While Figure 1 shows a situation that may not be desirable, because coverage of all zones is not uniform, you can rest assured that this situation does not occur often because there will likely be crossflow (pressure and fluid-flow communication between layers). Figure 2 is a schematic of how crossflow can occur.

With crossflow, such as shown schematically in Figure 2, and also in Figure 3, flow can be diverted to zones where there is no hydrocarbon to be swept. Figure 3 shows incipient crossflow from higher to lower permeability zones at a relatively short distance from the wellbore.

Water breakthrough in the high permeability layer has serious consequences. The water mobility and transmissibility will skyrocket in the high permeability layer due to relative permeability effects. Then, oil recovery and sweep efficiency can be seriously impacted.

To the PWRI specialist, crossflow can have three manifestations. There can be horizontal crossflow on shut-in, occurring at the wellbore. This has been demonstrated on Heidrun. There can be near-wellbore vertical interlayer flow, as is shown in Figure 3. There can also be crossflow deep in the reservoir due to permeability contrasts, or in-situ inhomogeneities. Figure 4 is an example, showing movement of fluid to the water zone (perforated only in the oil zone) because of more favorable relative permeability and gravity. Note in this example that flow is to a higher permeability zone.

The situation in PWRI is frequently more complicated since some form of vertical fracture may have often formed. Figure 5 schematically shows diversion of fluid to a lower permeability zone because of fracturing. Injectivity may appear to be adequate, but this cannot be considered exclusive of sweep efficiency. The inverse situation can also occur after water breakthrough in a high permeability layer.

The Conformance Crossflow Consideration

Beyond the normal operational and near-wellbore considerations when designing a PWRI project, there is an important and fundamental conformance crossflow consideration that should be addressed when designing and implementing a PWRI project involving a layered reservoir. If the formation for PWRI injection is layered without crossflow between the layers and there is good cement behind pipe and if water injection is being conducted at a pressure below which hydraulic fracturing will occur and there exits is an unfavorable mobility ratio, then where feasible and possible, the water injection rate should be proportioned into each reservoir layer in a manner to maximize and optimize oil recovery from the well pattern. Thus, in this manner, the design of the PWRI project can have a significant impact on the reservoir-engineering question of the inter-well oil recovery factor.

However, if the water is injected at sufficiently high pressure such that hydraulic fracturing occurs (e.g., thermally induced vertical fractures), then there is little that can be done conventionally to control the relative water injection rate into the various reservoir intervals. Thus, little can be done during the PWRI injection design to affect the inter-well sweep efficiency and the reservoir-engineering question of the inter-well oil recovery factor.

Likewise, if there is crossflow between the reservoir layers, then there is little that can be done conventionally during PWRI injection to control the relative water injection rate into the various reservoir intervals. Thus, in this case too, little can be done during PWRI injection to influence sweep efficiency and ultimate inter-well oil recovery.

The Challenge

You can imagine that there are a large number of parameters that can impact and that affect optimizing injectivity concurrently with hydrocarbon recovery. There are some methods that might be adopted for mitigating future problems. Some of these are described in the following section. One of the outcomes of the Layered Formations Workshop was to get operator input on how they would mitigate future conformance problems (to maintain injectivity and sweep). To restrict this exercise, four generic formation scenarios were delineated. These are shown in Table 1.

Table 1.   Generic Scenarios

Table 1 shows four generic injection target scenarios, each with two zones of different combinations of mechanical properties (hard, soft) and different permeabilities (low, high). What would you do to avoid future remedial activities in each of these cases? Remember NO FUTURE INTERVENTION is possible for this exercise. There may or may not be a possibility for in-well, near-well, or far-well reservoir crossflow. These possibilities are expected to have a significant impact on the following exercise. What are the practical means to ensure injection coverage of, and good conformance for, these reservoir packages? "How do you control the well during its life?" Part of this relates to starting the operations out correctly! "Are there situations where mechanical differences between formations can be overcome without using mechanical means for isolation? If so, how are these differences overcome?"

Also remember that there is a REQUIREMENT FOR MULTIPLE ENTRY POINTS. EVERYTHING ELSE WILL REQUIRE SACRIFICNG SOME PRODUCTION. The presumption is that injectivity and recovery require flow in both zones. CONFROMANCE IS AN IMPORTANT ISSUE AND AN IMPORTANT CONSEQUENCE OF THE DESIGN AND MANAGEMENT PROCESS.

Possibilities

The following items describe some of the possibilities that you could consider (upfront) to mitigate future injectivity and sweep difficulties.

1. "Brute Force Drilling:" If the economics favor it, an extreme possibility is to drill individual wells for targeting specific zones - dedicated injection per zone - not necessarily separate or individual wells.
2. Dual Completions: There are various variations on this theme. They can vary from isolation hardware (Y-block) to dual strings (long string/short string) to parasite strings (e.g., coiled tubing down the annulus) - long string/short string, dual string. The dual string solution avoids drilling a multilateral, but may still be too expensive.
3. Cased Multilaterals: Cased multilaterals could be used where control is possible. Individual legs could target individual zones. Controls might be required to be able to partition flow into specific legs. One possibility is to use multi-layer legs passing through a high permeability zone down into a low permeability zone. If the high permeability zone is below the low permeability zone, mechanical devices might be more appropriate.

Dual laterals are currently used on Magnus. The economic decisions may come down to how sustainable the injection needs to be over the life of the reservoir. Sometimes you find that over the life of a field you may not have to inject as much for voidage replacement as earlier on in the life of the reservoir.

4. Openhole Multilaterals: You can envision circumstances where openhole multilaterals may be a reasonable option. The restrictions could include that the formation integrity along the lengths and at junctions must be adequate to avoid collapse. Also, consider, either in cased or openhole multilaterals, whether or not individual controls are required or whether the target zones are adequately isolated naturally.
5. Drainholes: Can a single, horizontal, or at least optimized trajectory, drainhole be used? To optimize conformance, it may be possible to use different length drainholes in different zones and/or multiple drain holes.
6. Multi-leg Drainholes: Are multi-leg drainholes a possibility? They don't need to have as large a diameter to accommodate the flow control valves that might be required in a single horizontal drain hole. Soft rock may be easier for drainhole implementation because drilling is cheaper. Fracing in soft sands may be an issue.
7. Sidetracking: While there can be difficulties in monitoring, sidetracking may have economic advantages.
8. Wellbore Trajectory: There are demonstrated situations in Prudhoe Bay where there were substantial near-wellbore pressure losses in fractures due to tortuosity. Up to 400 psi differential pressure has been seen in wells with up to 60° deviation. Some of these can be overcome and used to your advantage by drilling in preferential directions, to make it more difficult to initiate at the heel.
9. "Wandering (snaking) Wells": With an appropriate trajectory, it may be possible to force a well to fracture in some places and not in others.
10. Climbing Wells: Beyond controlling a well's azimuth (direction relative to true north) you can consider exotic possibilities for wellbore inclination. Consider an unconsolidated, high permeability formation, overlying a tight, hard zone. The tight formation can be reasonably thick. A possible drilling approach is to have a slant well through the high permeability zone into the low permeability and then directionally drill back up into the high permeability zone. One problem with this well configuration is keeping multiple fractures in the low permeability zone from growing upwards into the high permeability zone. Another solution may be multi-laterals in each zone. There may be good injection but poor sweep and it would be difficult to thermally fracture the high permeability zone.

At Prudhoe Bay, in the WPWZ, there are some multilayered situations with crossflow and vertical fracture growth (through the tar mat) may be undesirable. Is there something that you can do during completion? What would be an alternative? One possibility would be to drill a well that is s-shaped and is subhorizontal through the zones above and below the tar mat. One of the problems can be formation damage during drilling as well as difficulty in cleaning perforations. You may want to create multiple fractures in the horizontal sections. The problem then becomes how to create these fractures? Alternatively, your may have to think of using multi-laterals. Regardless, one key in these horizontal sections is how to operate the well to get conformance. One of the difficulties is to keep the fractures from growing down through the tar mat.

11. Backflowing Before Injection Is Started: Backflow can be used as a form of mitigation, to clean perforations and ensure flow in multiple zones at the initiation of the injection operations. Backflow must be applied cautiously (i.e., in stable formations). It was established in the Soft Formations Workshop that backflow in unconsolidated formations is undesirable. Access to lower permeability formations may be encouraged if a backflow rate is established to remove perforation or other completion damage/debris.

Backflowing is an option only in situations where formation integrity is adequate

Examples have been provided (refer to the Layered Formations Workshop) where the importance of cleaning up perforations has been demonstrated. In one situation, the perforations were cleaned up before injection. They were backflowed and solids lifted to be certain that the perforations were clean. The permeability was between 200 and 300 md. The perforations had to be clean to allow water in for cooldown.

In general, especially in softer formations, you will want to have a soft shut-in to avoid water hammer effects. On the other hand, in some gravel pack scenarios, recall that water hammering has shown some possibility for temporary cleanup.

12. Profile Modification: This can be carried out in the injector or in the producer (if a zone is watering out). At the Layered Formations Workshop, there were observations on what is required to seal a hydraulic fracture, as opposed to a natural fracture. "TO SEAL A HYDRAULIC FRACTURE IT IS NECESSARY TO PUT IN GEL AND TO FOLLOW THIS WITH SOMETHING TO PLUG THE PERFORATIONS" (for example, CaCO₃). The argument was that, with a propagating fracture, complete sealing of the fracture was particularly difficult, let alone that experience (for example in cuttings reinjection and conventional refracturing) has suggested that delamination can occur and new injection can simply follow a new path possibly even within the previously created fracture. Therefore, positive shutoff is required.
13. Perforating Practices: One firm recommendation has been that blanket perforating should not be used, because it makes the possibility for effective remedial operations much more difficult. Do you have situations that refute this recommendation?
14. Perforation Pattern/Spacing: In situations where backflow is not desirable, efforts can be made to maximize separation of the perforations to minimize interaction between separate perforation tunnels. This practice has implications beyond backflow alone. If there are issues of formation integrity, they can be exaggerated if perforations are spaced too closely and disaggregated zones interact and are superimposed. In all cases that you might consider, the issue goes back to providing multiple points for fluid entry - fracturing, drain holes. In a low permeability scenario, you can consider creating multiple entry points with focused perforating, staged fractures, etc.
15. Perforation Balance: Underbalanced perforating is sometimes used to minimize damage during perforation. Do you have policies or are there regulatory or economic restrictions? "When should underbalanced perforating be considered?" "How should you decide on the degree of underbalance?" "What are the consequences of the degree of balance?"
16. Stimgun^TM: This, or similar propellant techniques might be considered for multiple fracturing in lower permeability zones.
17. Limited Entry: Controlling perforation size and phasing may encourage/inhibit local flow.
18. Cased Hole Stimulation: Presuming that you don't want to defer production in a high permeability zone, what are the options? You may be able to help partition the flow into the lower permeability zone by hydraulically fracturing (stimulation), or, some chemical stimulation may be appropriate - depending on the lithology. How do you guarantee that the fracture will stay open? Should you use proppant? Some guidance may come from looking at the rates that are achieved and the behavior of offsets.

Fracturing does not always work. For example, BP has tried to put propped fractures in the Sag River to get water in and try to get thermal benefit. They have found that these fractures can plug up when going from seawater to produced water in maybe 12 months. (As an aside to mitigation considerations, BP has also found out that these fractured wells clean up with seawater. Maybe the seawater pushes proppant away. They clean up with ~300 bbls of seawater. Solvents can also be used. Seawater flushes do not give a sustainable benefit and there is a need to flush with seawater every few months.)

19. Fracturing in Openhole: ARCO has reported initiating four or five fractures in openhole. There was an uncemented liner in the openhole and multiple propped fractures were generated. Are there methods for preferential reduction of breakdown pressures?
20. Scouring: Injection of sand slugs (to remove tortuosity) has been successful in some producers.
21. Wedge Fracturing: Some people would claim that with a clever frac you could flood both zones - for example, create a wedge frac by overdisplacing.
22. Higher Equivalent Well Density With Multiple Fracturing: How do you get an equivalently higher well density (multiple fracturing…)? In cuttings reinjection, there is good evidence that swarms of differently oriented fractures can be created at injectors. Can this be done in PWRI by periodic shut-in, etc., to overcome lack of injectivity in some zones or loss or injectivity with time?
23. Mechanical Control: There may be situations where it is only possible to guarantee adequate coverage of all zones with mechanical control. Mechanical control may not be required for the entire injection operation. Consider the following example.

Suppose that you have a long horizontal or an extended reach well where fracture initiation will preferentially occur at the heel. One possibility for encouraging some injection farther along the well is to set two ECPs at the toe, for pressure isolation. Pressure isolation may only required for a short period of time so that initiation of a fracture could be ensured at the toe, either from explicit fracturing or as a result of some cooldown operation.

In some formations and injection scenarios, you may not rely on thermal effects and you may well need mechanical isolation for long-term performance. "What are people thinking of for controlling conformance in soft formations? Are mechanical methods for guaranteeing conformance required from the start, in soft formations, to avoid the costs of remediation or the costs of trying to pre-condition certain layers. For example:

Cases have been reported where the differences in fracture toughness (or stresses) between zones can make it extremely difficult to guarantee conformance. Suppose that there is an upper zone with high stresses and high fracture toughness. A lower zone, below an intervening shale, has lower stresses and a lower K_IC. It might be determined that thermal manipulation of the stresses will not work because the contrast in stresses is too high. This was a situation (an actual BP Amoco example), where conformance was accomplished mechanically. It is unique because the mechanical completion was manipulated thermally (rather than the formation). An inflatable packer was set between the two zones. Produced water was used to keep the inflatable from unseating (i.e. to avoid unseating caused by cooldown). The upper zone was broken down and after a couple of days, injection was switched to seawater. With seawater cooling, the bridge plug eventually unseated and it was pulled. An inflatable had to be used because there was a 4 1/2-inch injection string with a 7-inch casing below (refer to the Layered Formations Workshop).

Are there specific situations when mechanical control may be the only currently effective means for guaranteeing adequate coverage of all zones? Various operators have tried to use orifice packers but they have sometimes been found to plug up.

24. Manipulation Of Breakdown And Propagation With Thermal Effects: There is good field evidence that cooldown can promote fracturing. Various methodologies are possible for thermal fracturing, particularly in consolidated formations. Consider some examples.

This is a Prudhoe Bay example. There was a 7-inch horizontal liner (~1,000 feet). A PLT had been run and the flow split was known (three zones?). As expected, cooling was much faster at the heel than at the toe. Early on in the injection, there was a large temperature differential between the heel and the toe. The concern was about thermal fracturing at the heel. Simulations were carried out, using a model with capabilities for temperature variation along the well and coupling with a fracturing model. It was established that even if there was only one layer and there is variation (e.g. upwards fining) and the well is not strictly horizontal, depending on the pressure and temperature, it may be possible to start the fracture at the toe, but the window may be small.

Elf has shown situations where radial flow mainly occurred in high permeability sandstone and fracturing was predominant in neighboring medium permeability sandstone. On the surface, one might have expected more flow in the high permeability layers. However, there was injection into all zones because of thermal fracturing and growth into the sands. THERMAL FRACTURING IMPROVED THE INJECTION PROFILE. Because of the contrast in modulus, thermal fracturing was induced in the hard rock. Some sands were soft enough that they didn't have substantial effects (no substantial thermal stress alteration). Some thermal effects can be used to your advantage.

Options for conformance are even more difficult when you go to soft rock because you may not have the thermal benefit.

25. Manipulation Of Breakdown And Propagation With Poroelastic Effects: Consider an actual example.

There was a high pressure difference across a shale separating two zones. The completion was a liner that was uncemented. It was desirable to avoid injection into a high permeability water zone. This was accomplished by setting a maximum bottomhole injection pressure. During injection, poroelastic effects caused the stresses to increase in the water zone and flow was diverted to the oil zone. "BP follows this practice more generally in converting producers to injectors by sidetracking "to ride the pressure wave."

26. Taking Advantage Of Different Formation Pressures: Consider the following (actual) example.

This is an extended reach/near horizontal well. It was in the same zone but the pressure varied from the toe to the heel. What are the possibilities for using different formation pressures for controlling fracture initiation preferentially? Fractures may initiate preferentially where permeability is lower. Stresses may be less in zones that have been depleted. It is important to have a feeling for your reservoir to be able to do this.

27. Recognize the Influence of S_o and k_ro: This is analogous to the Item above. Consider an example.

"The top layer in this field has a permeability of approximately 100 md (net-to-gross of 0.8) with an underlying interbedded zone with a net-to-gross of 0.4 and some local permeabilities up to a darcy. The upper zone has lower stresses. In the lower zone, s_hmin is 8000 psi, whereas it is 7000 psi (plus or minus) in the upper zone. Both zones have residual oil. The relative permeability to oil in the lower zone, k_ro, is 0.3. It is 0.5 in the upper zone. One possibility would be to use a level 5 or higher multilateral to allow you to physically isolate each zone. Do you perforate and complete or complete and then perforate?"

28. Dual Production/Injection Wells: This scenario may be possible, using a crossover packer with injection down the tubing and production up the annulus. Injection and production can support each other. Crossflow may be an issue. There is independent control.
29. Controlling Production: Consider a high/low permeability combination. Suppose that the low permeability zone has been fractured intentionally in the injector to assist in flow distribution at the injector. When water is produced from the higher permeability zone, choke it back or shut-in and continue to inject. You will need control valves in the producer - one for each zone being produced. Smaller flow control valves are being developed.
30. Annular Injection: Different zones can be accessed with the same or different fluids (produced water, seawater, gas...).
31. Zone Pressurization: In some lithology combinations, it may be possible to sweep a high permeability zone from the injector and then flood it (with control at the producer(s) to pressurize the low permeability zone. If k_v/k_h is not too low, flood the high permeability zone. Then, pressure up the high permeability zone and produce out of the low-pressure zone. Switch the manifold from production to injection. In Canada, Canadian Hunter has done this. Gulf steam-flooded and treated this as a horizontal fracture. Using crossflow to your advantage.
32. Manipulating the Injection Fluid: Gas injection in a lower permeability zone might inhibit crossflow. For gas injection, there would be some complaint about the cost of compression. Regardless, gas could be disposed in the low permeability zone and water injected in the high permeability zone.
33. WAG - One Well No Dual Completion: This is possible but it can be very difficult to operate.
34. Controlling Override: For a lower permeability zone, "What about gas riding upward and water downward?"
35. Cyclic Zone Production: Pull on the low permeability zone. Control the differential to avoid gassing out. Then pull on the high permeability zone. Cycle to avoid gas coming out of solution. This could be part of a strategy for overall reservoir management.
36. Plugging Material in the Injection Stream: Can you add material to the produced water to enhance plugging characteristics? There is chemistry to change the viscosity or the face plugging. Face plugging is not necessarily desirable because it reduces recovery.
37. SMART Wells: Are SMART injectors and/or producers a desirable consideration. Are there appropriate low-cost smart well technologies?
38. Deferred Production: This is one of the least desirable options to optimize injection conformance. It may be necessary to consider well control limitations on producers.