The overhaul of the reservoir model of the giant Forties oilfield – from data management to full field simulation – is described here by operator Apache North Sea’s Jeffery Pyle and Schlumberger’s Gordy Shanor.
Production from the Forties field began in 1975, peaking four years later at approximately 550,000b/d. By 2003, when Apache had taken over as operator, production had declined to about 42,000b/d. By 2008, field rejuvenation activities had increased production to an annual average of 61,700b/d. To date, Apache has invested in excess of $2.8 billion on facility upgrades, maintenance and infill drilling.
In 2008/09, Forties was the second largest producing field in the UKCS and in July 2009 monthly production averaged 73,500b/d, the highest monthly average since May 1999.
To achieve further improvements, Apache assembled a multidisciplinary team to prepare a detailed field development plan (FDP) – completed in May 2009 – that documented a five-year work program engineered to maximize field net present value over the remaining life of the field. In addition to addressing surface facility constraints and upgrades, the plan is linked to a new ‘live’ subsurface model and reservoir simulation workflows.
The Forties field, located in the central North Sea about 110 miles east of Aberdeen, is the UK’s largest oilfield. Its Paleocene sandstones are estimated to have contained approximately 5 billion barrels of oil originally in place; 2.6 billion barrels of oil were produced between 1975 and 2008. The field has more than 330 wells, each connected to one of five fixed platforms: Forties Alpha, Bravo, Charlie, Delta and Echo. The wells initially flowed to surface by natural lift but were later enhanced using gas lift or downhole electric submersible pumps (ESPs).
Initial rejuvenation program
Field production declined from a peak in 1979 of 550,000b/d to about 42,000b/d in 2003 when Apache became operator. After the acquisition, Apache immediately initiated an intensive rejuvenation program that included infill drilling, workovers and facilities upgrades. By 2008, 50 new targets had been drilled, and several projects completed to upgrade power generation, export pumps and water injection systems. The combined impact of these activities increased production in 2008 to a peak of 81,000b/d and an annual average of 61,700b/d.
Several challenges, common to most mature fields, were increasingly impeding further improvements in production. The infrastructure was ageing, infill targets were getting smaller and more difficult to predict, and drilling results were variable. Reservoir pressure was falling and water production was increasing. In addition, the capacity of surface facilities was an increasing constraint.
In 2008, Apache decided to take a fresh look at its Forties asset, spanning from the geology of the reservoir through to the surface facilities. The company assembled a multi-disciplinary team to prepare a detailed FDP to better assess the technical and economic issues facing the field and identify the options and priorities for short- and long-term projects that would deliver the greatest value in the future. The team was led by Apache and included personnel from Schlumberger with technical expertise in geosciences, reservoir modeling, drilling and completions, as well as experts from Petrofac to support technical issues related to facilities and production operations.
The FDP was built in three phases over an 18 month period. Phase one built a ‘base case’ FDP. Apache’s experience from the previous five years was reviewed, looking back to compare the actual versus predicted drilling target identification, drilling performance, well outcome, reserves estimates, completion design for sand control, effectiveness of the water injection, and suitability of the reservoir models. Phase one also assessed the infrastructure, defined remedial projects, and derived production forecasts assuming drilling of the existing inventory of infill targets and limitations of the current facilities.
Phase two identified and defined development options to increase production and assess impact on production reserves. The feasibility of incremental and enhanced recovery options was also tested.
Phase three combined the results of phases one and two to identify the highest value self-consistent set of options, and collated these into a five-year integrated activity plan – the FDP. Reserves and economics were estimated for the development options and compared with the ‘base case’ to determine the optimal way forward. This phase also included a rebuild of the reservoir model.
Immediate benefits
Phase one quickly led to improvements in target quantification and well planning, which resulted in better well results, drilling performance and increased production in 2008/09. The forensic lookback also discovered and highlighted a number of issues related to the existing Forties subsurface model that would need to be addressed before a robust FDP could be built. These issues included the modeling approach, the inter-relationships between the existing static and simulation models, and shortfalls in data reliability and interpretation methodologies in the vast historic data sets.
It became apparent that an updated reservoir simulation model was required to test and optimize water injection strategies and incorporate recent drilling results into the subsurface model more efficiently. Apache decided to build an integrated model using Petrel* seismic-tosimulation software, and to use ECLIPSE* reservoir simulation software to support decisions for future field development and optimization. This decision required the integrated subsurface team to perform a set of supporting technical studies.
These studies were intended to address data consistency and management methods, interpretation and modeling techniques, anisotropy and geomechanics, borehole stability, completions methodology and the data mining and integration functionality required for production and engineering work. This included pilot water flood, improved oil recovery (IOR), and enhanced oil recovery (EOR) feasibility studies. To determine how the full-field model should best be built, the process was first tested on a subset of the field around the Forties Charlie platform, focusing primarily on detailed seismic interpretation, depositional facies analysis and modeling in the Petrel software.
Verifying data quality
Initial reviews of existing databases highlighted issues in historical data accuracy, availability and accessibility across all disciplines. With more than 330 wells in the static model, the reliability of these data was vital to building an accurate model ready for reservoir simulation. In addition, several vintages of 3D seismic were used for subsurface imaging and 4D time lapse monitoring of the reservoir. Apache inherited seismic datasets acquired 1988-2000, and acquired new data in 2005. Finally, the existing petrophysical and seismic interpretations had been prepared using multiple software platforms, leading to various incompatibilities.
The InnerLogix* data quality management solution was implemented to analyze and quality check the existing Forties project, seismic and log data. Some legacy log data required digitizing and splicing, and some was found to be incomplete. An InnerLogix database was created with verified information during – and since – the model building process that has enabled better data handling across disciplines. This not only applied to historical information, but also updates with real-time and processed data from new wells and new interpretations. Data from InnerLogix was – and continues to be – used to generate the Forties Master Petrel Reference Project, which provides a single, common source for approved interpretations and model updates for all asset team members.
Building a new model
A full-field static model was generated from field-wide reinterpretations of structural surfaces, well data, petrophysical logs, formation tests and production data. Faults were also mapped and included in the Forties model for the first time, providing a better understanding of potential flow barriers. The implementation of repeatable workflows in the model building process allowed ‘start to finish’ model rebuilds with new drilling data or revised interpretations to be undertaken in minutes. Repeatable workflows were also used during the reservoir simulation phase to speed history matches, allowing rapid permeability adjustments. 4D seismic data was used directly in the dynamic simulation to honor sweep patterns while additional seismic data was used to better characterize variations in vertical transmissibility around the field.
Detailed data mining, production and salinity mapping, pressure analysis, anisotropy and borehole stability, IOR and EOR feasibility, and reservoir event studies progressed throughout phases two and three, providing input to the FDP. All information was integrated into the evolving static and flow simulation models. The final FDP report was completed in May 2009, incorporating all of the results of the supporting studies performed during the three phases. It was updated later in 2009 with direct modeling inputs from an updatable, integrated ECLIPSE simulation model built from new interpretations and consistent data, providing a ‘live’ document based on a ‘live’ model.
Future static model updates are now possible within days rather than months, and the database of information feeding into current drilling target generation work and field exploration is reliable, accessible and up-to-date. The revised static model also allowed a more informed opinion of in place volumes and therefore remaining reserves potential.
Apache is now drilling about one well per month in Forties. A new ‘4D snapshot’ seismic survey was acquired in 2010, and will provide further information about the reservoir that will be used to update the model.
Lessons learned
A key lesson learned from this project is the importance of maintaining coherent processes that continuously interrogate and integrate the vast amounts of knowledge and data involved in fields the size and age of Forties. Through effective data management, Apache is now able to easily load new information into the Petrel Reference Project and update the static and dynamic simulation models.
Data is the foundation of decisions – but only good quality data can form the foundation of good decisions. The model-centric FDP document and the living Forties model provide a solid foundation supporting future drilling and development decisions with higher confidence.
The ability to quickly and easily rerun models when new information becomes available enables the living model and FDP document.
A well integrated multidisciplinary team was also critical to the success of the project. The inclusion of personnel from service companies not only provided specialized expertise, but also enabled Apache to build the FDP while managing ongoing drilling and production projects. OE
About the Authors
Jeffery Pyle is a development geologist at Apache North Sea, identifying infill drilling locations through the integration of seismic, geologic and production data. He joined the oil & gas industry five years ago, having received a BSc in Geology from Reading University and a Masters degree in Petroleum Geology from Imperial College London.
Gordy Shanor is a senior technical consulting geoscientist in the Centre of Competence with Schlumberger Information Solutions, currently based in Stavanger, Norway. He has over 35 years of experience worldwide, and coordinated the integrated modeling team during the Forties FDP project.