ADVERTISEMENT

Multireservoir, Multinetwork Production Optimization of an Omani Reservoir

Topics: H2S/sour gas

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

Integrated reservoir and production modeling can be beneficial in conducting production forecasting for complex systems consisting of multiple reservoirs, fluid mixing, and complex production networks. In the complete paper, a new, fully coupled implicit tool was used to model an onshore Omani asset with multiple reservoirs, each featuring different fluids and multiple networks. A multifidelity approach was used throughout the modeling work flow, incorporating reservoir- and production-related uncertainty in the forecast and optimization processes.

Introduction

Petroleum Development Oman operates, in the south of the country, the Southern Sour cluster of high-pressure, highly sour oil fields. These fields are currently produced by what is known as the H Station. A second such facility, R Station, is under construction and will be onstream by 2019. These two facilities will be interlinked and will allow large-scale miscible gasfloods in more than 10 fields. The integrated nature of the production and injection systems is extremely complex and will be unique in terms of its scale and the technology deployed.

The project involves tying together multiple reservoirs and separating the reservoir fluids in the surface network, including the handling and separation of sour gas [the asset has a high hydrogen sulfide (H2S) content in reservoir fluids]. The separated acid and sour gases will be compressed at a high pressure and reinjected into the various reservoirs to create miscibility in the oil, thereby increasing its mobility and preventing a rapid decline in reservoir pressures.

Because the development is a critical one that requires correct modeling of fluid flow throughout the reservoirs, wells, and facility, the process of integrated production systems modeling (IPSM) becomes important. In the past, for such a project, implementation has been achieved by coupling several specialized tools for different aspects of the IPSM process in an explicit manner. The complete paper discusses the use of a new fully coupled, implicit solver to overcome the challenges faced during conventional IPSM work.

Work Flow

The work flow for developing the integrated model for the asset can be divided into the following broad but interdependent areas:

  • Reservoir simulation
  • Fluid modeling
  • Well modeling
  • Facilities modeling
  • IPSM solution

Reservoir Simulation. In the asset, nine high-fidelity discretized reservoirs and seven low-fidelity decline-curve reservoirs were modeled with the same tool. Most of the discretized reservoirs were designed to undergo gas injection, either to maintain pressure or for enhanced oil recovery (EOR). Some discretized reservoirs were designed only for primary production. The purpose of decline-curve reservoirs was to replicate primary production from certain fields where the geology and well configuration was simple enough to allow for the use of a lower fidelity.

Solving the Challenges in Reservoir Modeling. The problem is quite large because of the number of discretized reservoirs. Each reservoir was modeled with a compositional fluid model, with each fluid model containing 10 components. Apart from the complexity of compositional modeling, another complexity exists because of the common surface network shared by all the reservoirs, which creates interdependence among the reservoirs in terms of backpressure as well as changes in the compositions of injected fluids. In a conventional IPSM solution, modeling this interdependency reliably and ensuring material balance across the IPSM system can be extremely cumbersome. In the new approach, this interdependence is handled reliably and efficiently because of the fully coupled, implicit solution.

Multiple Fidelities in Reservoir Models. In this project, three kinds of fidelities were used in the reservoir modeling, defined in the following terms:

  • High fidelity implies the use of the original geological model without any simplifications.
  • Medium fidelity implies the use of an upscaled version of the geological model in which the upscaling is performed selectively so that the accuracy of the results is maintained.
  • Low fidelity implies the use of decline curves to model primary production from certain reservoirs in which geological complexity and interwell effects are minimal.

Fluid Modeling. Because each discretized reservoir is modeled with a unique equation-of-state (EOS) fluid, nine EOS fluids are in the asset. For simplification purposes, the decline-curve reservoirs are modeled with one of these fluids. Each of the nine fluids has 10 components. This gives a total of 90 components in the asset; all of these can exist in the shared production network at any given time.

However, several components are similar in properties and can be mapped to one another, resulting in fewer components in the network. In the end, each discretized reservoir was modeled with a unique EOS fluid model with 10 components and the surface network was modeled with 42 components. The properties of the integrated fluids depend on the amount of incoming fluids and changes in real time.

Maintaining consistent fluid definitions and mass balances throughout the IPSM system—from the reservoir to the production wells and facilities and back into the reservoirs by injector wells—can be very hard to achieve, especially if different tools are being used for each of the IPSM components. The implicit approach taken in this project, which uses a single tool, helped overcome this challenge in an efficient manner.

Well Modeling. A single tool was used for all the modeling work in this project, which means all the well- and facility-related modeling work was also performed on the same tool, thus mitigating the problems encountered when asset teams typically use different tools and simulators for solving different parts of the IPSM problem.

In the asset, more than 100 wells are present. Initially, a few wells were operational. During the course of field life, more wells become operational at various stages.

Multiple Fidelities in Well Models. In this project, a low-fidelity model was used for the wells. A low-fidelity well model only calculates the static pressure drop in the well. This was acceptable in this project because the wells are almost vertical in this asset. For other, more-complex well configurations, the option exists to use a medium-fidelity well model (for static and frictional pressure-drop calculations) and a high-fidelity well model (for static, frictional, and acceleration pressure-drop calculations). Lift tables also may be used to model the wells.

In terms of heat-loss calculations in the well, again, several options are available. However, in this project, wells were considered isothermal. Both the producer and the injector wells had both rate and pressure constraints defined. In addition, facility-based constraints were imposed on the producer wells so that their production was apportioned to meet the maximum capacity of the surface facility.

Facilities Modeling. The entire integrated facility model was created in the same tool so that there was no discontinuity between the fluid properties in reservoirs and wells and in the facility. Each timestep solution considered all the reservoir and fluid properties, well constraints, productivity and injectivity limits, and facility limitations and constraints.

The main characteristics of the facility model were as follows:

  • There is an existing facility (H Station) that produces through a separator train.
  • In addition, a planned facility (R Station) will have its own separator train. Some gas from H Station will enter R Station through crossover.
  • Both sections of the facility feature three-stage separation. In addition, the gas is stripped of the sour components (carbon dioxide and H2S) and the sweet gas is sent to sales. All other gas is reinjected into various reservoirs, either for pressure maintenance or for miscible EOR.
  • Compressors are used for gas reinjection.

The sweetening unit separates the sour components and directs the sweet gas toward sales while operating under maximum handling capacity.

IPSM Simulations. With these inputs, IPSM simulations were performed for a long duration, typically 30 to 60 years. During the first 12-to-15-year period, most of the reservoirs are in history mode and thereafter in forecast mode. Certain reservoirs undergo only primary production, whereas the others undergo miscible gas injection. Simulation time events were used to trigger the gas injection at desired constraints and dates.

At each timestep, all the operating wells bring the various reservoir fluids into the integrated facility, their volumes depending on their productivity and constraints. In the facility, fluid blending, three-phase separation, and gas sweetening take place and then sour gas is reinjected into the injector wells. A variety of numerical techniques are used for smooth running of the simulations.

Conclusions

  • The performance of various reservoirs under gas reinjection was modeled. This included depletion profile, dilution of oil by miscible gas, and degree of pressure maintenance.
  • Fluid continuity was maintained, which ensured the correct compositions within the facility as well as in the injected streams.
  • The production profiles for various scenarios were generated while prioritizing sales-gas production, maximizing oil production to facility capacity.
  • The work was performed in a continuous and collaborative manner by multiple users from multiple disciplines working from the same database with appropriate fidelities for various sections of the IPSM process.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 188377, “Coupled Multireservoir, Multinetwork Production Optimization of an Onshore Middle Eastern Reservoir With Sour-Gas Reinjection for Miscible EOR,” by Varun Pathak, SPE, Chelsea Palaschak, Anjani Kumar, SPE, and Rob Eastick, Computer Modelling Group, and Ralf Schulz, SPE, and Abdullah Al-Hadhrami, Petroleum Development Oman, prepared for the 2017 Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, 13–16 November. The paper has not been peer reviewed.

Multireservoir, Multinetwork Production Optimization of an Omani Reservoir

01 June 2018

Volume: 70 | Issue: 6

STAY CONNECTED

Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.

 

ADVERTISEMENT

ADVERTISEMENT

ADVERTISEMENT

ADVERTISEMENT

ADVERTISEMENT