Geologic model for the giant Hugoton and Panoma Fields Martin K. Dubois
by user
Comments
Transcript
Geologic model for the giant Hugoton and Panoma Fields Martin K. Dubois
Geologic model for the giant Hugoton and Panoma Fields Martin K. Dubois Alan P. Byrnes Geoffrey C. Bohling Midcontinent AAPG, Oklahoma City September 13, 2005 upscale porosity Objectives of modeling project Objective: Build 3D cellular model populated with lithofacies and petrophysical properties Purpose: 1. 2. Identify and quantify remaining gas in order to develop best field practices for efficient recovery. Study sedimentary response to rapid glacio-eustatic sea level fluctuations on an extremely gently sloped ramp (shelf). More specifically, and in conjunction with simulations studies Estimate original gas in place at well, region and field scales Reservoir connectivity at pore, flow unit, well, inter-well, region and field scales Differential depletion in stratigraphically separate reservoirs Production decline rates and EUR at ultra low pressures Status and outline Modeling project status: To be covered today: 9 Model workflow 9 Major lithofacies and depositional model 9 Large scale geometry of Hugoton and Panoma 9 Lithofacies in maps and cross sections Township scale models have been built and tested by numerical simulation Components are in place for building field-wide cellular model and work is underway (Field 3D model not yet complete but plenty to see) Thinly layered, alternating carbonate and siltstone reservoir in 13 marine-nonmarine sedimentary cycles Herrington Krider Winfield Towanda 550 ft (Hugoton) (Panoma) Chase Group Council Grove Gp. Wolfcampian 130 Miles Hugoton and Panoma Stratigraphy L. Permian Ft Riley Florence Wreford Funston Crouse Middleburg Eiss Morrill Cottonwood Neva Geomodel Workflow (static model) Gather data CORE & ELog Var. Neural Net NODE WELLS Stochastic Methods 3D MODEL 1400 “Node” Wells Train Neural network and predict lithofacies in non cored wells (nodes) Lithofacies in core tied to log and geologic constraining variables Fill volume between node wells using stochastic methods Develop dynamic model through empirical relationships Empirical Relations & Free Water Level Permeability, Water sat., Rel. Perm. Differential Pressure, Corrected MatBal OGIP Dynamic Model & Simulation Relative Permeability (fraction) Model, Facies, Phicorr 1 0.1 0.01 w-10 md g-10 md w-1 md g-1 md g-0.1 md w-0.1 md g-0.01 md w-0.01 md g-0.001 md w-0.001 md 0.001 0.0001 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Dubois, Byrnes etal, 2003 Water Saturation (fraction) Lithofacies from Core to “Node” Wells 80 mi 130 km Current training set Other Some wells have both Chase and Council Grove core 130 mi 210 km 27 mi 43 km Training set for neural network lithofacies prediction Well 1/2 foot count intervals 8 3952 Chase 10 4593 Council Grove 8545 ½-foot intervals with lithofacies tied to log and core properties Lithofacies predicted at 1369 “node wells” Neural Network Training and Predictions Chase All 1369 Wells Predicted Sandstone 8% 4% 6% 2% Coarse Silt 28% 23% 23% 20% Fine Silt 24% 4% 13% 8% Siltstone 9% 7% 8% 10% Carb Mdst 7% 5% 5% 4% Wackestone 18% 13% 14% 19% Fxln Dol. 4% 2% 3% 4% Packstone 15% 17% 15% 23% Grainstone 4% 1% 2% 0% M-Cxln Dol. 0%* 12% 6% 4% Sandstone 0%** 12% 5% 6% Predicted Council Grove Training Tr ain Pr ed ic Facies Continental Distribution of eleven lithofacies in training set t Lithofacies in training set and predicted in wells 42% 30% Marine 27% 33% 31% 37% * Insufficient training sample. Combined with Fxln Dolomite ** Insufficient training sample. Combined with Siltstone. Distribution of lithofacies predicted in 1369 wells is similar to that in training set. M-Cxln MM SS 5% Dol 6% Cont SS 6% Grnst 2% Cont Crs Slt 23% Pkst 15% Fxln Dol 3% Cont Fn Slt 13% Wkst 14% Mdst 5% Mar Slt 8% Phyloid Algal Bafflestone Close-up Core Slab Nonmarine Shaly Siltstone 0.5 mm 4.6% 0.000024 md Nonmarine Crs Siltstone-vfg SS Cm Core Slab Thin Section Photomicrograph T h i n S e c t i o n Photomicrograph L8 Core Slab 20.6% 1141 md Dolomite 13.9% 1.1 md C l o s e -u p C o r e Slab Close-up Core Slab 10.8% 0.30 md Thin Section Photomicrograph Council Grove Lithofacies Cm L6 L1-2 Pellet Grainstone 13.0% 2.53 md 0.5 mm L0 0.5 mm Core Slab 1 ’s 0 l ain l a id M K (1 t la F M ’s 0 ) s ile T astlP n o C ago L lM ound ga yl.A h P e a Id liz d e e D o p M s lo e d ito n la o (M d a ife s rR fte ,I is o rv n .) c Marine Siltstone Thin Section Photomicrograph 10.4 % 0.01 md Thin Section Photomicrograph 0.5 mm L7 L8 0.5 mm L3 (time slice) 0.5 mm Silty Wackestone Thin Section Photomicrograph Thin Section Photomicrograph M-CG Oncoid-Peloid Packstone 21.2% 32.3 md Close-up Core Slab Cm L5 Skeletal Wackestone 2.1% 0.11 md 0.5 mm Close-up Core Slab Thin Section Photomicrograph L4 3.4% 0.0024 md 0.5 mm Core Slab Unique Chase Lithofacies Two additional lithofacies plus same nine as in Council Grove but in different proportions. No phylloid algal facies. Close-up Core Slab 2 cm Close-up Core Slab Crs XLN Dolomite (CG oo-grnst) 22.3% 275 md Marginal Marine FG Sandstone 20.8% 48.2 md Thin Section Thin Section (time slice) Dolomitized medium to coarse-grained ooid and bioclastic grainstone are the dominant reservoir facies in Chase Present Day Structure 100 mi (160 km) Panoma KE YE SD OM E Hugoton Reservoirs of Hugoton and Panoma Fields were deposited on a very gently dipping shelf. Relief was much less than it is today. Top Council Grove Shelf Margin Chase Base of Council Grove VE = 200X VE = 100X Chase and Council Grove Core facies 700 ft 215 m Council Grove Chase Field Margin Continental VE=700X Marine Silt Grain support carb. Sand Mud supported and silt 85 miles 135 km Shelf Margin 9 Carbonate thins toward updip field margin 9 Redbeds thin basinward 9 Eolian sands at west margin 9 Council Grove thinnest at midshelf Gross interval 0’ 2 2 Net “Continental” Net Marine 70 ’ 15 0’ Similar sedimentation patterns in Chase and Council Grove 0’ 7 1 37 0’ ’ 40 0’ 22 ’ 30 25 0’ Series of slides based on facies predicted by Nnet in 1369 wells 0’ 1 2 Th Mi inn ds es he t lf Council Grove (thru B5_LM) 30 40 ’ 0’ Chase Mean Lithofacies in Marine Intervals Entire Chase Entire Chase 8 Color Bar Scale 4 Chase to Ft Riley Council Grove to C_SH 8 8 4 4 6.5 4.5 Facies 3-10 Mean F = 6.7 SD = 0.9 F10 dominates west margin 5.8 6.7 Facies 3-9 Mean = 5.8 SD = 0.6 Marine 3 Siltstone 4 Carb Mdst 5 Wackestone 6 Fxln Dol. 7 Packstone 8 Grainstone 9 M-Cxln Dol. 10 Sandstone Facies 3-9 Mean = 5.8 SD = 0.9 F9 dominates south Facies 3-9 Mean = 5.4 SD = 0.4 F6 dominates to NE F7 dominates to SE Continental 0 Sandstone 1 Coarse Silt 2 Fine Silt Shown are the mean code value for lithofacies predicted by neural network models in 1350 wells Main “Pay” Lithofacies in Chase (F7-9) Herrington Krider Winfield Towanda Krider only PhiH for F9 (Herrington through Gage) 200 0.8 0 0 Phi x H for Facies 9 Cutoff phi >15% Net thickness Facies 7 thru 9 Net / Gross Facies 7 thru 9 Accumulation of coarsegrained bioclastic-ooid sand associated with bathymetry of embayment near the shelf margin Krider Ooid shoal facies in Stevens County A A’ 10 foot divisions Core A 2 cm A’ Close-up Core Slab Crs XLN Dolomite (CG oo-grnst) 22.3% 275 md Thin Section 1 Coarse Silt 3 Siltstone 4-5 Mdst-Wackestone 7 Pack-Grainstone 9 M-Cxln Dol. 10 Sandstone Cottonwood (B5_LM) Phylloid Algal Mounds Net H, F7-8, Phi >10% Phyloid Algal Bafflestone 20 A 0 L8 Core Slab 20.6% 1141 md 0 1 Sandstone Coarse Silt 2 3 Fine Silt Siltstone 4-5 Mdst-Wackestone 6 Fxln Dol. 7-8 Pack-Grainstone A’ A Core A’ Crouse (B1_LM) fine-crystalline dolomite lithofacies F6-8, phi > 8%, Net/Gross 0.8 0 1 Sandstone Coarse Silt 2 3 Fine Silt Siltstone 4-5 Mdst-Wackestone A 6 0 Fxln Dol. 7-8 Pack-Grainstone B A Core 0.5 mm B Thin Section Photomicrograph Dolomite 13.9% 1.1 md C l o s e -u p C o r e Slab Cm L6 Core Neva (C_LM) 0 1 Sandstone Coarse Silt 2 3 Fine Silt Siltstone & Sandstone 4-5 Mdst-Wackestone 6 Top Council Grove Neva Fxln Dol. 7-8 Pack-Grainstone Net thickness, phi >15% Fine-grained sandstone in lower Council Grove is pay in Texas County Eolian sandstone Council Grove Continental sandstone thickness 120 Cum. Prod. 1.5 BCF 20 Dubois and Goldstein, 2005 Summary Township scale models have been built and tested by numerical simulation Components are in place for building field-wide cellular model (underway) Neural network models are proving effective in facies predictions and building an accurate geomodel We anticipate being able to successfully delineate remaining gas in place in the Hugoton and Panoma Fields Acknowledgements We thank our industry partners for their support of the Hugoton Asset Management Project and their permission to share the results of the study. Anadarko Petroleum Corporation BP America Production Company Cimarex Energy Co. ConocoPhillips Company E.O.G. Resources Inc. Medicine Bow Energy Corporation Osborn Heirs Company OXY USA, Inc. Pioneer Natural Resources USA, Inc. also geoPlus (Petra) and Schlumberger (Petrel)