Shaya adopts Delfi for faster run times, enhanced collaboration, and better decision-making

As part of its ongoing operations, Shaya Ecuador S.A. successfully used two applications running on the Delfi™ digital platform: Petrel™ E&P software platform and On Demand Reservoir Simulation. This enabled Shaya to make more informed operational decisions by drastically reducing modeling and simulation times, increasing the resolution of the geological data captured and improving its understanding of reservoir behavior.

The challenge

The area under study is a mature oil field in Ecuador with considerable complexity. The field geology is an estuarine environment with large degrees of vertical and lateral heterogeneity. Historically, 17 individual static models were built to represent the geology and to account for limitations in simulation run times. Due to the different static and dynamic models across multiple projects, and with limited computing resources, it was not possible to account for commingled opportunities. Reservoir response to the waterflood could not be well established, and it was difficult to meet project deadlines and analyze all the available data appropriately.

The solution

Deploying Petrel platform and On Demand Reservoir Simulation on Delfi enabled Shaya to leverage the scalability of the cloud and open new analytical horizons.

The Delfi digital platform brings together a collection of digital solutions for petrotechnical workflows, accessed from profiles covering the entire exploration to production (E&P) life cycle, hosted in the cloud, and available on demand. Market-leading solutions such as the Petrel platform and the INTERSECT™ high-resolution reservoir simulator run with significantly faster compute times and provide access to new tools that increase flexibility and productivity.

The results

By adopting the Petrel platform and On Demand Reservoir Simulation on Delfi, the 17 individual static models could be integrated into three regional sector models covering the north, middle, and south of the field. The lateral and vertical geological complexity could be accounted for with an increased number of cells and stratigraphic layers, giving a more accurate representation of the field.

The Shaya team was able to create a high-resolution model with detailed physics that more accurately represented the reservoir’s geological heterogeneity, while doing a better job of capturing reservoir response to the waterflood. The new approach also enabled them to assess commingled opportunities. Cross-domain collaboration within the team was greatly improved as all team members were able to work in the same environment with the same set of data.

Sensitivity and uncertainty tests

By leveraging the high-performance computing power of the Delfi digital platform, a larger number of sensitivity and uncertainty scenarios were tested with drastically reduced run times:

  • Run time for uncertainty analysis and volumetric calculation was reduced from 2 hours to just 6 minutes for a reservoir model and from 1 week to 6 hours for a regional model.

  • Updating of each regional static model was reduced from 1 day to 2 hours.

  • Regional scale models, which could not be simulated using on-premise resources, were easily simulated in the Delfi digital platform.

  • Resolution of the regional model was improved over the previous sector models—from 12.2 million cells (100 m × 100 m × 1 m) to 54 million cells (50 m × 50 m × 0.5 m).

Geological modeling

Previously, for the central area of the field, there were five static models and six reservoirs with 269 wells (253 producing and 16 water injectors). The five static models were integrated into a single regional model with seven reservoirs and the same number of wells. Before-and-after comparisons are shown below for 3D grid dimensions and grid cell resolution.

3D grid dimensions, km

Table showing 3D grid dimensions in kilometers

Model resolution

Table showing this models resolution of the number of cells per reservoir (in millions)

Reservoir simulation

Computerized image or a reservoir simulation before and after
Before and after views of the dynamic model

Before:
For each region, coarse grid reservoir models (100 m × 100 m with 1 m cell thickness) were simulated separately and combined to produce a forecast for field development.

After:
On Demand Reservoir Simulation on Delfi was used to run a unique fine-grid reservoir model (50 m × 50 m with 0.5 m cell thickness) that could capture the detailed physics and high resolution required for waterflood optimization. This new tool enabled users to capture the interactions between reservoirs for optimum analysis and field development planning.

Sensitivity analysis

The chart below shows the time it took to finish a specific sensitivity analysis (evaluation of well performance) with about 100 runs. As shown, the use of the Petrel platform and On Demand Reservoir Simulation on Delfi significantly decreased the time required to perform the analysis by 50% to 75%—allowing more time to evaluate additional scenarios.

Sensitivity analysis graph
Performance comparison between on-prem and Delfi of an uncertainty analysis

Find out how the Petrel platform and On Demand Reservoir Simulation on Delfi can help your organization

Contact Us