ADVERTISEMENT

Integrated Asset Model Enables Simulation of Complex, Multifield Asset

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

This paper presents an innovative application of the integrated-asset-model (IAM) approach to simulate a surface network collecting many complex fields and multiple constraints. The use of a last-generation high-resolution reservoir simulator (HRRS) made it possible to perform the reservoirs-network coupling directly by a field-manager (FM) process, which acts as an orchestrator for a variety of reservoirs and network-simulation instances. No third-party controller application was required because the FM of the HRRS directly managed the communication between the different reservoir models and the surface-network model.

Introduction

The ability to predict reservoir behavior has always played a fundamental role in the definition of a development project, and numerical simulation can be considered the first tool for a quantitative evaluation of reservoir performance. In the past decades, improvements in information technology have increased calculation capabilities, enabling development of dynamic simulators able to solve for more-complex reservoirs and dynamic processes. During the production-forecast phase, for example, technical constraints must be considered, primarily related to hydrocarbon rate (i.e., gas sale limitation, gas/water injection limits, flaring policies, limitations in export pipelines, and treatment capacity) and pressures (i.e., minimum required pressure to reach a treatment, separator working conditions, and export-pipeline design).

The increasing development of fields located in remote or offshore areas and the growing interest in near-field exploration make it necessary to exploit common surface networks. In these situations, the previously mentioned technical constraints and limitations must be defined for the entire asset, thus accounting for the intrinsic connection between fields that become interdependent. With standard standalone reservoir simulation no longer able to identify and address these kinds of problems and constraints, integrating the reservoir and network is key to an optimized development strategy.

The IAM brings together reservoir, well, and surface-facility models in a unique system for reservoir and well optimization. This methodology ensures that the interactions between all the components are accounted for correctly and, by coupling dynamic reservoir and facilities models into a single IAM, addresses backpressure, the mixing of different fluids, and flow assurance.

Currently, two approaches are available to perform the integration. The first is based on combining the reservoir models with a dedicated network simulator with optimization capabilities in a fully integrated model. The second approach is focused on using the capabilities offered by some commercial simulators in which the network solver, on the basis of precomputed hydraulic tables, is included in the simulator together with some features of reservoir coupling.

Theoretical Background

An IAM combines different building blocks into a common simulation system. Reservoir, surface-network, and process models are dynamically coupled, removing the fictitious boundaries between surface and subsurface. This option is especially valuable when several integrated fields are produced and managed under the same constraints applied to the entire asset, such as a floating production, storage, and offloading unit, or a common surface network. The reliability in the consideration of the various network constraints within the IAM improves as the model moves closer to reality. Therefore, the IAM is built with the aim of providing a production forecast that incorporates the physics of both the reservoir and the infrastructure.

In this application, a fully integrated asset model based on an HRRS and an external steady-state network simulator were used. The supporting logic is that a section of HRRS dedicated to field management performs the ­reservoir-network coupling using the network solver to calculate and set additional constraints on the basis of network deliverability (rate or pressure). The inflow performances of the wells are provided by the reservoir simulator to the network solver as boundary conditions; then, the maximum allowed rate and minimum pressure are passed back to the HRRS. The reservoir simulator sets up the well-working parameters and the reservoir is simulated until the next coupling time. The FM acts as an orchestrator for the entire process. This approach does not require a third controller application to allow the reservoir-network coupling with the network solver. The HRRS manages the communication directly, acting as a controller program through a dedicated FM that provides a framework to build and simulate the needed field development scenarios. The HRRS consists of two dedicated sections, with one acting as an FM controlling well schedules and field or well constraints, while the other, the proper reservoir simulation engine, models all interactions between cells (i.e., geometrical characteristics, well completions, pressure/volume/temperature properties, and petrophysical properties). Splitting reservoir simulation into separate processes enables effective management of the coupling approach, allowing flexibility in selecting the suitable network model.

A dedicated feature was developed to allow the communication between the last-generation HRRS and the external steady-state network solver. The approach is suitable for any commercial network simulator.

Although this method is more-­complex and computationally heavier than a simplified solution based on prebuilt tables for network resolution, which is available in some commercial reservoir simulators, it provides a more-accurate and -reliable calculation of pressure losses in the pipelines. While in the prebuilt tables the pressure losses are precalculated in defined conditions and then parameterized as a function of the variables, the network simulator is able to calculate the instantaneous pressure drop in the pipelines by considering the fluid and the surrounding environment temperatures in each network node, the mixed fluid density according to the different rates, and the real complexity of the network, including internal loops, ending points, and pipeline routing.

IAM Implementation

The proposed flexible IAM approach was preliminarily tested on a single reservoir model to optimize computational efficiency with respect to the necessary process details in terms of memory usage and simulation run time. The methodology was then implemented on the full asset: three low-permeability reservoirs with horizontal multifractured wells interconnected to a complex surface network (Fig. 1), constrained by limited gas market demand and zero flaring policy.

Fig. 1—The IAM approach was used to simulate an entire asset consisting of many complex fields and multiple constraints successfully.

 

The test case consists of two producing oil and gas fields. Field A comprises two reservoirs: an oil reservoir saturated with gas cap, and a gas-­saturated reservoir. Field B is a gas-saturated reservoir. The two main fields are simulated by means of three different reservoir models, having different production levels separated hydraulically. Production is delivered to shore by means of a complex surface network in which many platforms are interconnected. Thus, the minimum delivering pressure from a single platform is strongly dependent on the backpressure related to the other production platforms. Because further development is planned, taking into account reliable pressure constraints and system optimization in simulations is mandatory. An IAM was built by coupling the three reservoirs to a network simulator acting as the main driver for dynamic interdependent pressure constraints.

Conclusions

  • The IAM approach was used to successfully simulate an entire asset, enabling the company to maximize oil recovery in its future scenarios by accounting for overall system constraints, such as gas demand and total gas limits.
  • The IAM approach provided a flexible method to analyze different development options and well- and pipeline-routing configurations to maximize oil production, improving asset gas management.
  • The three dynamic models were successfully coupled, addressing overall asset and facilities constraints. The comparison between the resulting production profiles with the standalone model simulations, constrained by fixed minimum tubinghead pressure (THP), clearly showed the effectiveness of the proposed IAM approach. With THP calculated with IAM according to the actual flow conditions, the proposed methodology resulted in a strong improvement, especially during the tail-end production phase, that affects ultimate recovery and reserves estimation. With the proposed approach, asset performance could be properly evaluated by correctly taking into account the backpressure of the multiple interdependent platforms. Moreover, the application of HRRS made it possible to run the reservoir simulations in an efficient way on a high-performance computing cluster to speed up the overall process.
This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 192603, “New-Generation Integrated Asset Modeling: High-Resolution Reservoir Multimodels Coupled With an External Steady-State Network Solver,” by Pietro Selvaggio, Fabrizio Freni, Roberto Rossi, Danielle Christian Di Giorgio, and Ivan Colombo, Eni, prepared for the 2018 Abu Dhabi International Petroleum Exhibition & Conference, Abu Dhabi, 12–15 November. The paper has not been peer reviewed.

Integrated Asset Model Enables Simulation of Complex, Multifield Asset

01 July 2019

Volume: 71 | Issue: 7

ADVERTISEMENT