Conduct a search of the oil and gas technology environment today and you’ll find “production optimization” to be one of the most commonly used phrases.

From the latest downhole gauges to the best in artificial lift and well control technologies, all

Figure 1. In a three-dimensional view of the horizon model, a continuous surface method of horizon modeling may have difficulty with the triangular areas between the lambda faults and the main fault. The fault block-based method of horizon modeling does not encounter this problem. (All graphics courtesy of Roxar)
operators are looking to increase their performance, production and profits, and this against a backdrop of increasing maturity and complexity of reservoirs, increased scrutiny of financial reporting, and a growing shortage of staff.

The volumes of work and the stakes have never been higher; fewer people are available to take on these challenges.

The structural model
Yet, while production optimization tends to focus on fields when it has gone live, many of the key decisions tend to start at the field development planning phase.

One of the most overlooked areas of reservoir management which has the potential to have a strong impact on production optimization is the structural model.

For many exploration and production (E&P) professionals, the structural model is simply a platform on which the rest of the field development planning process is conducted. With dynamic processes such as seismic interpretation, simulation and history-matching taking place around the structural model, too often the structural model is considered a time-consuming and labor-intensive process, which simply brings the different disciplines and geological properties together.

To take this attitude, however, is to disregard one of the most significant productivity enhancement opportunities available in reservoir management today.

The structure of a field is one of the most critical components of the reservoir model and can account for the greatest uncertainty in terms of in-place reserves.

Get the structural model wrong and reserves calculations, production predictions and field development planning will be impacted. Inaccuracies due to complex faults, for example, will directly impact the net present value (NPV) or result in incomplete geological models going forward for simulation.

On the other hand, get the structural model right and you are looking at a much greater return on investment from data acquisition and interpretation expenditure, increased productivity from your workforce, and increased production optimization from your reservoirs.
Take the North Sea — a classic example of the repercussions of getting your reservoir model wrong. Here, reserves have increased or decreased by more than 50% in more than 40% of the fields. This has led to the requirement to drill 60% to 80% more wells than originally anticipated in the field development planning process, directly affecting the financial return of each project.

A question of time
Time is also crucial when generating structural models. Time over-runs can lead to delays in production, the missing of bid dates or incomplete models going to simulation. The geological modeling phase — which many E&P professionals regard as the key step for new interpretation value and knowledge to be added — is stripped to the bare minimum.

During the production phase, there can also be long delays in structural model updates as new production data is incorporated into the model. Field development events such as drilling, coring and seismic acquisition are all taking place and providing new data on the reservoir. This data is, however, not being effectively used to improve or refine the reservoir model and in turn improve field development planning decisions such as well placement or production optimization strategies.

Up to the task?
Given the importance of the structural model, are today’s structural modeling tools meeting the challenge? The answer would have to be no.

Today’s structural modeling process can be long, cumbersome and labor-intensive — the
Figure 2. In a three-dimensional view of the horizon model created from the fault model a continuous surface-based horizon model would not be able to represent the repeat section under each of the three reverse faults. The fault block-based method of horizon modeling does not have this limitation.
domain of a few specialist reservoir geologists or geophysicists. The length of the process is also highly dependent on the type of field and the number and complexity of the faults. The tendency has been that the higher and more complex the number of faults, the longer the process takes.

And even after this long process, the faults may not be properly represented. Current methodologies have limitations on the types of fault intersections and the number of faults that realistically can be modeled. The structural framework is often a compromise between the actual structure and what the modeling system allows, particularly in areas with large numbers of Y-intersections, low angle faults or reverse faults.

Sophisticated geostatistical techniques may be commonly used to create facies and petrophysical models, but the underlying structural frameworks often do not correctly portray the true structure.

Building the framework
To meet the challenges of structural modeling, Roxar has developed a new technique for structural framework building.

The approach to model building is not just a new method of creating a fault framework but also applies to the creation of the geologic model and the reservoir grid. The process involves three steps — creating the fault framework, creating the geologic model and creating a reservoir grid.

The method does not require compromises or simplification in types of faults, numbers of faults or types of intersections; does not require compromises or simplifications in horizon modeling of complex shapes such as repeat sections; and does not require compromises or simplifications in creating the reservoir grid, such as limiting the types of intersections that can be made.

Whereas other methods (pillar methods and binary tree methods, for example) have difficulty in modeling low-angle faults, nested Y-faults and self-truncating faults, the new methodology has none of these limitations. Sections of faults are bounded by other faults which cross or touch the fault, and these sections may be manipulated independently of one another.

The simplicity of building and editing the fault relationships, creating the stratigraphic model, and building the reservoir grid means that al members of the asset team can easily update a model, test different interpretations and use the model for both geologic and engineering applications.

Fault surfaces, for example, can be automatically adjusted to well picks, may die out laterally or in depth, and can be truncated by unconformity surfaces. Horizon surfaces can also be generated directly from interpreted seismic data or calculated using well and thickness data.
Horizontal wells are also addressed, with such wells often containing a limited number of well picks, resulting in horizon surfaces erroneously crossing the horizontal section of the well. The service company’s structural modeling can use the zone information from a zone log to control the horizon surfaces and ensure that the structural framework honors all the information from horizontal wells without the addition of pseudo data.

From months to weeks
The speed of this technique, along with a streamlined workflow, an intuitive graphical user interface, optimized defaults and minimal manual editing, can reduce cycle time from months to weeks. This frees up productivity to build more scenarios and reduce uncertainty.

And the fact that the entire asset team can now access the same structural representation provides consistency across the disciplines and increased productivity. The interactivity of the model-building process allows an asset team to test a variety of interpretations where appropriate or to include the details necessary to make informed reservoir management and production optimization decisions.

The structural model’s accuracy is also improved. It is now possible not only to create fault frameworks of complex truncations but also to take these frameworks through to reservoir gridding without having to simplify or alter the fault relationships. The geologic structure does not have to be compromised or simplified in order to fit the constraints of the modeling systems.

New 3-D grid building, designed to work with the new structural framework building, will also ensure the building of the best quality grids. The result is simulation-friendly grids, suitable for accurate predictions of production where as much structural complexity as possible is incorporated into the grid.

A living model
Just as an accurate and fast structural model is essential to optimizing production at the field development planning process, real-time updating of the model can be crucial during the actual production phase.

Newly acquired data is processed, interpreted and updated into the model on a frequency that enables new reservoir development events to be positively affected.

Uncertainties in structural modeling and geological property modeling can all be simultaneously evaluated, ensuring that the full impact of these uncertainties is captured in the structural model.

The result is a structural model that integrates all available data, including seismic, well log and other geological data, and attempts to quantify all structural and reservoir property uncertainties. Ease of use and intuitive structural modeling will go a long way toward making this a reality.

The user’s choice
Although further back in the field development planning phase, higher quality, faster and updatable structural models are crucial components of production optimization today.
With recent developments, the level of structural complexity modeled is now a user choice rather than one imposed by the technology. The result will be quicker and more accurate characterizations of the reservoir, better decision-making, and maximum reservoir performance.