A Systematic Multidisciplinary Approach for Optimization of Brownfield SAGD Projects – Part 2
WRITTEN BY: MARK SAVAGE, P. L. Geo. and SHAHBAZ MASIH, P. Eng.
This article is Part Two of a two-part paper which focuses on a multidisciplinary optimization approach to drilling new well opportunities in existing pads. In June 2020, Part One addressed optimization of existing pads by modification in well completions and operational strategy. This article will focus on how a multidisciplinary approach can be used for adding infill wells, pad extension wells and/or farmer wells based on updated reservoir characterization, heat transfer and other observations from operational performance data. It will also discuss how optimization of a brownfield project can be achieved for performance forecast, well design and completions, facilities modifications for anticipated changes in steam requirement and treatment of produced fluids and operational optimization of new wells by using a multidisciplinary approach.
Reservoir heterogeneity can present a challenge, e.g. uneven heat distribution along the well length and in the reservoir during operation and optimization of Steam Assisted Gravity Drainage (SAGD). Based on analysis of geoscientific and well performance data, this inefficiency may lead to modification in well completion and operational strategy, as discussed in Part One.
Optimization of brownfield projects directly targets increased oil rate and decreased steam oil ratio (SOR). The geological models (geomodels) used for numerical simulation of production and SOR forecasts, economic justification, and designing well pads and facilities depend on the quality of data from strat wells, seismic, geoscientific interpretation and well performance. Analysis of geomodels and reservoir engineering interpretations is used to evaluate the optimal number of wells and well placement for the drainage area (Figure 1). However, the operation of SAGD wells and interpretations of 4-D seismic, vertical observation well (Obs) data, temperature logs and continuous geoscience and reservoir/production engineering learnings, will give rise to updated reservoir characterization and/or geomodelling (Figure 1). Supplemental subsurface data (4-D seismic and Obs data) can help identify reservoir pay thickness suitable for new well placement.
Because regions of untapped heat transfer exist in a brownfield project, the benefits of continuous steam injection might not be realized due to limitations of existing well placement, steam generation, and fluid handling infrastructure. An updated reservoir characterization can identify regions of untapped heat transfer and lead to increasing brownfield oil rate and decreasing SOR through effective placement of new wells and/or upgraded infrastructure (Part One).
These optimization changes may lead to adjustments in well completions, steam requirements, surface facilities, and future well completions (Figure 1).
Figure 1 Impacts of Reservoir Characterization Updates
Operational Data Analysis and Changes/Updates in Reservoir Characterization
Production performance of a SAGD operation depends on effective steam usage and heat transfer to bitumen. Steam mobilizes the bitumen along the steam chamber interface to facilitate gravity flow to the production well. In early stages of a SAGD well pair’s life, temperature logs provide insight into heat distribution along the well length. Temperature logs along the production well length are the first indication of variable behaviour of permeability or heat conductivity in different reservoir facies. The reservoir facies originally considered to be permeable may in fact be a barrier or a baffle to convective heat propagation. Low quality reservoir facies may direct steam in unexpected ways, and lead to inefficient heat distribution and a higher SOR. Obs well data (temperature, pressure and/or saturation logging) is a useful tool to determine convective and conductive heat transmission in the reservoir. An additional advantage of Obs well data is its positioning relative to heat sourced from SAGD injection well. Unlike temperature logs, Obs well data is not distorted by proximity to a heat source or to intermittent changes in steam injection or to short term shutdowns. 4-D seismic is another data type that aids in the analysis of reservoir heat distribution. The investigation volume measured with 4-D seismic is more expansive than either horizontal well log or vertical Obs well log data. Obs well temperature and saturation data and production well temperature data are used to calibrate 4-D seismic data. Analysis of calibrated 4-D seismic can be used to map steam transmission and heat transfer in the reservoir to identify new brownfield drilling opportunities.
Production data analysis may indicate premature coalescence of steam chambers and heat entrapment in places where existing SAGD well pairs cannot access mobilized bitumen. The analysis of 4-D seismic and Obs well data can identify where heated mobilized bitumen occurs that well pairs cannot access, e.g. outside the drainage area or between and below current producers. Identification of inaccessible mobilized bitumen presents opportunity for drilling new infill, pad extension, and farmer wells (Figure 2). Incremental production provided by the additional new wells requires deployment of increased steam capacity or reallocation of existing steam from suboptimal performing wells. Learnings from operation of existing wells, operational data analysis, and updates in reservoir characterization can also identify supplemental optimization changes to brownfield well design, follow up well completions, and artificial lift.
The steam requirement increase may necessitate modifications in steam/power generation system, water supply system, water treatment system and water disposal system (Figure 2). The increase in produced fluid may lead to changes in oil treatment system, bitumen storage and delivery system, produced gas volume, water treatment and disposal systems (Figure 2). The description of these major systems and the disciplines responsible have been described in Part One of this article. The discussion section of this article will describe three types of additional brownfield wells and corresponding operational learnings that lead to the addition of these wells. It will also describe how new learnings can impact operation/modification of major components of a SAGD project (Figure 2) and how to handle those changes and finally, operational optimization of new wells in a multidisciplinary approach.
Figure 2 Additional wells and their impact on main components of SAGD system
Example 1 – Infill Wells
Infill wells are horizontal producers placed between producing SAGD well pairs to access a wedge of heated bitumen (Figures 3 and 4). The timing and placement of infill wells (horizontal and vertical) are a function of:
- reservoir characteristics (saturation and permeability),
- reservoir architecture (thickness, structure, and degree of heterogeneity), and
- interwell spacing and performance (steam chamber conformance and production) of existing well pairs and facility steam and fluid handling capabilities.
Figure 3 Infill Wells – Simplified Plan View Figure 4 Infill Wells – Conceptual Cross-sectional View
Optimal timing and number of infill wells per pad involves multidisciplinary input from geoscience, reservoir and production engineering, drilling and completions (D&C), and operations and maintenance (O&M). O&M and production engineering manage the existing SAGD well pairs’ daily production; for example, steam requirements, production rate, and fluid processing. The optimal timing of an infill well’s first production is influenced by O&M along with production and reservoir engineering in assessing future steam and fluid handling demand versus available steam and fluid handling capacity. Steam and fluid handling capacity is influenced by actual performance of the existing well pairs’ fluid rates and SOR, versus the predicted and potential forecasted range (Figure 5), and by the limitations of surface facilities. If SOR is higher than predicted, steam availability may be limited, thereby influencing the timing and number of infill wells.
Figure 5 Conceptual SAGD Well Pair Performance – Predicted (Black line) vs Actual (Solid Dark Green line) vs Forecasts (Dashed Light Green lines)
Reservoir and production engineering analyze daily operational data from SAGD well pairs and Obs wells (Figure 6), to track, interpret, and estimate steam chamber development and the coalescence timing of adjacent chambers.
Figure 6 SAGD Obs Well Temperature Data over Time (Source: AER D54 2019 8668) Gamma Ray (Black line), temperature timelines: timeline 1(Gray line), timeline 2 (Purple line), timeline 3 (green line), and timeline 4 (red line).
Coalescence of steam chambers in adjacent well pairs is indicated by similar steam chamber pressure trends or equal steam chamber pressures. This can lead to heat transfer outside the drainage area and provides little or no contribution to bitumen recovery. Continuous steam injection for 3 to 4 years can lead to accumulation of heat and mobilized bitumen in the reservoir between two adjacent well pairs that is inaccessible to existing producers and leads to increased SOR. Reservoir characterization updated with revised interpretations from 4-D seismic and Obs well data can indicate favourable conditions for the placement and production of new infill wells. Geomodels (Figure 1), updated by geoscience can help reservoir engineering improve field production forecasts, and estimates of steam and production handling requirements; for example, the amount of additional oil and water to be treated, additional diluent and disposal requirements, and the additional capacity for bitumen storage and delivery (Figure 2). Typically, all these variables lead to a first oil window during the early to mid plateau stage of a drainage area’s productive life. D&C should involve the collaborative input of reservoir and production engineering and geoscience in the well design and selection of artificial lift.
Infill well placement planning is a collaborative effort involving geoscience, reservoir and production engineering, and D&C. The horizontal placement of infills is typically equidistant between producing well pairs (Figure 3). Vertical placement is influenced by architecture and characteristics of reservoir geology, SAGD well pair producer trajectory and corporate stand-off philosophy (i.e. height above base of pay or bottom water), Figure 7. Key reservoir considerations influencing the infill’s vertical placement are the base of pay structure (elevation and directional trend) and nature of the contact (mudstone/non-pay, Devonian carbonates, or bottom water), presence of high mobility zones, e.g. transition or lean zones, and the degree and placement of reservoir heterogeneity (Figure 7). Operational considerations influence the stand-off height considering the impact of heat loss on SOR and water cut.
Figure 7 Infill Wells – Vertical Placement Relative to SAGD Well Pair
Optimal vertical well placement is unique, it is rare if two infills are positioned the same (Figure 7). Vertical placement influences infill economics and the proportion of accelerated versus incremental reserves. The lower the infill well is placed relative to the SAGD producer, the more incremental reserves can be produced. Economic and operational analysis will justify the number and timing of infill wells. The plan of each infill well trajectory involves collaborative work by geoscience, reservoir and production engineering, and D&C. To optimize the infill’s vertical position requires integrated interpretation of geoscience data, updated geomodel and reservoir engineering simulations, steam chamber conformance and heat transfer analysis from 4-D seismic, Obs well data and temperature logs and, existing well pair performance.
Operational Optimization of Infill Wells
Infill well performance depends on its level of connectivity to heated bitumen between coalesced steam chambers. For start-up of infill wells, establishment of connection may require multiple periods of cyclic steam stimulation and demand availability of temporary infrastructure (e.g. temporary steam lines and supporting well completions) for switching the well from injector to producer. The existence of high mobility transition and lean bitumen zones in the vicinity, reservoir heterogeneity, and the amount of innate heat due to steam injection in the adjacent injectors may impact the start-up process.
Infill well start-up is considered almost complete when it has established communication with drainage caused by the overlying steam chamber as indicated by produced fluid temperatures, production continuity and other operational data. The infill well creates its own drainage mechanism and will alter the steam chamber shape of adjacent well pairs. The heat source for infill well production comes from steam injectors in adjacent well pairs. Fluid rates and SOR performance data of adjacent well pairs and infill wells should be analysed in combination for optimization purposes. The connection of an infill well to the steam chamber may lead to changes in steam chamber pressure and thereby change steam injection rates in the adjacent injectors. Changes in steam injection rates and production from an infill well can also lead to necessary changes in operational and artificial lift design in adjacent producers due to their impact on volumes drained from the steam chamber. Analysis of steam allocation by reservoir engineering is required where steam shortages or other plant issues that pre-exist an infill well could impact steam availability and fluid handling capacity. In conclusion, additional production from an infill well may cause changes in all major components of a SAGD brownfield project (Figure 2).
Example 2 – Pad Extension Well Pairs
Pad Extension well pairs are SAGD well pairs (injector and producer) placed outside the original drainage area to access incremental bitumen (Figures 8 and 9). The timing and placement (horizontal and vertical) of these wells are a function of multiple parameters, as is in the case of infills.
Figure 8 Pad Extension Well Pair – Plan View Figure 9 Pad Extension Well Pair – Cross-Section View
To assess the wells’ optimal placement and performance forecast requires a multidisciplinary effort, as is outlined for infills in the previous section. A pad extension well pair is usually drilled to access part of the reservoir that was previously considered not suitable for SAGD operation. Updates in reservoir characterization data from temperature logs, Obs well data and 4D seismic data and operational data may identify an opportunity for a new well pair outside the original drainage area. For example, 4-D seismic, operational data and/or updated facies characterization may indicate thicker pay close to the original drainage area.
The existing production data can also be used for history matching, for reservoir parameters, and for fine-tuning and generation of performance type curves in reservoir simulations. Optimal timing of first production is predicted with O&M and production engineering assessment of future steam and fluid handling requirements, compared to availability. The steam requirement for a pad extension well pair could be like the offset well pairs, and considerably higher than an infill well. If existing well pair SOR is higher than predicted, steam availability may be limited, thereby influencing the timing for start-up of pad extension well pairs. The extra volume of steam and produced fluids may require modifications in the steam generation system, oil and water treatment systems, the disposal system and storage and delivery systems (Figure 2).
The updated reservoir characterization data, geomodels and learnings from operational data may lead to modifications in well completions, e.g. placement and type of flow control devices and artificial lift design, for both injector and producer of the extension well pair (Figure 1), and may require coordination between multidisciplinary teams as explained in the infill wells section. The placement of extension well pairs is influenced by reservoir geology and the trajectories of existing well pairs, and is also handled by a multidisciplinary approach as explained in infills section.
Operational Optimization of Pad Extension Wells
Start-up and SAGD operational phase of the well life cycle can also be enhanced from the learnings of other wells in the same pad or from other pads in the same field (Figure 1). The multidisciplinary approach for optimization of the pad extension wells can be utilized as explained in Part One of this article, previously published.
Example 3 – Farmer Wells
Farmer wells are sidetracked horizontal producers placed below the SAGD well pair producer to access heated bitumen (Figures 10 and 11). The heated bitumen can be identified from Obs well and 4-D seismic data. The presence of untapped mobilized bitumen below the producer may also be identified from updated geoscience reservoir characterization and learnings from drilling (for example, stand-off height) and operations of the existing well pairs.
Figure 10 Sidetrack Farmer Well – Longitudinal Cross-Section View
Figure 11 Sidetrack Farmer Well – Perpendicular Cross-Section View
The optimal timing for farmer wells is the wind down stage. The horizontal and vertical placement, completion and operation of farmer wells require a multidisciplinary effort similar to infill and extension wells, and as explained in the previous infill well section. The key difference for farmer wells is the interpretation and mapping of the heated bitumen zone below the producer’s elevation and the zone’s areal extent. The interpretation of heated bitumen regions using 4-D seismic with 3-component geophones profoundly improves optimization of farmer well placement. In general terms, and dependent upon reservoir quality, saturation, and vertical placement, the ability to put a producer well lower across a drainage area can equate to an additional range in recoverable bitumen of 300,000 to 800,000 barrels.
Operational Optimization of Farmer Wells
To our knowledge, farmer wells have not yet been deployed in the field, so operational predictions are made based on the author’s experience from infills, extension wells and SAGD well pairs. Similar to an infill well, the start up of a farmer well will depend upon the amount of heat present along the well’s trajectory, and access to drainage from an overlying steam chamber. The closer proximity of a farmer well than an infill well, to its SAGD well pair can heighten communication with bitumen mobilized by conduction below the overlying steam chamber. If a farmer well is placed in immobile bitumen, it may need cyclic steam stimulation to establish communication with the overlying steam chamber. Operating a farmer well can be similar to operating a SAGD producer well, once fluid temperature and production rate indicate farmer well start up is complete. Any production from farmer wells will increase the steam chamber volume. As a result, in order to achieve a targeted steam chamber pressure, planning for an increased steam volume may be required.
Impacts on the main components of a SAGD brownfield operation (Figure 2), can be optimized by using a multidisciplinary approach, as was discussed in Part One of this article.
Optimization of brownfield pads or individual wells comes in many forms and with multiple options. The appropriate optimization components can vary across a drainage area or an entire field. The optimal strategy requires collaborative input from geoscience, reservoir and production engineering, D&C and O&M.
The potential of interpreted 4-D seismic to identify and enhance assessment of prospective incremental reserves and improve well placement of one of the three well types (infill, pad extension or farmer) significantly outweigh the cost of seismic acquisition and processing. Ongoing updates of the reservoir characterization process during the ramp-up and plateau stages with new geoscientific data and operational learnings can enhance economic viability of remedial action.
A collaborative multidisciplinary process with defined roles and responsibilities will enhance understanding of which additional well type (infill, pad extension or farmer wells) is most advantageous for a drainage area.
Input provided by Gordon T. Stabb, P. Geol., towards writing this article is greatly appreciated. The authors thank Gordon for his long serving support of CHOA and our industry.
Mark Savage is an APEGA registered P.L. Geo. member. He has been in the oil sands industry since 2000. Mark started his oil sands career with Petro-Canada working on the Lewis, MacKay River and Fort Hills projects. Since leaving Petro-Canada in 2008, he has been actively engaged in various in-situ oil sands assets with Ivanhoe Energy Ltd., Statoil Canada Ltd. and Athabasca Oil Corp. He has collaborated on and provided hands-on leadership of geoscience, operations, and development teams on multiple SAGD projects.
Shahbaz Masih is an APEGA registered professional engineer with M.Sc. from University of Regina (Canada) and B.Sc. from UET Lahore (Pakistan), both in Petroleum Engineering. He is currently a machine learning engineer at Intelius Analytics Ltd. and is pursuing a Master of Data Science and Analytics from the University of Calgary. He has worked in the Canadian oil industry for more than ten years with Cenovus Energy, Statoil Canada Ltd., and Sunshine Oilsands Ltd. Highlights of his professional expertise include analytics by using data science and big data tools, reservoir simulation for SAGD, CSS and solvent assisted processes for process optimization and project development, holistic optimization of SAGD operation and its well completions, reservoir management by using a multidisciplinary approach, ESP’s design, operation and failure analysis, production management in coordination with other stakeholders, and an NSERC research project focused on porous media flow in SAGD.