Kansas Geological Survey, Open-file Report 2012-13
A.P. Byrnes, S. Bhattacharya, and J.R. Daniels
KGS Open File Report 2012-13
Thin (1-10 m thick), heterogeneous, shallow-shelf carbonates of the Arbuckle (Arb), Mississippian (Miss), and Lansing-Kansas City (L-KC) formations in Kansas account for over 73% of the 6.3 BBO cumulative oil produced over the last century. For these reservoirs basic petrophysical properties (e.g., porosity, absolute permeability, capillary pressure, residual oil saturation to waterflood, resistivity, and relative permeability) vary significantly horizontally, vertically, and with scale of measurement. Many of these reservoirs produce from structures of less than 30-60 ft (10-20 m), and exhibit vertical variation in initial saturations and relative permeability properties. Being located in the capillary pressure transition zone, these reservoirs exhibit vertically variable initial saturations and relative permeability properties. Rather than being simpler to model because of their small size, these reservoirs challenge characterization and simulation methodology and illustrate issues that are less apparent in larger reservoirs where transition zone effects are minor and most of the reservoir is at saturations near "irreducible" water saturation. Understanding how capillary pressure properties change with rock lithology and, in turn, within transition zones, how relative permeability and residual oil saturation to waterflood change through the transition zone is critical to successful reservoir management as reservoirs mature and enhanced recovery methods are planned and implemented.
Major aspects of the proposed study involve a series of tasks to measure data to reveal the nature of how wettability, drainage and imbibition oil-water relative permeability, capillary pressure, and electrical properties change with pore architecture and initial water saturation. A second goal is to utilize the data to model shallow shelf carbonate reservoirs and to explore how the properties observed influence reservoir production in transition zone environments. Tasks involved collection of oil and rock samples from carbonate fields around the state (Task 1). Basic properties of the rocks and oils were measured. Comparison was performed between crude and synthetic oil wettability and evaluation made of how wettability is influenced by pore architecture (Task 2). Drainage and imbibition oil-water relative permeabilities were measured on rocks representing the range of porosity, permeability, and lithofacies (Task 3). New petrophysical models were developed and used to construct theoretical reservoir architecture models and geomodels for both analysis of the nature of production in transition zone environments for "type" reservoir architectures and for two reservoirs previously simulated using simpler models (Task 4). Using the theoretical and real geomodels, coring locations in a Lansing-Kansas City and Arbuckle field were selected (task 5). In these fields cores were obtained, analyzed, and evaluated within the context of the geomodels (Task 6). A technology transfer program for data and findings included providing data through a web-based database, publication, and talks given a several professional organization meetings.
Analysis of preserved oil samples collected from 31 wells across western Kansas indicates that oils from the L-KC (n = 34), Miss (n = 53) and Arb (n = 30) average 39 API, 36 API, and 35 API and these western and eastern Kansas oils are statistically similar. Utilizing an empirical relationship developed from previous work oil-water interfacial tension (@60°F) averages 31+1 dyne/cm (error represents 1 standard deviation). Utilizing a selection of these oils with core plugs from the three formations, Amott wettability tests indicate that the Arb and Miss exhibit neutral wettability and that the LKC can be characterized as exhibiting low intermediate oil-wetness. Testing using an isoparaffinic oil indicated similar wettability results.
Porosity and permeability data were compiled from previous work at the Kansas Geological Survey and loaded into a new Rock Catalog database for public access. To supplement these data new cores from wells in the L-KC (n = 7), Miss (n = 2) and Arb (n = 1) were analyzed in addition to the two new cores obtained in the study fields. These data provide the basis for robust permeability-porosity trend relationships. Trends for all three formations exhibit variance in permeability at any given porosity of approximately a factor of 2-2.5 orders of magnitude. They also show that knowledge of lithofacies significantly improves predictive accuracy. Lithofacies-specific porosity-permeability relationships were examined and improved. Of perhaps greater importance, vertical porosity and permeability profiles in the Arb and L-KC wells were analyzed. In the Arbuckle the highly cyclic nature of the peritidal sequences results in similar high frequency cyclicity of high and low porosity and permeability intervals on a scale of 0.1-3 m. In the L-KC the relation between permeability (0.001-400 md) and porosity (0-34%) is significantly influenced by the connectivity of the oomoldic pores complicating the use of porosity as an effective predictor of permeability without information about lithology. The nature of the lithology and the permeability porosity relationship changes vertically through beds as thin as 2-3 m. Of equal importance, work in this study reveals that in the L-KC the Archie cementation exponent (m), used in wireline resistivity log analysis, exhibits significant vertical change over bed thickness (2<m<5). A new relationship between cementation exponent and porosity, parametric in permeability, aids in understanding cementation exponent in oomoldic limestones. This relationship can also be utilized to improve permeability prediction. Cementation exponents for the Arb and Miss equal 2+0.1, in agreement with the commonly utilized value of the standard Archie model.
Air-mercury and air-brine capillary pressure relationships for the L-KC and Miss reveal that for typical reservoir structural closures in Kansas of less than 10-20 m all rocks except the most permeable are in a capillary transition zone over most or all of the reservoir thickness. These relationships indicate that initial oil saturations significantly vertically due to lithology, porosity, permeability, and height above the free water level. Capillary pressure curve models were developed for the L-KC and Miss that provide the ability to estimate capillary pressure properties based on input porosity. Capillary pressures for these carbonates can be modeled using modified Brooks-Corey equations where the threshold entry pressure and pore size heterogeneity dimension can be predicted using permeability. These models can be utilized to estimate vertical water saturation variation and aid in geomodel construction.
Drainage and imbibition oil-water relative permeability measurements were performed on L-KC oomoldic limestones and Miss moldic porosity mudstone to grainstone lime-dolomites. For these rocks, residual oil to waterflood (Sorw) increases with increasing initial oil saturation (Soi) for a given rock type due to enhanced trapping by emplacement of oil in fine pores. The Land (1968) equation trapping characteristic, C, increases with increasing porosity resulting in less trapping with increasing porosity. This relationship, coupled with increasing "irreducible" water saturation (Swi) with decreasing porosity and permeability, results in a systematic change in Sorw with porosity/permeability and Soi. With Soi decreasing with depth in the transition zone, proper modeling of kr in the transition zone requires a family of relative permeability (kr) curves that reflect changes in kr with changing Soi. Utilizing a family of kr curves in reservoir simulation shows that both oil and water recovery are greater than predicted from models utilizing kr curves with a constant Soi and Sorw. Oil recovery is higher because Sorw(Soi) is lower and water recovery is higher because water saturation (Sw) increases with proximity to the oil-water contact. These results validate and expand on use of the Land equation in shallow-shelf carbonates and help to explain both the high oil recovery and the high water production rates that are often evident in these reservoir systems. Oil relative permeability varies with k and Sorw(Soi). Comparison of models utilizing Sorw(Soi) with models utilizing a constant Sorw values for the entire reservoir indicates that models using Sorw(Soi) predict more oil and more water production.
Comparison of numerical flow simulation of reservoirs modeled using fewer vertical cells and simpler relationships with the same reservoirs modeled using the relationships developed in this study indicates that reservoir performance prediction differs between the approaches and that models using the relationships developed in this study may improve reservoir performance prediction and management.
KGS_OF_2012-13.pdf (51.6 MB)
To read this file, you will need the Acrobat PDF Reader, available free from Adobe.
Kansas Geological Survey, Energy Research
Updated Aug. 9, 2012
Comments to email@example.com
The URL for this page is http://www.kgs.ku.edu/Publications/OFR/2012/OFR12_13/index.html