Reservoir Characterization Process - Vol. 3

Reservoir Characterization Process – Volume 3 – Simulation and Risk Stages

This post completes my discussion of the reservoir characterization process. I refer you to my November 28th post reviewing the data collection stage and the December 14th post reviewing the petrophysics and geomodel stages.

My oil industry experience is in the geoscience discipline working in or leading geoscience and multi-disciplinary teams focused on the development and production of oil and gas assets. I have conventional, unconventional and oil sands experience. This post will focus on my experience, both hands-on and leading the reservoir characterization process. 

The reservoir characterization process (the process) is a fundamental component for many asset subsurface evaluation and assessment deliverables. The extent of the detail required for the process should be determined by the overall project scope and schedule. 

I will describe the process that I have used for major decision gate (DG) milestones, DG1 (concept), DG3 (project sanction) and for the resource/reserve assessment and validation. A thorough understanding of the assets potential and uncertainty can be gained by following the process outlined below and both the Nov. 28th and the Dec. 14th posts. For the purpose of this post I will discuss the considerations from a steam assisted gravity drainage (SAGD) perspective. With some modifications the workflow and/or considerations cited could be used across various asset types and project scopes. 

A key component of the process is the multi-disciplinary team’s composition and mandate. As a minimum, the team should be comprised of the following disciplines: geology, geophysics, petrophysics, geomodeller and reservoir engineering. Depending on the project specifics other specialists may also be required: hydrogeologist, risk analyst, finance, etc. For the best results, the team needs a work environment that encourages collaboration and openness to ensure the end product represents and quantifies a holistic range of uncertainty, risk and opportunity.  

There are five stages to the process workflow: 

• Data Collection
• Petrophysical Analysis
• Geomodel scope, discussion and construction
• Simulation, proxy curves and forecast
• Risk Analysis, documentation and mitigation

The products from each stage of this workflow will heavily influence the degree of uncertainty captured in the resultant performance forecasts. These forecasts can be used to assess and defend the in-place resource and the recoverable reserves, the greenfield or brownfield expansion project’s optimal facility design capacity, the number of wells required for a field/pad start-up and the schedule for subsequent sustaining pads. The latter should include learnings and required modifications resulting from the producing wells performance, e.g. update 3-D geomodels and simulations to adjust the sustaining well/pad schedule.

For the remainder of this post I will discuss the Simulation and Risk Analysis stages of the reservoir characterization process. 

                                                Simulation Stage

The learnings and recommendations that I discuss are from directly working with and/or leading reservoir engineers throughout my career. The simulation derived forecasts is the essential component of any reservoir evaluation and assessment process. This stage of the process provides the forecasts and the potential uncertainties that will be utilized in the generation of the project’s economic assessment.

Outlined below are a few of the key discussion points that I have experienced for the simulation stage of the process, my recent experience and learnings with regard to reservoir simulation has been associated with the steam assisted gravity drainage (SAGD) process . This is not meant to be a description of a recommended workflow for your simulation process. In my experience most companies have their own defined simulation workflow that is followed by the reservoir engineers in order to produce a forecast and/or history match.

As discussed in my prior posts (Nov. 28th and Dec. 14th) on the reservoir characterization process, the simulation stage of the process is the culmination of a data intensive workflow that requires ongoing input and collaboration from all members of the multi-disciplinary evaluation team. In my experience it takes the multi-disciplinary technical review of the simulation’s cross-sectional images (saturation, permeability, temperature, etc.) to determine the validity of the cited realizations performance.

As with both the petrophysical and geomodeling stages, early discussion and decisions during the project scope phase should confirm the simulation grid (sim grid) cell size and the area of interest, well, pad or full field. Keeping the sim grid cell size the same as the geomodel cell size will allow for a better representation of the reservoir’s architecture but bear in mind that the project schedule and/or the computing capability may not allow for this.

The relative permeability and the pressure-volume-temperature (PVT) data inputs are a critical set of parameters for the simulation stage in addition to the parameters populating the geomodel. I have seen PVT results vary across a field and across a region; I encourage you to collect PVT data for your specific reservoir and field. Confirming the appropriate relative permeability and PVT inputs prior to initiating the simulation process is vital to both the project schedule and the final forecast. If there is a range in either parameter this should be included in the uncertainty analysis.

Another early consideration and decision is which probability realizations will be used in the simulation process. If time is not a factor it is ideal to run all the geomodel realizations, but this is typically not the case. I recommend using the P10, P50 and P90 sim grids as the minimum number of realizations to run. This will allow some understanding regarding the range of uncertainty attributed with the reservoir.

In the SAGD dominion the key parameters influencing the ultimate performance are the degree of heterogeneity, vertical permeability, temperature and saturation. The saturation, vertical permeability and heterogeneity are obtained from the geomodel and are typically represented in an upscaled sim grid. The temperature in which the field is operated is a function of the reservoir depth and the fracture gradient of the cap rock. I have experienced cap rock fracture gradients ranging anywhere from 14 kPa/m up to 21 kPa/m, the higher the value the better. The higher fracture gradient value allows the operator to use a higher operating pressure and therefore a higher temperature. Optimization of the operating pressure strategy should occur over the well’s life and/or at the different stages of development: circulation, ramp-up, plateau, decline and blow-down. The optimization of the operating pressure is an easy and low cost method to reduce the field’s steam-oil-ratio (SOR). Optimization of the operating pressure and temperature should be evaluated and confirmed prior to completing the production forecast.

The vertical permeability (Kv) and heterogeneity are codependent parameters. An increasing level of heterogeneity (mud bed volume) will reduce the flow unit facies (facies) Kv, which in turn can impede the steam chamber’s vertical development and the reservoir’s ultimate recovery. This is a simplistic explanation regarding a very complex reservoir architectural phenomenon. It is further complicated by the deposition of younger incising channels creating uncertainty regarding the inter-well correlation length of the facies. The geomodel and the resulting reservoir sim grid can capture most of this complexity but this is only with the inclusion of 3-D seismic. The cannibalization and stacking of these channels will directly influence the simulation forecast results by manipulating the steam chamber’s lateral and vertical development.

In more heterogeneous reservoirs the horizontal permeability (Kh) may be the dominant parameter controlling the initial steam chamber development, until the facies’ Kv value reaches the threshold to promote vertical steam chamber growth. Understanding this concept when analyzing the simulation derived forecasts will aid in the quality assurance process.

The sim grids are populated with the facies and corresponding petrophysical parameters which are derived from the upscaled geomodel realization. If there is the need to manipulate any parameter to achieve a match or a desired forecast shape, I recommend starting with the Kv/Kh ratio values. The hydrocarbon pore volume parameters have the least amount of uncertainty and therefore should be the last parameters manipulated in the simulation process. The multi-disciplinary team should discuss and agree upon the allowable range of parameter modification for each facies unit before manipulating the sim grid values. Understanding the Kv threshold value at which a facies becomes an impediment to steam rise is critical to ensuring the forecast remains representative. Some understanding of this threshold value can be gained from the producing pads observation well data which is collected from the wireline saturation logs, temperature data, post-steam flood core and the 4-D seismic. Depending on when this data is collected in a pad’s production life the results may only account for the convective heating process affects. Post steam-flood core has demonstrated that the conductive heating process also contributes to the SAGD pads performance; see my Oct. 28th post titled Are We Discounting Recovery Potential.

Before finalizing the forecast the multi-disciplinary team should review the simulation’s forecast saturation and temperature results at various time periods over the field’s production life in order to indentify inconsistencies and anomalies. Understanding the volumetric and resource/reserve contribution by facies is another check that should be completed before finalizing the forecast.

                                                  Risk Analysis Stage

The learnings and recommendations that I discuss come from directly working on and/or leading a number of reservoir characterization projects throughout my career. These are the subsurface risks that can be associated with the SAGD reservoir. Below are a number of risk analysis discussion points that I have experienced during my career. This is not meant to be a description of a recommended workflow for your risk assessment process.

A project risk analysis by all participating disciplines and cross-functional groups should be conducted, documented and reviewed/updated regularly. If possible use a common risk tool across all participating cross-functional groups. There are several commercial options available but an Excel spreadsheet will also work. Be sure that all groups also use a common set of guidelines to quantify the individual risk probability and consequence.

Risk comes in many forms and with varying degrees of potential impact (consequences) and chance of occurrence (probability) on a project’s outcome.

The subsurface risks that can materially impact a project’s forecasted recovery are as follows: misrepresentation of the in-place resource, simplification of the reservoir architecture, the simulation derived recovery factor, sterilization of resource due to a regulatory ruling, the water source and water disposal potential.

The risk of misrepresenting the in-place resource is related to two reservoir components: pay height and/or area. The risk for both of these parameters is directly related to the depositional environment (fluvial – estuarine) and the associated facies. Specifically the presence, vertical positioning and lateral continuity of the mud bed dominant facies (Inclined Heterolithic Stratification, IHS). In my experience some of this risk and associated uncertainty originates from the ability to consistently define the reservoir facies from core and/or the geophysical wireline logs. For Tier 1 (high quality) reservoirs this risk is typically minimal and most often restricted to the upper few meters of the pay interval. For the more challenging reservoirs with increased heterogeneity this risk can be more substantial. This risk is further heightened in the inter-well regions.

In the McMurray Formation’s pay interval the hydrocarbon pore volume parameters typically have a minimal range of variance. The risks associated with the data collection and data analysis process are related to the equipment (calibration) and the analytical (sample selection and preparation) process. The saturation and porosity geophysical wireline equipment has a minor error range associated with the data acquisition (calibration and wellbore condition). These data related risks are accounted for during the petrophysical analysis stage of the reservoir characterization process.

It is my recommendation that the in-place resource volumetrics be generated using two different techniques: 2-D mapping and the 3-D geomodel. Each technique will generate a P10, P50 and P90 volumetric range. This allows the team to understand the range of volumetric variance attributed with each technique and can then be compared to one another. If there is a large discrepancy between the two techniques this can alert you to a potential area that may require further investigation, such as the geomodel’s pay surfaces.

The risk of reservoir architecture simplification is very real and is heavily influenced by the depositional environment and the data available, specifically well density, the coverage of core, borehole logs and 3-D seismic. The fluvial estuarine fining upwards depositional environment is a very complex system with heightened lateral variance. To further complicate matters, it is rarely only one stacked channel system. Numerous younger channels incise and cannibalize the older channels, introducing greater uncertainty regarding the lateral continuity of the IHS beds and increasing the vertical heterogeneity. This degree of reservoir complexity can only be captured with representation and integration from multiple data sources, wells and seismic. This data should be used to develop a conceptual geological model representing the evolution of the reservoir’s architecture. For Tier 1 (high quality) reservoirs, this risk of simplification is typically minimal and is often restricted to the top portion of the reservoir. For the more challenging reservoirs with increased heterogeneity this risk of simplification can be more substantial and the risk is further heightened in the inter-well regions.

It is my recommendation that the reservoir architecture incorporate all available well and seismic data to develop a conceptual geological model. The use of 3-D seismic can assist in identifying large geomorphologic and inter-well features that can influence the reservoir’s performance. This conceptual geological model can be used to assist in the construction of the 3-D geomodel with object based features, such as abandonment channels and/or channel versus regional facies boundaries, etc. You can reduce some of the simplification and misclassification associated with electrofacies by using core and borehole image logs to derive the flow unit facies.

The risk associated with the simulation derived recovery factor is a cumulative effect of the reservoir characterization process. Any bias, misinterpretation or incorrect piece of data that is incorporated into the 3-D geomodel and sim grid will influence the predicted recovery factors. Other than the two risks outlined above, there are two other parameters that require further scrutiny. The vertical permeability and the Kv/Kh ratios utilized for the individual facies, specifically the IHS facies should be reviewed. Another area for investigation is the conductive heating contribution, which will be reservoir and/or pad specific as referenced in my Oct. 28th post titled Are We Discounting Recovery Potential.

It is my recommendation that an uncertainty analysis be conducted and incorporated into the simulation’s final results to help account for the identified risks attributed with the key data. The uncertainty analysis can incorporate up to five parameters (permeability, realization, OBIP, relative permeability, etc.). Another check to conduct is a review of the simulation saturation cross-sections at various time periods of a pad’s production life with a focus on the IHS facies. Be sure to cross reference the facies specific saturation changes against similar facies performance in the producing pads.

The risk associated with sterilization of the resource due to a regulatory ruling is difficult to quantify both from a risk probability and consequence perspective but this potential risk should not be ignored. Suffice it to say, the probability associated with this risk may be low but the consequence of sterilization is high. This risk can result from an issue associated with the drilling of a single vertical or deviated evaluation well (single drainage area) up to an operational issue (full field). This can be further complicated by the presence of a potable aquifer or in an extreme case when a surface event has an effect on other stakeholder interests.

It is my recommendation that you review your drilling and field operations best practices. For the drilling review you should focus on the maximum allowable angle of deviation and the well path trajectory for the evaluation wells. When evaluating the field operations best practices, a focus on the maximum operating pressure and ensuring that the regular operations don’t exceed the pressure threshold.

The risk attributed with the water source and water disposal is the lack of areal extent and the resource is in a finite supply. A potential secondary risk associated with the water disposal component is the possibility of a pressure effect on the reservoir. The SAGD process can be a water intensive process, so ensuring that there is an adequate water source for the duration of the field’s development life is critical. With the increasing number of proposed SAGD projects and the present day drawdown on several of the key aquifers it has been estimated that there may only be 20+ years of available source water.

The high water recycle rate, +90%, presently demonstrated by most of the SAGD producers will aid in minimizing the strain on the aquifers. The introduction of solvent, non-condensable gas co-injection or other emerging development techniques will also aid in reducing or even eliminating the water requirements.

The water disposal risk comes in two forms: limited areal extent and the potential to directly influence the reservoir’s operating pressure. This risk is further enhanced when the disposal zone is also a direct contact bottom water zone for the producing asset.

Having multiple disposal wells positioned as far from the producing pads as possible can help manage the operating pressure impact. The preferred disposal solution would be a deeper disposal zone, such as the Devonian Keg River Formation, Prairie Evaporite Formation salt caverns, etc.

In closing, the reservoir characterization process is a critical component of the overall asset evaluation and assessment process. Ensure that you spend adequate time defining the project scope and ensure that you discuss, debate and design the key deliverables for each of the stages. Always include the multi-disciplinary and cross-functional groups in all stages of the project.  Ultimately, it is the integration of the different data sources, the petrophysical analysis and the geomodel coupled with open and the on-going collaborative communication that will lead to a more representative range of simulation results, including opportunity, uncertainty and risk. 

This is my last reservoir characterization process post and I thank you for sharing your comments, feedback and suggestions. As always, your thoughts and insights are welcomed and appreciated.