Integrated Water-Injection Analysis Uses Salinity as Tracer in Giant Brownfield
In this paper, an integrated work flow is proposed for brownfields where oil production is driven mainly by water injection. Produced-water salinity plays a key role, acting as a natural tracer and, thus, helping avoid additional costs for new data acquisition.
When dealing with giant brownfields, a systematic approach is mandatory to succeed in identifying the main production mechanisms amid the geological and dynamic complexity of the reservoir. In this paper, an integrated work flow is proposed for brownfields where oil production is driven mainly by water injection. Produced-water salinity plays a key role, acting as a natural tracer and, thus, helping avoid additional costs for new data acquisition.
Production-data analysis (PDA) is a well-known method of investigating field performance. Nevertheless, its application in a giant brownfield can be difficult. Analysis can be very complex and time-consuming because of the large quantity of data and data heterogeneity and reliability, especially if the data are gathered through different monitoring systems and technologies.
Considering this, a systematic integrated work flow has been developed that focuses on brownfields where water injection has played a key role in oil production.
The work flow allows a comprehensive analysis of the water injection through the following steps:
- Identification of injector/producer connections
- Fluid-path conceptual models
- Evaluation of the current water-injection efficiency
- Water-injection optimization
Phase 1, which includes the first two steps, is based mainly on advanced PDA. Integrated analysis of production, pressure, and salinity data allows for the detection of the evolution of injector/producer connections. The main fluid paths in the field then are identified and used to drive and integrate the geological knowledge of the field. The result is a more-robust 3D model, representative of the main highlighted production mechanisms.
Phase 2, which includes the third and fourth steps, instead is driven by streamline analysis traced on the dynamic model, which allows for the evaluation of the efficiency of the current water-injection strategy. Water injection then is optimized on the basis of a redistribution of the injection rates, promoting the most-efficient connections to the detriment of the least efficient. The consequent modification of the water paths allows for contact with new, unswept areas, leading to an increase in oil recovery with limited investments.
This paper shows the application of the work flow in the context of a wider brownfield-rejuvenation project. In particular, the paper is focused on the setup of easy-to-understand conceptual models and considers how they drive geological evaluations (e.g., faults, layering, contacts) in the realization of a new 3D model. A dynamic model able to reproduce the events since the preliminary history matching confirms the validity of the approach.
Conceptual Models: Work Flow
One of the main drivers of the PDA was salinity. The salinity of produced water can be treated as a reliable natural tracer because of the strong difference between the formation-water values (between 200,000 and 300,000 ppm) and the injected seawater (approximately 40,000 ppm).
All salinity values were plotted together vs. time (Fig. 1 above), showing the wide spread existing in the reservoir (gray dots). Different behaviors can be seen between wells producing only formation water throughout their production lives (blue dots) and those where the breakthrough of injected water has occurred (red dots). In fact, the former group maintains the salinity of the produced water always in the range of the formation water, while the latter group shows clear evidence of a decreasing trend toward lower values because of the breakthrough of injected seawater.
Salinity is integrated with static pressure and production data for all the single producers of the analyzed zone.
Given the field’s long production history, PDA is divided into time frames according to clear generalized behaviors or specific events characterizing the zone (e.g., depletion period, pressure support, injection startup, infilling).
In each stage, the analysis is furthermore integrated with repeat-formation-test (RFT) data to confirm the pressure trend and highlight the presence of eventual vertical heterogeneities inside the reservoir.
From this point on, the analysis proceeds on a well basis. This analysis not only highlights well-to-well connections but also allows information related to the aquifer presence, pressure-support efficiency, and breakthrough time to be inferred. Those results, once integrated in the geological context, lead to a more-complete picture of the fluid movement in the subsurface.
Conceptual Models: Applications
Conceptual models proved to be powerful tools to understand the field behavior and its changes over time as a consequence of depletion or other adopted development strategy. Furthermore, all this information can drive some critical assumptions during the creation of a detailed 3D model, such as fault-system characterization, fluid contact, and aquifer support.
Fault System. Fluid-path conceptual models can provide valuable information about the fault system. Whenever there is evidence of connections between injectors and producers though faults, the fault is allowing the fluid to pass through. Dramatically different behavior seen in other wells reveals that another nearby fault is sealing.
A very detailed geological analysis of the area was performed, coupling geological evidence with PDA outcomes. The results highlighted that communication between wells completed on dynamically separated levels was possible by sands superposition, through a complex system of faults.
Oil/Water-Contact Interpretation. A second important application of conceptual models is the validation of oil/water contacts, a critical issue because of a lack of robust information from logs and RFTs.
The connections between injectors and producers show good communication in the producing zone. Evidence suggests the presence of just two main oil/water contacts, despite the complex fault system and the historical interpretation, which claimed seven different fluid contacts.
A dedicated petrophysical analysis revised all the contacts and confirmed the conceptual model, which was incorporated in the new geological 3D model.
Vertical Separation Between Zones. Conceptual models improved the knowledge of vertical communication in the reservoir. For example, different dynamic behaviors were highlighted in two zones that had historically been considered in communication.
To reduce water production, the lower part of the interval was excluded in some wells on the basis of the assumption that the bottom aquifer was advancing. Salinity values showed a dramatic reduction as soon as production came only from the upper zone.
RFT data, although present only in a couple of producers, confirmed the analysis. The geological view of the area was then updated, and a vertical separation between the two zones was included in the new geological model.
Geological Zonation. Anomalies in well zonation are another outcome of the conceptual models. For example, Well 4 was directly entering a formation named “Lower” without crossing the shallower “Upper” formation, which was a strong disagreement with surrounding wells. The low salinity values showed that Well 4 was clearly producing injected water, like the other nearby wells of the area. Because the open injector is completed only in the upper zone, the analysis confirmed that Well 4 also should be open in the upper zone, and the zonation was updated accordingly in the new geological model.
Well-Integrity Issues. A simplified overview of the fluid paths in the reservoir helped eliminate production-data misinterpretation. On the basis of water-breakthrough evidence, a preliminary conceptual model was set up that showed aquifer arrival. Nevertheless, significant incongruities were identified. Although the initial period clearly indicated formation water, the range of values was anomalous, either lower or higher with respect to the aquifer salinity for the zone.
A more-detailed analysis of salinity values resulted in the identification of the following three well behaviors:
- Wells in line with the reference salinity of the zone (approximately 250,000 ppm)
- Wells with higher salinity (greater than 300,000 ppm)
- Wells with lower salinity (approximately 200,000 ppm)
Coupling this information with well-completion analysis revealed that many older wells showed well-integrity issues, either casing leakage when crossing a salt zone or packer leakage. On the basis of that information, only the wells with a salinity of approximately 250,000 ppm were considered representative of aquifer breakthrough, and the fluid-path conceptual model was reviewed.
Commingling Analysis. As anticipated, the work flow is based on the analysis of dedicated producers only, excluding commingled wells. However, once the fluid conceptual model for each zone is defined, it can be used to estimate the contribution of each reservoir through commingled wells.
The primary results of the analysis in the complete paper are easy-to-understand conceptual models where the evolution of fluid paths is clearly identified, providing valuable support for the geological vision of the reservoir. The result is a dynamic model able to reproduce the main fluid movements since the first history-matching runs, demonstrating the robustness of the approach and validating the integrated analysis performed.
The work flow was applied to a giant Egyptian oil field characterized by more than 60 years of production and approximately 500 drilled wells. It helped to improve the understanding of several static and dynamic critical issues: fluid contacts, well integrity, aquifer support, effect of faults, well zonation, and contribution of commingled wells.
This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 185872, “Integrated Water-Injection Analysis for a Giant Brownfield,” by A. Ortega, S. Renna, G. Fanello, C. Callegaro, SPE, and I. Bergamo, SPE, Eni, and O. Yehia, SPE, Petrobel, prepared for the 2017 EAGE Annual Conference and Exhibition/SPE Europec, Paris, 12–15 June. The paper has not been peer reviewed.