Analog

Key Takeaways

• A reservoir analog must be selected on the basis of geological similarity across multiple dimensions simultaneously, including depositional environment (same facies associations and energy regime), diagenetic history (comparable burial depth, temperature, and cementation), structural setting (similar trap style and stress regime), fluid type (oil vs gas vs condensate), and driving mechanism (solution gas drive, water drive, or compaction drive), because similarity in only one dimension does not predict the other properties that collectively determine well performance: A common analog selection error in WCSB Cardium horizontal well evaluations is using the Pembina Cardium D3 pool (the most productive and best-documented Cardium pool, with over 2,000 horizontal wells and 30+ years of production history) as the analog for Cardium wells 100 km north in the Gilby or Ferrier areas, which have comparable porosity and apparent lithology but different shoreface orientation (relative to dominant longshore transport direction), different water saturation (higher Sw at Gilby due to shallower reservoir depth), and different capillary pressure characteristics (finer grain size at Gilby reduces permeability below the Pembina baseline). Wells drilled at Gilby using Pembina type curves overestimated EUR by 30 to 60% in multiple operator programs through the early 2010s before field-specific type curves were established, illustrating the importance of multi-attribute analog screening rather than single-attribute similarity matching.
• Outcrop analogs supplement subsurface data by providing direct measurement of geological properties inaccessible from wellbores, including lateral extent of sand bodies, connectivity between discrete channel sands, fracture density and orientation in natural exposures of reservoir-equivalent rocks, and stratigraphic architecture at the decametre scale that determines sweep efficiency in waterfloods and polymer floods but cannot be resolved by 3D seismic or well logs alone: The Cardium Formation outcrop in Raven Canyon (northwest of Rocky Mountain House, Alberta) exposes 8 to 14 m of Cardium shoreface sandstone in continuous cliff sections extending 12 km along strike, providing measurements of sand body widths (500 to 2,800 m across depositional strike), shoreface gradient (0.1 to 0.3°), lateral facies variation (shoreface crest, shoreface flank, offshore transition), and fracture density (0.2 to 0.8 fractures per metre in the shoreface crest facies, 0.05 to 0.15 per metre in the offshore facies) that cannot be obtained from subsurface well data alone. Chevron Canada's Cardium 3D seismic reinterpretation program in 2014 used Raven Canyon outcrop measurements to constrain the width distribution of Cardium D3 shoreface bodies in the Pembina area, improving the geological model for infill well placement and waterflood pattern design that was previously based on implicit assumptions about body geometry never tested against outcrop observations. Outcrop analog data is compiled in the WCSB by the Alberta Geological Survey in publication series "Atlas of Devonian and Cretaceous Outcrop Analogs" and by university research groups including the University of Calgary CREWES and NSERC programs that specifically target WCSB-relevant outcrop geometry measurements.
• Production analogs — type curves derived from the historical production performance of existing wells in a geologically similar area — are the primary analog tool for reserves evaluation and economic risk assessment in WCSB tight oil and shale gas plays, with statistical distributions of EUR, initial production rate, and decline parameters compiled from existing well populations providing the quantitative input for Monte Carlo probabilistic reserves estimates required by National Instrument 51-101 (NI 51-101) for public company disclosure: For the WCSB Montney Formation, the AER compiles well-by-well production data in the Alberta Energy Regulator Production Records database, enabling operators and evaluators to extract Montney type curves by township range section (TRS), Montney sub-unit (Montney A, B, C), and well configuration (lateral length, stage count, proppant volume per stage). A typical WCSB Montney type curve for a 2,500 m lateral with 50-stage completions in the Gold Creek Montney A area (northwest Alberta) shows initial rate of 4.5 to 8.5 MMscf/d, year-1 decline of 60 to 75%, and a b-factor of 1.2 to 1.6 in the hyperbolic decline model, with a P50 EUR of 8 to 14 BCF and a P90 to P10 EUR range of 5 to 22 BCF from a population of 180 comparable wells. NI 51-101 requires that the analog population be statistically representative of the proposed well design and geological setting, that the evaluator document the basis for analog selection, and that the EUR distribution be probability-weighted to produce P90, P50, and P10 estimates for the reserves certification — a rigorous process that distinguishes NI 51-101 analog-based reserves from the simpler type-curve extrapolations used in internal economic assessments.
• SAGD (steam-assisted gravity drainage) analogs for new oil sands project evaluations in the Athabasca and Cold Lake areas compare steam-to-oil ratio (SOR), bitumen recovery factor, production ramp-up profiles, and capital cost per barrel per day of capacity across operating SAGD pads with comparable net pay, reservoir quality, and overburden characteristics, enabling project proponents to estimate operating costs and recovery efficiency before drilling sufficient delineation wells to establish project-specific performance data: The Alberta Energy Regulator's SAGD benchmarking database (published annually in the AER's In Situ Performance Presentations for Oilsands) compiles pad-level SOR, production rate, steam injection rate, and recovery factor for all active Athabasca SAGD operations, covering approximately 250 SAGD well pairs at 25 operating pads from Cenovus Foster Creek, CNRL Kirby, MEG Energy Christina Lake, and smaller operators. New SAGD project evaluations use the benchmark database to select the three to five most analogous operating pads (matched on McMurray net pay 15 to 30 m, bitumen viscosity 500,000 to 5,000,000 mPa·s at reservoir temperature, overburden thickness 200 to 400 m, and absence of top water), then use the P50 SOR (typically 3.5 to 5.0 Sm³ steam per m³ bitumen) and P50 recovery factor (40 to 60% OOIP) from the analog pad database to predict the capital efficiency (typically CAD 35,000 to 55,000 per bbl/day of bitumen production capacity) and operating cost (typically CAD 12 to 22 per bbl SOR-dependent) for the proposed project.
• The key limitation of analog-based performance prediction is that analogs describe the outcome of a specific geological setting after the fact (post-drill), while the geoscientist is using them to predict outcomes before the fact (pre-drill), creating a selection bias toward accessible, productive fields where sufficient well data exists to build a type curve, systematically excluding the range of outcomes from failed exploration or sub-commercial development that a true population-based analog should include: The analog selection process in WCSB petroleum exploration tends to systematically draw from prolific fields (Pembina Cardium, Montney Gold Creek, Duvernay Kaybob) where hundreds of wells provide statistically robust type curves, while the undrilled acreage being evaluated may be geologically intermediate between the productive analog field and a less-productive or dry area with few wells and therefore no published type curve. This "oasis fallacy" in analog selection — using only the best-documented successes as comparators while the comparanda may include portions of the geological system that are barren or sub-economic — is a systematic upward bias in mean EUR estimates. Probabilistic analog frameworks that explicitly incorporate the geologically unfavorable analog cases (in addition to the productive ones) and weight them by geological similarity to the undrilled target produce more accurate pre-drill EUR distributions than approaches that use only the most favourable or best-documented analog fields. The SPE PRMS (Petroleum Resources Management System) guidance on analog-based resource assessment explicitly requires that analog selection include both productive and non-productive cases in the comparator set, with the relative proportion of productive to non-productive analogs informing the geological chance of success (COS) assigned to the undrilled prospect.