In preparation for the SPE Applied Technology Workshop on Use of 4D-Seismic and Production Data for History Matching and Optimization—Application to Norne (Norway), 14–16 June 2011, a test-case study (Norne E-segment), based on field data of an offshore Norway brownfield, was organized to evaluate and compare mathematical methods for history matching and strategies for optimal production or enhanced oil recovery (EOR). The integrated-data set provided an opportunity to discuss emerging and classical history-matching and optimization methods after being tested with real field data.
Introduction
The Center of Integrated Operation in the petroleum industry at the Norwegian University of Science and Technology (NTNU), in conjunction with SPE, organized an SPE Applied Technology Workshop on the use of real data from the Norne field. The workshop attracted 80 delegates and international speakers from more than 10 countries. This workshop addressed a comparative case study that used real field data that included time-lapse-seismic data.
The purpose of reservoir management is to control operations to maximize short- and long-term production. The process consists of life-cycle optimization based on reservoir-model uncertainties and model updating by production measurements, time-lapse-seismic data, and other available data. Time-lapse-seismic data help determine reservoir changes that occur with time and can be used as a new dimension in history matching because they contain information about fluid movement and pressure changes between and beyond wells.
The well-production schedule and history for the period from December 1997 through December 2004 were provided as observation data for the history match. A previous full-field calibration was performed by the operator to match the history up to 2003. The reservoir attributes calibrated in the previous history match included fault-transmissibility multipliers, regional relative permeability parameters, and large-scale (absolute) permeability and porosity heterogeneity using regional and constant multipliers. These attributes defined a global history match for a single (structural) reservoir description. The exercise was to improve the match and then perform a recovery optimization.
Five groups participated in this exercise, of which four presented their results during the workshop. The number of participants was limited because the Norne database has a license limitation regarding commercial companies. The participants in this comparative study were expected to produce a history-matched model, preferably using an integration of production and time-lapse-seismic data, and to produce an optimal production strategy for the remaining recoverable resources.
Norne Field
The Norne field is in Blocks 6608/10 and 6508/10 on a horst block in the southern part of the Nordland II area in the Norwegian Sea. The rocks within the Norne reservoir are Late Triassic to Middle Jurassic. The current geological model has five reservoir zones—Garn, Not, Ile, Tofte, and Tilje. Oil is found mainly in the Ile and Tofte formations, and gas is found in the Garn formation. The sandstones are at a depth of 2500 to 2700 m. The porosity ranges from 25 to 30%, and permeability varies from 20 to 2,500 md. The data consist of near-, middle-, and far-stack 3D-seismic data acquired in 2001, 2003, and 2004.
The first package included the E-segment of the Norne field, with subsequent benchmarks to include larger parts of the field. Further, seismic data were separated to suit the requirement of coverage of only the E-segment. The E-segment was chosen because it has the highest-quality seismic data of the entire field. The E-segment of the Norne field had 8,733 active grids and five wells as of the end of 2004 (i.e., two injectors and three producers).
Description of the Exercise
The exercise was defined 6 months before the workshop. This benchmark case covered 1997 to 2004 for history matching and 2005 to 2008 for recovery optimization. The 2004 simulation model containing all information and properties was provided. Production and injection data from 1997 to the end of 2004 and 4D-seismic data for the same period were provided. These data were the basis for the history match performed by participants. The defined workflow was as follows:
- Download the existing Norne model and import it into the participant’s reservoir simulator.
- Participants history match the model until the end of 2004 and predict the production (oil, water, and gas) performance until the end of 2008.
- From the obtained history-match results, participants create an optimal production strategy for the remaining recoverable resources for the stated period. Participants could also suggest techniques to enhance recovery because a significant amount of the recoverable reserves had been produced by the end of October 2008.
- The format for the production strategy should contain time, bottomhole pressure (BHP), or flow rates for the wells.
- Constraints
- For each injection well, the maximum flowing BHP (FBHP) =450 bar.
- For each producing well, the minimum FBHP=150 bar.
- For each injection well, the maximum water rate=12 000 std m3/d.
- For each producing well, the maximum liquid rate=6000 std m3/d.
- Maximum water cut=95%.
- A maximum of two wells can be sidetracked to increase recovery.
- Economic parameters
- Oil price=USD 75/bbl.
- Discount rate=10% at the reference time of January 2005.
- Cost of water handling and injection=USD 6/bbl.
- Cost of gas injection=USD 1.2/Mscf.
- Cost of a new sidetracked well =USD 65 million.
- Participants could assume their own parameters related to EOR methods (e.g., surfactants, polymers, and low-salinity waterflooding).
- Discuss and compare results of the achieved recovery factor.
General Methods
Four groups presented their results in the Applied Technology Workshop in June 2011 in Trondheim, Norway. A summary of the methods applied for history matching and recovery optimization by each group is shown in Table 1. To perform history matching, Stanford University started with dimensionality reduction of the reservoir parameters by use of principal-component analysis (PCA) and then applied particle-swarm optimization for history matching. For the subsequent optimization, they used a derivative-free method—the Hooke-Jeeves direct search (HJDS).
The group from Texas A&M University first engaged in multiscale reparameterization of the permeability field by use of the grid-connectivity-based transform (GCT) and then calibrated the reduced permeability to production data using a quasi-Newton method. Thereafter, they applied a streamline-based method to integrate the 4D-seismic data. Last, the Texas A&M group increased recovery and optimized the production forecast by draining the oil pockets through sidetracking, and then by applying a streamline-based method to equalize the arrival time of fluid-phase fronts at all producers.
The group from NTNU applied manual history-matching techniques that included qualitative use of time-lapse-seismic data. They then optimized production by oil-pocket drainage through the addition of new wells and low-salinity waterflooding. Method details for each group are explained in Appendix A of the complete paper.
Results
There was no winner or loser in this comparison study; no one knows the answer exactly. The outcome and experience were more useful. The group that attained the highest recovery factor also attained the lowest net-present-value (NPV) increase because their methods to increase recovery included the more-costly use of new wells. Although all of the groups used different strategies and methods, the results differ only slightly, which indicates that the applied approaches may be realistic. A summary of the results is shown in Table 2. One main challenge in reporting the combined case-study results was the different formats and notations used by the participants because there was no specific guidance on how to present the results defined before the work began. The participants’ specific notes about the methods and results are detailed in Appendix A of the complete paper.
Conclusions
The workshop provided an opportunity to address state-of-the-art technologies within the area of optimization, focusing on production history and 4D-seismic data and on the interplay between these diverse types of data. The workshop enjoyed active discussion and contributions throughout. This case study served as the first benchmark for the use of data from the E-segment of the Norne field.
A single conclusion for the Norne field characterization and optimal production schedule was not achieved. However, bringing the applied suite of approaches together was valuable. The use of seismic data in this case was not as expected; more use of seismic data, both qualitatively and quantitatively, will be required in future cases. Because the case study was defined only 6 months before the workshop, most participants considered the time provided as too short and, therefore, recommended that up to 1 year of study time be used for future comparative case studies.
This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper SPE 157112, “Results of the First Norne Field Case on History Matching and Recovery Optimization Using Production and 4D-Seismic Data,” by Richard Rwechungura, SPE, NTNU; Eric Bhark, SPE, Texas A&M University; Ola T. Miljeteig, SPE, NTNU; Amit Suman, SPE, and Drosos Kourounis, Stanford University; Bjarne Foss, SPE, NTNU; Lars Hoier, Statoil; and Jon Kleppe, SPE, NTNU, prepared for the 2012 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 8–10 October. The paper has not been peer reviewed.