimental Study for Mechanical Formation Damage
- A.C. Soares (Petrobras Research Center) | F.H. Ferreira (Petrobras Research Center) | E.A. Vargas Jr. (Catholic U. of Rio de Janeiro)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- December 2002
- Document Type
- Journal Paper
- 480 - 487
- 2002. Society of Petroleum Engineers
- 1.6 Drilling Operations, 5.3.4 Integration of geomechanics in models, 1.2.2 Geomechanics, 1.6.9 Coring, Fishing, 3.2.5 Produced Sand / Solids Management and Control, 3.2.3 Hydraulic Fracturing Design, Implementation and Optimisation, 4.1.2 Separation and Treating, 1.8 Formation Damage, 5.5 Reservoir Simulation, 4.1.5 Processing Equipment, 2.4.3 Sand/Solids Control
- 3 in the last 30 days
- 501 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
This work presents triaxial experiments that were carried out in different stress paths, with measurements of P- and S-wave acoustic velocity and permeability, to evaluate mechanical formation damage. In some reservoirs, large deformations occur during oil production because of the effective stress increase. This may cause permanent damage to the reservoir as the rock structure changes, lowering the permeability and the final oil recovery. The experiments showed that plastic behavior is always present and that a "cap" model may represent the yield region. This methodology allowed a near-wellbore evaluation and permeability investigations using a stress-path concept.
Usually, reservoir engineers do not take into account the effect of in-situ stresses on production, considering that permeability is constant during reservoir oil production. Nevertheless, this assumption is not valid for limestone and unconsolidated sandstone (e.g., deepwater reservoirs in the Campos Basin), and a better understanding of the phenomenon is required.
Our initial experiments were used for pore-collapse studies1-3 on a Campos Basin limestone. Some fields presented an additional decrease of oil production, which probably was caused by pore collapse. The oil recovery of the fields was lower than expected. This limestone had presented a very ductile behavior in rock mechanics tests for hydraulic fracturing analysis. Based on laboratory results and field observations, the pore-collapse hypothesis became stronger. Therefore, it was decided to perform further experiments based on the same methodology employed by Smits et al.1
Because the highest variation on stress state occurs near the wellbore, and the previously used experimental methodology did not represent this region, it became necessary to establish a new procedure for a better understanding of the stress-strain-permeability behavior on production.
A series of experiments was then planned for another Campos Basin limestone field. These tests were to be performed under different stress paths. Permeability and ultrasonic wave velocity were measured simultaneously. This new methodology would make it possible to identify pore-collapse occurrence throughout the reservoir.
The analyses of the experiments were based on a cap model. Plastic strain was present along all the stress paths, and permeability changed constantly before reaching the cap.
Based on the results of the geomechanical experiments, a function relating permeability and pore pressure was fitted to experimental points. Additionally, a computational model was developed to evaluate the mechanical damage. This code considers a well in an infinite reservoir and permeability as a function of pore pressure.
The initial experiments were performed with samples from a field in the Campos Basin. The oil production and recovery values were lower than the ones predicted by reservoir simulation. A ductile, heterogeneous, porous limestone (20 to 35% porosity) comprised the reservoir.
The pore-collapse study for this field was carried out with Smits et al.'s1 experimental procedures for Ekofisk field. The aim was to obtain a trend line based on a 1D compression test, well known in soil mechanics as an oedometer or uniaxial strain test. The test consists of applying an axial load while strain in the horizontal directions is prevented. Therefore, the axial strain is exactly equal to the volumetric strain. The strain condition of the uniaxial strain test is similar to the condition that occurs in the field during production4 (i.e., there are no horizontal displacements in the formation). This statement assumes that sedimentary basins are horizontally confined. In these experiments only compressional ultrasonic wave velocities (P-wave) were measured. This was done to identify a possible rearrangement of the rock structure through sound velocity variation. Permeability was also measured, and it was possible to observe its dependence on the settlement process. Because of limitations in the triaxial cell, permeability was measured only in the axial direction. An inert oil was used as pore fluid to avoid fluid/rock interaction. The tests were performed in a rock mechanics test system with 2700 KN of axial load capacity and 82.5 MPa of confinement pressure.
The result of a typical uniaxial strain compaction experiment in which pore collapse was observed is shown in Fig. 1.
The stress-strain diagram is characterized by nonlinear behavior at the beginning of the test. At this interval, an increase in the P-wave velocity was observed because of the closure of fissures5 generated by removing the core from its original in-situ stress conditions. A pronounced decline in permeability was also observed at this point. Soon after, the stress-strain curves showed linear behavior. For this linear portion, the slope of the P-wave velocity vs. the axial stress curve decreased. On the other hand, the permeability continued to decrease, but at a smaller rate. Close to collapse, the stress-strain curves turned nonlinear, with larger deformation for a small variation in axial stress. At this moment, the P-wave velocity was almost constant and was approximately equal to 4000 m/s, indicating a rearrangement of the pore structure, while the permeability decreasing rate was higher. Finally, the stress-strain curves reached a new linear configuration, showing that the structure had reached a new state of equilibrium. For this situation, the increase of the P-wave velocity rate became higher, and the permeability was close to zero. The pore-collapse stress was defined at the beginning of yielding in the stress-strain curve (approximately 60 MPa).
The uniaxial strain tests performed showed, at first glance, a great dispersion of results. However, when plotted and grouped according to the samples' initial porosity, the results showed good agreement, and the data formed a logical sequence. It was observed that the pore-collapse stress was related to the initial porosity. Therefore, the larger the porosity, the smaller the stress that leads to pore collapse. When putting all the axial stress-strain curves together in the same graph, a trend line could be identified. This concept is presented in Ref. 1, with examples of some fields where this type of study was done. Fig. 2 compiles the results of our uniaxial tests.
|File Size||1 MB||Number of Pages||8|