Although the recent commodity price fluctuations have exposed the role of geo-politics, world economies and commodity trading in the life cycle of assets, a limited number of field development studies have considered the impact of commodity cycles on the development of in-fill wells. These commodity price fluctuations have resulted in a reduction in drilling and completions activities across the United States. This further implies that in most unconventional plays throughout the US, in-fill well completion (and thus production) will be delayed. This article highlights some of the considerations that should be made when evaluating the well completion strategy with focus on two unconventional plays, the Bakken and the Eagle Ford.
Figure 1. Typical fracture geometry variation expected in a stage.
The key objective of studying the Bakken and Eagle Ford was to model the well performance of the parent wells with the aim of predicting in-fill well performance. Petrophysical and Geomechanical models were developed using data-rich parent well data-sets. Based on the calibration data available (core and logs) numerous static models are generated to capture the range of possible interpretations that honour the calibration data. These models were validated with more calibration data in completions and reservoir domains to reduce the uncertainty space via history matching real data (production or fracture treatments). Once, the parent well models have been validated (by matching well performance history), in-fill modelling and optimisation can proceed. Since the two assets (Bakken and Eagle Ford) are at significantly different portions of the development cycle, the Bakken dataset has the luxury of modelling and matching the performance of the parent and in-fill wells whereas, the Eagle Ford portion of this study focuses primarily on forward modelling and optimising the in-fill well completions.
Figure 2: Comparison of sliding-sleeve (single fracture) and plug-and-pert geometries (dominant cluster of three).
Bakken system parent well modelling
Petrophysical evaluation indicates that the average of porosity and water saturation is 8% and 50% respectively for the Middle Bakken. Average Klinkenberg permeability for the entire Middle Bakken is 0.02 mD. Mercury injection capillary pressure (MICP) data indicated that irreducible water saturation was between 30 - 40% for rock with and residual oil between 30 - 40%. Petrophysical evaluation suggests that the Upper part of the Three Forks had oil potential.
Average reservoir parameters for the upper Three Forks facies are 8% porosity, 60% water saturation and permeability < 0.02 mD. Irreducible water saturation was estimated as 45 - 55% for good and poor quality rock respectively; residual oil is estimated at 40 - 55% for good and poor quality rock respectively. In this area, the middle and lower Three-Forks have higher water saturation, with very low permeability streaks (< 0.007 mD).
Figure 3. Pressure and production history match - the effect of bashing noted at the end of history.
A key parameter in the iteration of various mechanical earth models (MEMs) was the pore-pressure profiles. A pore pressure transition region into the Lodgepole (Scallion formation), was created where a linear pressure increase from salt water gradient up to the estimated pore-pressure of the Upper Bakken Shale. For the TRFK, a linear decrease of pore-pressure from over-pressured in the Lower Bakken Shale to normal pressure was generated within the first 100 ft. The pore pressure remained elevated across the pronghorn and first two TRFK benches, and then reduced back to a salt-water gradient for the remaining TRFK benches.
Fracture modelling - Middle Bakken
For the Middle Bakken: Micro-seismic and the diagnostic fluid injection test (DFIT) data together with interference information from 355 offset fractures with 47 communication events, constrained the expected geometry height growth and an expected fracture length, in both the MB and TRFK. For the Three Forks: micro-seismic data and interference information from 134 fractures with 28 communication events constrained an expected geometry height above and below the TRFK, and a minimum expected fracture length in the MB and the TRFK.
Analysis on the Bakken wells in the study area indicated high (> 1500 psi) net-pressures and low breakdowns. This implies that a plug-and-perf (PnP) completion methodology could yield success by enabling multiple perforating clusters (within a stage) to breakdown, especially when high pump rates (typical of slickwaters) are utilised (Figure 1). When parent well sliding-sleeve (SS) treatments were modelled (Figure 2), a large single dominant fracture was observed (via fracture modelling and production modelling). When parent well PnP treatments were modelled (in the same area), propagation of multiple fractures was confirmed by both fracture and production modelling. Fracture modelling indicated that two dominant fractures were propagating.
Figure 4. Pressure and production history match - the effect of bashing noted at the end of history./small>
For the Bakken model, the upper and lower shale were modelled, together with the Middle Bakken. Daily bottom-hole pressures (BHP) were computed and calibrated to BHP gauge data on offset wells to select the appropriate flow – correlation. Compaction curves were generated from core data and utilised to model fracture conductivity and permeability degradation. The SS Middle Bakken well has a permeability of >0.01 mD, maximum fracture half-length of 250 ft and a dimensionless fracture conductivity that reduces from seven to four. Flowing bottom-hole gauge data was used to constrain the history match (Figure 3). Compaction curves from core and proppant were utilised to model degradation of proppant conductivity and rock compaction (permeability/porosity reduction). An ‘additional damage’ factor (using compaction curves) had to be added after two years of production to match the fluid level measurements. This damage increased with time. The initial hypothesis was proppant crushing (since effective stresses were close to 6000 psi – crushing pressure of sand).
Fracture and production modelling - Three Forks
Fracture modelling in the Three Forks indicated that the SS treatments connected into the Middle Bakken due to the creation of a large single dominant fracture. Production modelling confirmed that the connection to the Middle Bakken was stress sensitive and deteriorated within the first three months of production (Figure 4). Flowing bottomhole pressure and water cut indicated that this well was connected to the Middle Bakken for a significantly longer time than expected (Aman, 2010). The SS Three Forks well had a permeability less than 0.005 mD (top 10 layers of the Three Forks), fracture half-lengths ranging from 100 - 150 ft. and a dimensionless conductivity greater than 10 (Figure 4). The short fracture half-lengths observed were consistent with fracture modelling. In the PnP Three Forks treatments, connectivity to the Middle Bakken was poor due to the addition of multiple clusters per stage (via PnP) and subsequent reduction in job volume per fracture propagating in a stage.
Optimisation with no parent well influence
Once the multi-domain iteration was complete and models with quantified uncertainty ranges built, predictive modelling could begin. For the fracture length and number of stages to be optimised, the bounding (offset) wells were used to ensure accurate forecasting on long-term performance of the middle well.
Forward modelling (optimisation) runs indicated that a hybrid cluster approach could be utilised to maximise fracture propagation success. Optimal job size volumes, fracture treatment rates and stage count were determined for the Middle Bakken and Three Forks. The results indicate that the maximum production is obtained by increasing the number of fractures in a well from X to 4X. Thus, if a SS completion system was utilised, 4X sleeves with entry-points must be installed. If a PnP approach is utilised, fracture modelling has demonstrated the pump rate required to propagate two to five clusters and the resulting job size requirements.3 Optimisation also shows that past 4X fractures, further production uplift can be achieved from increased fracture length (deployment of modified job design).
Bakken system in-fill well/depletion modelling
The in-fill interference of a parent Middle Bakken well (by an offset Middle Bakken well) was fracture modelled and production matched (Figure 3). Production matching of the in-fill drilling and interference of this parent well (by an offset well) confirmed that scale was the damage mechanism since a post-interference match was re-established once this ‘additional damage’ was removed. This observation and the scale treatment history of this well resulted in the hypothesis that the relatively fresh water from the fracture treatments dissolved the scale in the parent well. Numerous cases of in-fill interference have been reported in the basin over the past few years. To date the impacts of interference on the parent well have been primarily related to additional ‘stimulation’. Modelling results indicate that it is unlikely to have sufficient amounts of proppant to travel those distances even under asymmetrical conditions. Thus, scale dissolution may also be contributing to the sustained production enhancement of the parent wells.
As a result of poor productivity in the Three Forks interval, minor fracture asymmetry was observed from the in-fill well. This is due to the poor productivity of the Three Forks interval in the area. However, interference is caused from an upper reservoir (Middle Bakken). To date little research has been performed to understand the productivity and the extent of depletion in the Upper and Lower Bakken Shale.
As production reduces stress but increases effective stress it can be expected that simultaneous fracture operations (driving efficiency) are created to form a similar asymmetry and lack of increased containment. Completing Middle Bakken wells, followed immediately/shortly by an in-fill Three-Forks will result in containment.
Figure 5. Example of depletion profile from asymmetric propagation and fracture geometry as a result of the in-fill drilling between two parents wells in the Eagle Ford.
Eagle Ford system parent well modelling
Workflows and procedures for the characterisation and testing (calibration requirements) of rocks that are both the source rock and reservoir have evolved significantly from those used in conventional plays. Since, pressure, permeability and saturation are much more challenging to measure in thin-bedded unconventional reservoirs using standing logging tools only, multiple measurements and models must be created with the aim of reducing uncertainty.
Petrophysical evaluation indicates that the average of porosity and water saturation were greater than 9% and less than 18% respectively. The petrophysical model results show that the permeability ranged from 100 – 900 nD.
The Ben Eaton stress model with a variable anisotropic Biot model was utilised using calibration from core. An iterative approach engaging fracture modelling, constrained by log responses (that imply possible pore pressure transitions) resulted in an estimation of pore pressure gradient from the Austin Chalk into the Buda.
Fracture and production modelling
In the Eagle Ford dataset, since little in-fill drilling has occurred; limited interference or micro-seismic data exists. A similar workflow utilised in the Bakken Three Forks was utilised to understand fracture geometry and constrain production history match parameters. High net-pressures (>1200 psi during treatment and > 500 psi in the DFIT) in the Eagle Ford enabled the breakdown of multiple clusters. Modelling indicates that some optimisation opportunities still exist within the multi-cluster fracturing scenarios since only 30 - 50% of the fractures within a stage can create dominant fractures. Modelling indicates that if a lateral porpoises across ‘critical’ stress units, the propagation and connectivity within the reservoir changes significantly. Production history matches honoured the relative change in fracture half-lengths within a stage due to limited entry or stress shadowing. Productive fracture half-lengths varied from 50 - 200 ft and permeability ranges from 70 - 500 nD. More than two years of production history was matched on two parent wells.
The importance of layering effects causing pinching and preferential propagation in most unconventional plays are forcing operators to realise that draining a thick reservoir column with multiple laterals can improve production performance significantly.
Eagle Ford system in-fill well/depletion modelling
Figures 3 and 4 demonstrate the early impact of in-fill drilling and interference in the Bakken and Woodford. Overestimation of vertical connectivity across the entire column can result in the under-estimation of the depletion sink.
Once production history matching was performed on the parent well, the new stress state was re-computed to model the impact of fracture propagation between two parent wells. Fracture modelling indicates that production from the upper benches can create improved containment with similar effectiveness using smaller designs and create weak points that result in height growth that would otherwise not be observed under virgin conditions. Based on the forward models, in-fill treatments were designed to minimise asymmetrical effects and direct interference with the offset wells. Figure 5 shows the impact of a larger design-optimised design not shown. Preliminary production results are encouraging and continued monitoring, testing and evaluation will take place.
This study demonstrated that parent well modelling was critical to understanding which wells are within the region of interference. The success of the in-fill well completion design is determined by the ability to characterise the current production system in order to understand the extent of depletion. Modelling was utilised to understand the changes required in operational activities when performing in-fill development. The observation of the potential challenges and re-design of the completion program has resulted in significant success. The introduction of a model based approach to improve decision making will reduce the cycle time between the initial wells drilled and the optimal development strategy.
SM Energy: Nathan Nieswiadomy, Brent Bundy, Sarah Edwards.
Sanjel Corporation: Rafif Rifia, Kristina Kublik, Santhosh Narasimhan, James Gray, Olubiyi Olaoye, Hamza Shaikh.
1. US Energy Information Administration. http://www.eia.gov/oil_gas/rpd/shale_gas.pdf, (April 13, 2015).
2. Cherian, B.V, Nichols, C.M., Panjaitan, M.L., Krishnamurthy, J.K., Sitchler, J., ‘Asset Development Drivers in the Bakken and Three Forks’, SPE 163855; SPE Hydraulic Fracturing Technology Conference, The Woodlands, TX, USA, (4 - 6 February, 2013).
3. Mukherjee, H., Poe Jr., B.D. Heidt, J.H., Watson, T.B., and Baree, R.D., ‘Effect of Pressure Depletion on Fracture Geometry Evolution and Production Performance’, SPE 65064, SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, (22 - 25 September, 1995).
4. Ajani, A., Kelkar, M., The University of Tulsa, ‘Interference study in Shale Plays’, SPE 151045, SPE Hydraulic Fracturing Technology Conference, the Woodlands, Texas, (6 - 8 February, 2012).
5. Waters, G.A., Dean, B.K., Downie, R.C., Kerrihard, K.J., Austbo, L., McPherson, B, ‘Simultaneous Hydraulic Fracturing of Adjacent Horizontal Wells in the Woodford Shale’, SPE 119635; SPE Hydraulic Fracturing Technology Conference, The Woodlands, Texas, (19 - 21 January, 2014).
6. Wheaton, B., Miskimins, J., Wood, D., Lowe, T., Barree R., ‘Integration of Distributed Temperature and Distributed Acoustic Survey Results with Hydraulic Fracture Modelling: A Case Study in the Woodford Shale’, URTeC 1922140. SPE/AAPG/SEG Unconventional Resources Technology Conference, Denver, Colorado, USA, (25 - 27 August, 2014).
7. Lalehrokh, F., and Bouma, J., Talizman Energy USA, ‘Well Spacing in Eagle Ford’, SPE 171640, SPE/CSUR Unconventional Resources Conference – Canada, Calgary, Alberta, Canada, (30 September - 2 October).
8. Ganpule, S., Srinivasan, K., Izykowski, T., Luneau, L., and Gomez, E., ‘Impact of Geomechanics on Well Completion and Asset Development in the Bakken Formation’, SPE-173329. SPE Hydraulic Fracturing Technology Conference, the Woodlands, Texas, (3 - 5 February, 2015).
9. Heller, R., Vermylen, J., and Zoback, M., ‘Experimental investigation of matrix permeability of gas shales’, AAPG Bulletin, v. 98, no. 5 (May, 2014), pp. 975 - 995.
10.Narasimhan, S., McCleary, M., Fluckiger, S., Gray, J., Cherian, B., Shaikh, H., ‘Using the wrong method to estimate stresses from depletion causes significant errors in predicting wellbore integrity and fracture geometry’, SPE 173310, SPE Eastern Regional Meeting held in Morgantown, West Virginia, USA, (13 - 15 October, 2015).
11. Aman, B. Williston basin Petroleum Conference- 2010. Bakken-Three-Forks connectivity study. http://wbpc.ca/pub/documents/archived-talks/2010/Brad_Aman_-_Continental.pdf
12. Waters, G.A., Presentation at The Society of Petroleum Engineers’ Gulf Coast Section http://www.spegcs.org/media/files/event/dcf55310/. InfillDevelopment-Handout.pdf
13.Cherian, B.V., Stacey, E.S., Bressler, S. et al. 2012.Evaluating Horizontal Well Completion Effectiveness in a Field Development Program. Paper SPE 152177 presented at the SPE Hydraulic Fracture Technology Conference, The Woodlands, Texas, USA, (6-8, February).
14. Eva Peza, Erik Kvale, Richard Hand, Wallace Harper, Ram Jayakumar, Eric Wigger, Barry Dean, Ziad Al-Jalal, Sameer Ganpule, 3-D Integrated Workflow for Understanding the Fracture Interference and Its Impact into the Gas Production of the Woodford, URTeC 1923397, Unconventional Resources Technology Conference, Denver, Colorado, (25 - 27, August 2014), pp. 247-272.
15. Eagle Ford Shale Is Getting 'Thicker' - Upper Eagle Ford Emerging As A Viable Second Development Zone, (August 15, 2014) SeekingAlpha
16.Gupta ,J., Templeton-Barrett, E., Yao, Y., Choi N., Zielonka M., Burnham H., Jackson S., El-Rabaa W., Albert R., Integration of Fracture, Reservoir, and Geomechanics Modeling for Shale Gas Reservoir Development, SPE-164018-MS, SPE Unconventional Gas Conference and Exhibition, Muscat, Oman, (28 - 30, January).
17. Walls, J.D, and Sinclair S. W., Eagle Ford shale reservoir properties from digital rock physics. Rock Physics and Formation Evaluation, First Break –EAGE. Volume 29, Issue 6, (June, 2011).
Edited from an article byBilu V. Cherian, Sanjel Corporation, Mathew McCleary and Samuel Fluckiger, SM Energy
Read the article online at: https://www.oilfieldtechnology.com/special-reports/06062016/parent-well-modelling/