Energy Policy Act of 2005 (Ultra-deepwater and Unconventional Resources Program)
Reservoir Connectivity and Stimulated Gas Flow in Tight Sands
The goal of this project is to develop new concepts for predicting the location of high saturation/high permeability sweet spots and optimizing well completion and stimulation procedures in tight sand reservoirs, by carrying out a multi-disciplinary analysis of data from the Mamm Creek field in the Piceance Basin of Colorado and to conduct a regional analysis of the entire Piceance basin to allow extrapolation of key results (such as parameters that control sweet spots) to exploration targets elsewhere in the basin. Specific project activities in pursuit of this goal include reservoir modeling, rock mechanics modeling, seismic analysis, microseismic studies of fracture propagation, and regional stratigraphic and structural mapping. Based on the results of this study of Mamm Creek and some aspects of neighboring fields, the project team hopes to add a second phase to the project, which would be a full-scale fracture experiment to test the validity of the model predictions.
Colorado School of Mines (CSM), Golden, CO 80401
University of Colorado – Boulder, Boulder, CO 80309
Mesa State College, Grand Junction, CO 81501
iReservoir, Inc., Littleton, CO 80120
Natural gas produced from tight sandstones has become a significant contributor to the nation’s energy supply over the past twenty years. Unfortunately, predicting the distribution of gas saturation and natural fractures within tight sandstone formations remains difficult, and selecting the best well completion and stimulation strategies for a given set of tight gas reservoir conditions can be a hit-or-miss prospect in many cases. This project is developing concepts and methods that can be used by the E&P industry to better predict natural gas sweet spots and optimize well completion and stimulation strategies in tight gas reservoirs.
The project is carried out in a multi-disciplinary way and is focusing on geologic, geophysical, and production data from the Mamm Creek field of the Piceance Basin, Colorado. It is also addressing stratigraphic and structural issues across the entire Piceance basin. Well and seismic data for the project were provided by Bill Barrett Corporation and Whiting Petroleum.
The project is performed by a diverse team of engineers, geophysicists, and geologists from CSM, University of Colorado, Mesa State College, and iReservoir. The Colorado Energy Research Institute at CSM is overseeing and coordinating the entire project, and students and post-docs carry out many of the project’s technical tasks.
The deliverables for the project include: 1) static reservoir models for key producing zones of the Mamm Creek field; 2) cross-sections correlating outcrops, well logs, and regional seismic data; 3) maps of fracture distribution and an explanation of the tectonic context of the fractures; 4) reservoir simulation models based on the static geological model and performance forecasts; 5) a predictive model for static crack propagation; 6) a draft model showing the interplay of structure, stratigraphy, fractures, and fluid flow in the Mamm Creek field; and 7) annual reports and a final report including all results and outputs.
This project is likely to result in new concepts and methodologies for predicting natural gas sweet spots in tight sandstones similar to those found in the Mamm Creek field. The results should enhance the ability of E&P companies to predict the location of higher permeability zones with significant gas content in tight sandstones throughout the Piceance Basin and perhaps beyond. The project should also result in recommended completion and stimulation strategies to improve gas production. This could raise per-well efficiency, lower overall development costs, and increase recovery of natural gas from tight reservoirs.
In the near term, improved sweet spot prediction for the Piceance Basin has the potential to accelerate the addition of some portion of the basin’s estimated resource of 24 Tcf recoverable gas to the nation’s domestic gas supply. The cumulative, long-term benefit will depend on the number of wells that are drilled, completed, and stimulated using concepts and tools resulting from this project. An incremental increase in domestic gas production would result in increased tax revenues, royalties, and regional economic benefits.
Technology transfer. The team has built a partnership of 10 companies that are now participating in this project (5 more than at the time of proposal submittal). They are sharing an increasing amount of data from throughout the Piceance basin.
Development of static reservoir models. CSM, University of Colorado, and Mesa State College are utilizing well logs, outcrop data, LiDAR data, and high-resolution orthophoto data to develop static reservoir models for key producing zones in the Mamm Creek field. Work is well underway with the reservoir models, using data obtained from nearly 400 wells.
Analysis of basin stratigraphy and structure. This task is progressing very well, and is being greatly aided by the support from the 10 operators in the basin who have generously shared data, expertise and cash support above and beyond the Federal funding. The researchers have assembled a regional well log data base of 3300 logs across the basin.
Fracturing of rocks in theory and in the field. Progress to date has been significant. To this point, the studies have focused on two areas: determining the T-stress parameter in a common rock fracture specimen and the development of a discrete element modeling code.
Task 1 – Project Management Plan. The Project Management Plan consisting of a work breakdown structure and supporting narrative that addresses the overall project as set forth in the agreement has been completed.
Task 2 – Technology Status Assessment. A Technology Status Assessment report describing the state-of-the-art of the proposed technology has been submitted.
Task 3 – Technology transfer. A partnership of 10 companies are sharing an increasing amount of data from throughout the Piceance basin. There have already been several meetings with each company individually and three group meetings. Another group meeting for all of the 10 members of the industry consortium will be held on May 19th, 2009 in Denver.
Task 4 – Development of static reservoir models. CSM, University of Colorado, and Mesa State College are utilizing well logs, outcrop data, LiDAR data, and high-resolution orthophoto data to develop static reservoir models for key producing zones in the Mamm Creek field. Resulting 3D models will be generated in Petrel and will include lithology, structure, porosity, and permeability of the chosen intervals. Rachel Shaak, Alicia Hewlett and Sait Baytok (students of Matt Pratner at CU Boulder) are already well underway with the reservoir model, with data obtained from nearly 400 wells in the Mamm Creek field. For added information on the sedimentary facies of the reservoir sand bodies they are incorporating outcrop data previously collected by Bruce Collins along the Grand Hogback a few miles to the east of the field. They have also started building the Petrel reservoir model from the 3D seismic data volume, and will develop volume models in the time domain, depth domain and by principal components. In this framework, they will refine the geometry of the gas bearing reservoir sand bodies using object-based modeling and paleocurrents from image logs. Permeability and porosity will be distributed in accordance with the model objects. This task is progressing well, and it is important that the resulting geostatic model be essentially completed before start of the dynamic modeling (fluid flow simulation).
Task 5 – Analysis of basin stratigraphy and structure. This task is progressing very well, and is greatly aided by the support from the 10 operators in the basin who have generously shared data, expertise and cash support above and beyond the RPSEA funding. The main objective of this task is to generate a regional framework to predict and extrapolate key findings in the Mamm Creek field (e.g. what makes a sweet spot) – and elsewhere in the densely drilled Colorado River corridor – to other more distal and underdeveloped parts of the basin. Comprehensive and reliable stratigraphic and structural frameworks are the key to such predictability. We have assembled a regional well log data base of 3300 logs across the basin (more will be added. Chris Schwendeman, Joe Nicolett, Renee Foster and Mike Liebovitz (students of Paul Weimer at CU-Boulder) are building the stratigraphic framework of the bulk of the Cretaceous strata across the basin layer by layer. Chris Schwendeman is reconstructing the stratigraphy of the lowermost units of interest – the top of the Dakota Sandstone to the top of the Isles Formation. Even though the total number of well logs in the basin is great, only 74 wells penetrate deep enough to reach the top Dakota. Chris has been able to map out the marine sandstones tongues of the Isles Formation in the Piceance basin. Joe Nicolett (yet to start) will be responsible for the next stratigraphic interval – from the Cameo Coal (base of Williams Fork Formation) to the top of the informally named “Big Kahuna.” Renee Foster is building the detailed correlations in the I-70 corridor area, and will tie this to detailed outcrop work along the western margin of the basin. This outcrop work has been already completed as part of an earlier outcrop-focused project along the western basin margin. The informally named “Big Kahuna” (very thick sandstone) is a very important reservoir unit, accounting for 40% of the cumulative production in this basin to date.
Yet, there is still more production in the cumulative thinner-bedded units distributed throughout the rest of the Williams Fork Formation. It is still an unsettled question whether the “Big Kahuna” is a major incised valley fill or a multistory aggradational channel (this distinction has production implications). Mike Leibovitz is responsible for the stratigraphy of the uppermost productive interval, from the top of the “Big Kahuna” to the Ohio Creek Formation, which includes the zone of the Price Coal and the adjacent “Upper Kahuna.” Some wells are completed and producing all the way up into the Ohio Creek Formation. Completion of the stratigraphic framework will have a significant impact on our understanding of gas charge distribution and basin-wide gas migration pathways.
Bruce Trudgill at CSM is conducting a structural analysis of the basin that will build on the stratigraphic framework outlined above and apply some of the same data. It is clearly most useful when tied to the most significant stratigraphic horizons. In addition, potential field data will also be brought into the structural analysis. A key structural finding for parts of the Piceance basin (at Rulison field) is the vertical change from extensional structures at the Dakota level changing upward to compressional structures at the top of the Williams Fork. Such structural ‘inversion’ can be explained by transpressional and transtensional stress along strike slip faults, but work to establish the definite mechanism will continue. The second aspect of this structural study addresses directly the issue of predicting fracture density and orientations, applying Badley’s “Trap Tester” software. The work now moves into the phase of testing the predictions of these models and relating them to observed spacing, orientation and density of fractures in the producing fields.
Task 6 – Reservoir flow simulation tied to rock properties. iReservoir.com is conducting the research related to task 6. The project team will create a flow simulation model to examine the effect of fractures on fluid flow in the Mamm Creek field. This task also includes seismic anisotropy analysis by iReservoir to identify potential bypass zones. This task has not yet started, awaiting the completion of v. 1.0 of the geostatic reservoir model (Task 4).
Task 7 – Fracturing of rocks in theory and in the field. This task is conducted by John Berger, Graham Mustoe and Paul Martin (CSM) and Scott Buechler (student at CSM). Progress to date has been significant. To this point, the studies have focused on two areas: determining the T-stress parameter in a common rock fracture specimen and the development of a discrete element modeling code. The T-stress in the local stress-field around a crack tip (and potentially other higher-order stress field terms as well) can have substantial effects on the fracture of rock. The next step will to apply this technique on Brazilian disk test specimens - Brazilian disks are widely used in fracture studies of rock.
The team is also developing a complete bonded code capable of producing macroscopic properties (Young’s modulus and Poisson’s ratio) using local particle stiffness parameters, and will drive this project towards incorporating the results into a hydro-mechanical fracture model for tight gas sandstones.
Task 8 – VSP analysis of reservoir attributes. This task was originally referred to as “building on RCP’s earlier work in the Piceance basin.” Tom Davis and his colleagues in the Reservoir Characterization Project at CSM conducted an earlier study of the Rulison field, and collected a shear wave VSP (vertical seismic profiling) data set as part of that project. Such data sets are very rare. In this project, Davis’ student Praj Mazumdar is developing a powerful new approach to the analysis of such data, and applying this to the task of seismically defining sand body attributes in the Williams Fork Formation. Shear waves give better impedance contrasts along shale/sandstone boundaries than do P waves, and the shorter wave path (and lack of near-surface interference) associated with down-hole placement of geophones provide higher fidelity in the signal. For example, the data reveal small faults that are not detectable with surface data acquisition.
Task 9 – Micro-earthquake location by wave-interferometric imaging. This task aims to study the feasibility of using interferometric wave-equation imaging to examine fluid movement and fracture growth during hydro-fracturing operations. Paul Sava and Roel Snider at CSM are the co-principal investigators, and a student will come onboard this summer. By using a “Bayesian” probabilistic approach to the data analysis, they expect to more realistically overcome the many issues related to signal-to-noise ratio during acquisition of the very faint and intermittent micro-earthquake signals. The next step is to dig more deeply into the physics of the phenomena.
Task 10 – Electrical methods for monitoring hydrofracturing. CSM will integrate theoretical, numerical, and experimental approaches to evaluate the self-potential (SP) method for monitoring fluid flow during hydro-fracturing. Andre Revil at CSM is the principal investigator on this project. The tasks underway include: 1) coupling of streaming potential and multiphase flow within a new simulator, including hydromechanical aspects of the problem, 2) development of a framework of inverting borehole and surface streaming potential data sets to locate the position of water fronts and fracturing events (this also will be probabilistic, or Bayesian, in nature), and 3) development of a controlled sandbox experiment to prove the validity of this approach. This project task is in its infancy.
Task 11 – Azimuthal AVO and attenuation analysis for fracture characterization. This CSM project under the leadership of Ilya Tsvankin (student coming onboard this summer), is predicated on the observation that the presence of aligned fractures makes a medium (a rock body) anisotropic on the scale of seismic wavelengths. As we have known for some time, the anisotropy measures both the effects of fractures and pores, and it also appears to provide direct detection of the top of continuous gas; a potentially very valuable application that deserves to be addressed further. There is also a question whether we can invert AVO data into fracture density as well as orientation, which might potentially generate additional applications for this exploration tool.
Project Start: September 19, 2008
Project End: September 18, 2010
DOE Contribution: $2,894,256
Performer Contribution: $4,634,000
RPSEA – Kent Perry (email@example.com or 847-768-0961)
NETL – Gary Covatch (Gary.Covatch@netl.doe.gov or 304-285-4589)
CSM – Dag Nummedal (firstname.lastname@example.org or 303-273-2506)