An AVO and Seismic Attribute Analysis of The

University of Nevada, Reno

An AVO and seismic attribute analysis of the

SouthernSan Emidio Geothermal System,

Northwestern Nevada

A thesis submitted in partial fulfillment of the

requirements for the degree of Master of Science in Geology

by

Joseph Dierkhising

Dr. John Louie / Thesis Advisor

Dr. Graham Kent / Thesis Advisor

May, 2015

Abstract

Geothermal fields in the Basin and Range province are often found in highly faulted and fractured subsurface zones that provide the necessary permeability for vertical fluid flow. Frequently, subsurface structure and petrophysical properties vary substantially from what is predicted from surficial mapping alone. Advanced seismic reflection techniques, including amplitude versus offset (AVO) and seismic attributes, can yield the additional information needed to characterize known geothermal resources and explore for currently undiscovered resources. This study utilizes 2D reflection seismic data to explore the geophysical signature of the San Emidio geothermal field, located in NW Nevada. A strong correlation is found between the highly faulted geothermal field in the San Emidio basin and the geophysical response observed in P- and S- wave derived attribute sections. A subvertical region of high Poison’s ratio, associated with less consolidated softer material, as well as locally low P- wave velocities, associated with higher porosity zones, is found in the vicinity of the geothermal well where known high permeability faults have been detected. The energy amplitude versus offset response along the 2D seismic line shows low AVO-gradient values where faults are interpreted in the seismic data and surficial fault scarps are extended into the subsurface. The development and application of these techniques to geothermal resources in the Basin and Range provide opportunities to identify new subsurface clean energy resources.

Acknowledgements

Include in final draft…

• John Louie, Satish, Optim, Graham, Kyle Reeder, Kyle Gray, Andrew Sadowski, Ken Hoffman and Jasbinsek
• Fellow grad students, department members
• Family and non-academic friends

John Louie was instrumental in assisting me with the functionality and use of his codes, scripts and Viewmat processing software. John also provided invaluable inspiration, guidance and support during the course of this project. SatishPullammanappallil provided much assistance producing the velocity models, obtaining data, and understanding key project details. Funding of this research by Optim Inc. was crucial in completing this project. Graham Kent provided valued and necessary input on seismic processing and interpretation. My fellow office mate and masters student Kyle Reeder, also working on another aspect of this project, provided critical feedback and stimulating discussions on a daily basis.

Table of Contents

• 1.0: Introduction (Thesis)
• 2.0: Paper

-2.1: Introduction

-2.2: Geologic and Tectonic Setting

-2.3: Methods

-2.4: Results and Interpretation

-2.5: Discussion

-2.8: Acknowledgements for Research Article

• 3.0: Conclusions (Thesis)

-3.1: Future work

- 3.1.1:Look at vibroseis correlations

- 3.1.2: Collect better data for pre-critical AVO work

- 3.1.3: Collect and integrate well data

- 3.1.4: Are there any other future student projects left?

- 3.1.5: What could these future projects consist of?

• 4.0: References
• 5.0: Appendix

-5.1: San Emidio seismic acquisition details

-5.2: Updated down-going p-wave Zoeppritz Matrix

-5.3: AVO Modeling Matlab script

-5.4: Empirical AVO Matlab script

-5.5: Functions listed

List of Figures

Figure one: Satellite image of the San Emidio geothermal field including the surficial alterations, range names, well locations, power plant location, seismic line location and wind mountain epithermal deposit mine location.

Figure two:Image of the Vp and Vs models.

Figure three: Image of the Vp/Vs and Poisson’s ratio sections.

Figure four: Image of the attribute sections, AVO-I, AVO-G and RpRs.

Figure five: Kirchhoff pre-stack depth migrated image of the seismic line with the production well location, interpreted pyramid sequence horizon, and interpreted faults overlain.

Figure six: AVO RMS-Energy sections.

Figure seven: Theoretical AVO response out to the critical angle using the full Zoepppritz equations, Frasier and Richards approximation, and the Shuey approximation. The horizontal axis is labeled with both angle and offset distance.

Figure eight:Theoretical I/G crossplot including our best estimate and the I/G variability due to faulting.

Figure nine: A crossplot indicating the change in parameters influence over I/G pair crossplot location.

Table one: The amplitude preserving processing schemeused in this study.

Introduction (Thesis)

This project began as an effort to expand geothermal resource exploration efforts to increase energy production at the San Emidio geothermal field. In 2011 the current field lease owners and power plant operators, US Geothermal Inc., were awarded funds to explore the ability of advanced geologic and geophysical methods to detect geothermal resources in the San Emidio basin. Funding for this project was made possible by an award from theAmerican Recovery and Reinvestment Act (ARRA) for the U.S. Department of Energy Validation of Innovative Exploration Technologies in the Geothermal Technologies Program.

The geophysical exploration and software company Optim Inc. was awarded a contract to lead the advanced seismic methods portion of the project. 10 approximately 3.2 km 2d reflection seismic lines were collected along the southwestern edge of the San Emidio basin where the known geothermal resource exists. 5 lines were collected near the existing geothermal field and 5 lines were collected to the north where a right step in the range front and an epithermal mineral deposit hint at the existence of undevelopedgeothermal resources.

Optim, Inc conducted the initial seismic processing needed to produce high quality p- and s- wave velocity models used for interpretation and to produce the pre-stack depth migrated images of the subsurface.

In 2011 surficial geologic mapping along with the pre-stack depth migrated images were used by University of Nevada Reno masters student, Gregory Rhodes, to make structural interpretations of the San Emidio geothermal system.

This project continues the geothermal exploration and reservoir characterization efforts in the San Emidio geothermal system. High resolution compressional and shear wave velocity models were computed by SatishPullammanappallil of Optim Inc., and given to me as ASCII text files.

I then modified a C Shell script written by John Louie, of the Nevada Seismological Laboratory and the University of Nevada Reno, to convert the ASCII files into smoothed Intel binary files required for migration. Next, I produced travel time plots by modifying a travel time generating script written by John Louie. Pre-stack depth migrations were produced using a C code, written by John Louie. I specified the migration parameters used by the migration code by modifying a parameter file accessed by the C code.

Processing of the original shot gathers used in the migrations, and further processing of the migrated common image gathers, was done by me using John Louie’s Viewmat software.

I also wrote a library of Matlab scripts and functions to: (1) take information about rock properties at San Emidio to produce a model of the theoretical seismic response at the geothermal reservoir, (2) take the p- and s- wave velocity models and produce various seismic attribute sections, and (3) cross plot empirical data.

John Louie was instrumental in assisting me with the functionality and use of his codes, scripts and Viewmat processing software. SatishPullammanappallil provided much assistance producing the velocity models, obtaining data, and understanding key project details. Graham Kent provided valued and necessary input on seismic processing and interpretation.

This paper will be published in _____ under the title, “____”.

BEGIN PAPER

The following chapter is my submission to ___ publication. I am first author on this paper. My co-authors are: John Louie, and Graham Kent of the Nevada Seismological Laboratory, University of Nevada, Reno, Nevada and SatishPullammanappallilof Optim Inc., Reno, Nevada.

Introduction

The San Emidio Basin is located in the northwestern Basin and Range province, immediately northeast of the better known Pyramid Lake Basin and approximately 105 km northeast of Reno, Nevada.The San Emidio geothermal system is currently producing approximately 9.0 MW from 4 (maybe 5 now? Who can I ask?)production wells located in a highly faulted zone in the southeastern edge of the San Emidio Basin, currently under lease by US Geothermal Inc. ( US Geothermal plans to expand production at San Emidio by drilling additional production wells (Is this still the case?). Resources to explore and develop the geothermal system in this area remain limited. Consequently, an understanding of the geothermal system, reservoir geology and existing data is critical to optimize the development of this field and minimize future costs.

The current San Emidio power plant is the second power plant to occupy the San Emidio basin. The first one was a 3.6 MW power plant known as the Empire Geothermal Plant ( In 2012 the current power plant went online with a design capacity of 11.8 MW ( Future development at San Emidio is intended to reach the full capacity of the existing power plant (Reference? USG?) (Is this paragraph even necessary?).

The current production zone is located within a highly faulted zone in the southern portion of the San Emidio geothermal resource area(Teplow et al., 2011). To the north of the proven geothermal resource is a hard-linked right step in the range front (Rhodes, 2011). In the vicinity of this right step there are indications of a potential high-grade geothermal resource including a high density of faulting and epithermal mineral deposits (Rhodes, 2011). Careful characterization of the known reservoir in the southern part of this system may yield clues to identifying high quality drill targets to the north.

Faults and fractures oriented approximately orthogonal to the least principle stress direction are favorably oriented for subvertical fluid flow in highly permeable fault zones (Barton et al., 1995; Faulds et al., 2006). It is believed that favorably oriented faults and fractures in areas of high fault and fracture density likely produce the permeability required for deep fluid circulation within the San Emidio geothermal system (Rhodes, 2011).

Normal faults, favorably oriented for subvertical fluid flow, that have a fault width of greater than 15 cm are considered to be large aperture faults (LAFs) in the San Emidio geothermal system (Teplow et al., 2011, Ferrill, D.A., Morris, A.P., 2003). Production wells at San Emidio have encountered LAFs in the proven geothermal reservoir(Teplow et al., 2011). For this reason, accurately identifying LAFs at depth is a crucial step to identifying new geothermal resources in this basin and many other basins in this region.

Until recently, amplitude versus offset (AVO) methods have been used primarily to explore and characterize gas accumulations in clastic reservoirs (Rutherford, S.R., Williams, R.H., 1989). The development of a comprehensive AVO classification schemeprovides the framework to characterize AVO response in geothermal reservoirs, independent of hydrocarbon specific responses (Young, R.A., LoPiccolo, R.D., 2003).

Pioneering studies, using AVO methods to characterize geothermal resources in faulted hard rock reservoirs,have demonstrated the utility of AVO in seismic studies (Cameli et al, 2000). An increase in fracture density is observed to be associated with a decrease in formation density and seismic p-wave velocity (Cameli et al, 2000). Traditional AVO methods, designed for clastic hydrocarbon systems, are often inappropriate for hard rock geothermal systems and subsequently new AVO methods must be developed. A new AVO-attribute has been tested and shown to be effective in detectingfractured geothermal reservoirs in intrusive basement rock, common in many geothermal systems (Aleardi, M., and Mazzotti, A., 2014 (Personal communication until published)).

The availability of long offset (~3.3 km) 2D reflection seismic data, at the producing San Emidio reservoir, provides the unique opportunity to observe and model the seismic response of such systems in the highly faulted, extensional basins of the northwestern Basin and Range province. Well constraints on the depth and location of the known reservoir will be used to observe and characterize the AVO response with the intent that these results will help provide crucial information about geothermal prospects to the north.

Geologic and Tectonic Setting

The San Emidio basin is formed between the Lake Range, to the east, and the Fox Range, to the west. The eastern San Emidio basin, where the known geothermal resource exists, consists of a Mesozoic basement, Tertiary volcanic and sedimentary rocks, and Quaternary alluvium, lacustrine sediments and hydrothermally altered rocks (Rhodes, 2011). The Mesozoic sequence, collectively referred to as the Nightingale formation and exposed along the western side of the Lake Range, consists of metamorphosed and folded low-grade argillaceous phyllite, with some slate, schist and interbedded carbonate, sandy and volcanic horizons (Moore, 1979; Wood, 1990). The overlain Tertiary volcanic and sedimentary sequence consists of the middle Miocene Pyramid sequence volcanic rocks and late Miocene sedimentary rocks correlated to the Truckee Formation (Drakos, 2007; Moore, 1979). Quaternary sediments, including alluvial fan deposits and Pleistocene Lake Lahontan silt, sands, tufa, and silicified sands are observed at the surface along the western edge of the Lake Range (Teplow et al., 2011).

2.2.1: Structural framework

The San Emidio fault zone consists of a network of north and north-northeast striking normal faultsthat dip to the west and west-northwest respectively. The western flank of the Lake Range is bound by the north striking, west dipping Lake Range normal fault.The Lake Range consists of a series of well exposed, primarily east-tilted, north-trending fault blocks bounded on the west by the moderately to steeply west-dipping Lake Range fault. To the west, on the hanging wall of the Lake Range fault, Quaternary sediments overlie the Tertiary and Mesozoic basement complex.

An approximately 1 km right-step in the northern section of the Lake Range fault is connected by an east-northeast striking sinistral-normal oblique-slip fault. The regional extension direction, inferred from geodetic data, and the slip direction of the oblique-slip fault linking the right-step in the Lake Range, taken from preserved slickensides along the fault scarp, both trend to the west-northwest (Teplow et al., 2011). An epithermal gold and silver deposit resides in the intensely silicified intersection of multiple normal faults extending northward of the right-step.

The roughly north-trending, west-dipping, curvilinear Holocene San Emidio fault is located approximately 1 km west of the Lake Range fault in the southern San Emidio basin. The current San Emidio geothermal power plant is producing from a reservoir located at the top of the middle Miocene Pyramid sequence, located at the southernmost surficial expression of the San Emidio fault (Rhodes, 2011).

Production wells in the San Emidio geothermal field are located in the basin, adjacent to fault and spring-related hydrothermal surface alterations and near surface carbonate and silica deposits. Additionally, the production wells are located near the intersection of multiple normal faults.

- 2.2.2: Kinematic analysis

Slip and dilation tendency analysis, in the San Emidio desert, suggests that north-northeast striking faults and fractures are favorably oriented for fluid flow due to west-northwest-directed extension (Rhodes, 2011). Both the Lake Range fault and San Emidio fault trend roughly northward and are favorably oriented for fluid flow at depth.

• Is there anything else you’d like to see here?

Methods

This study utilizes 2-d seismic reflection data from an ~3.3 km east-west trending line. Data was recorded using 3-component geophones with a ~17 m geophone spacing and a vibroseis source with a ~67 m shot spacing. The seismic line used in this study was chosen due to its close proximity to the known geothermal resource and the four productions wells in the San Emidio field. This east-west trending seismic line extends from the Lake Range fault to the east, through the north-trending San Emidio fault, into the center of the basin to the west.

- 2.3.1: Velocity Model: Vp, Vs and Derivative Sections

The fidelity of the results at every step in this study is contingent upon the accuracy of the p- and s- wave velocity modeling process. For this reason, a significant emphasis has been placed on determining the most accurate velocity models possible. The p-wave and s-wave velocity models were computed by Optim Inc. using a simulated-annealing algorithm that utilizes a Monte Carlo optimization scheme to invert first-arrival picks for velocities (Pullammanappallil, S.K., Louie, J.N., 1994). The complex structure at San Emidio requires a non-linear velocity optimization that avoids assumptions about structural geometry. The simulated-annealing algorithm iteratively converges on an optimized solution, while avoiding becoming fixed at local least-square error minima points. This method allows for an accurate velocity model in the structurally complex San Emidio basin. Travel time plots are constructed using a fast finite-differencing scheme that utilizes a solution to the eikonal equation (Vidale, 1988). Velocity data from the ~3.3 km seismic reflection line was split into a grid that is 400 elements wide by 200 elements deep. Each square element has a length of 8.382 m.