|
|
Case Study: Talisman Energy Improves Production Forecasting and Secures Future Production
"The availability of the keyword editor makes Petrel software a powerful tool for integrated studies"
Clas Normann
Talisman Energy
|
|
An integrated field study of Gyda South using Petrel and ECLIPSE software
Challenge
- Recalculate field reserves
- Better understand well behavior to stem rapid water breakthrough
- Improve production forecasting capability and fully tap field reserves
- Provide a means of team collaboration
Solution
Integrate the workflow across a multidisciplinary team using a combination of Petrel and ECLIPSE software for geophysical interpretation, geological modeling, and reservoir simulation
Results
- Produced a reliable simulation model that was used to optimize injection rates and design a new producer well
- Increased understanding of the reservoir and confidence in the production forecasts based on the simulation model, as it remained consistent with the geological interpretation
Gyda South field
Gyda South, a Talisman Energy asset, is an onshore field located about 40 km southwest of Stavanger, Norway. The field is a four-way, dip-closed structure with a small amount of normal faulting. The reservoir rock is sandstone, deposited in a prograding shoreline depositional environment. Average reservoir permeability is approximately 100 mD. The field contains a light oil with a bubblepoint of approximately 350 bar and a gas/oil ratio of approximately 600 m3/m3. There was no primary gas cap in the field.
A single production platform with a total of 32 slots produces oil from the Gyda, Gyda Southwest, and Gyda South fields. Gyda South production is from one well, Well A-13, supported by two injection wells, the A-21 and the A-29. Initial production was approximately 12,000 stb/d. Productionexperience showed that high connectivity exists between Well A-21, the southerly injector, and the producing well. As a result, rapid water breakthrough occurred, prompting Talisman to conduct an integrated field study by a multidisciplinary team consisting of a reservoir engineer, a geologist, and a geophysicist.
Conduct integrated field study
A Talisman project team, including Clas Normann, Vegard Bruaset, and Lyndsey Dyer, undertook the integrated field study to recalculate field reserves, increase understanding of well behavior, and predict future production. From an existing geophysical interpretation, this team developed a new static model and a new dynamic model of the field. The dynamic models were built from the geological model without upscaling to better reflect geologic heterogeneity in the simulation model. A compositional reservoir simulation model was used with a 6-hydrocarbon component fluid description, with 17 layers representing the “A” and
“B” sands of the Gyda reservoir.
The simulation model was history-matched with current production data. Early versions of the model predicted bottomhole pressures in the producing well significantly higher than those observed in the field. The project team improved the match between the predicted and observed bottomhole pressures by reducing the pore volume in the model. The geologist and reservoir engineer worked together to identify and make modifications to the geological model, modifying the porosity distributions while taking into account known influences on porosity, such as diagenesis on the flanks of the field. To ensure overall geological consistency, appropriate equivalent multipliers were applied to the permeability and net-to-gross ratio.
These changes were then cascaded into the reservoir simulation model. This mode of working ensured that the simulation model remained consistent with the geological interpretation and increased the team’s confidence in predictions.
The high connectivity between the injector, Well A-21, and the producer, Well A-13, was modeled by enhancing permeability in parts of the model. The distribution of the permeability enhancement was based on seismic and other data. In the Gyda area, an instantaneous amplitude attribute was found to be a good guide to reservoir quality at reservoir depth.Other factors affecting the level of communication in the reservoir were gross reservoir thickness and an overall trend with depth. These factors were considered in mapping the level of permeability increase.
The Petrel Process Manager made it practical for the team to create and simulate multiple geological realizations to understand the implications of uncertainties associated with the static and dynamic models. The variables used by the geologist to generate the maps were parameterized and allowed to vary within the expected anges, making it possible to produce and simulate hundreds of realistic and equally probable geological realizations automatically. Aided by the Petrel keyword editor, the team was able to access ECLIPSE functions to incorporate a compositional fluid description for more realistic modeling. The history-matched model was used to forecast future Gyda South field production in a number of different scenarios and using different injection rates.
Optimize injection and plan new horizontal producer well
The Petrel simulation modeling results showed Talisman Energy that Gyda South production could be maximized by ceasing injection into Well A-21. The secondary gas cap, formed during production to date, was deemed sufficient to support production by expansion— continued water injection into Well A-21 would simply drown the production well. The modeling also showed that a significant amount of oil in place was not accessible using the existing production well.
Based on the findings, the team used the model to design a new horizontal producer well to tap the Gyda South reserves and increase future production. The collaborative environment—using a single application, Petrel software, as an interpretation and modeling tool as well as the front end to ECLIPSE software—enabled the different disciplines within the project team to keep interpretations and models synchronized throughout the study. This ensured consistency and increased the team’s confidence in predictions made using the simulation model.
|
|
|
