III. PROJECT OVERVIEW AND TIMETABLE

• Overview of the Science Plan
• Project Timetable and Hole Design

In this Section we provide an overview of the key elements of the drilling operations and science plan and the overall sequence of proposed activities. In Section IV we discuss in detail the specific work plans of the different researchers participating in this project as they relate to the activities outlined below. In Section V we present additional technical details on the drilling and operational plan.

Overview of the Science Plan

Rock and Fluid Sampling. Our sampling strategy has been designed to maximize the scientific return from this experiment, regardless of any operational difficulties that may be encountered, and to allow for continual improvement in our knowledge of the composition and structure of the fault zone during the experiment so that subsequent sampling operations can be carried out with a maximum of efficiency.

Rock samples will be obtained from the fault zone and adjacent crust in four ways:

1. During the initial rotary drilling phase, cuttings will be continuously collected, described and logged. The drilling budget includes additional geologists to work in the mud logging unit 24 hours a day to assure that appreciable cuttings are collected, accurately described and properly archived. These geologists will be trained and supervised by the principal investigators responsible for core analysis.
2. Eight, 10-m-long, spot cores will collected at convenient locations-five cores will be obtained from the from within the fault zone and three from the Salinian block adjacent to the fault zone.
3. Assuming that hole conditions permit, side-wall cores will be collected (principally from within the fault zone). We have budgeted for a side-wall coring technology-involving drilling a core out of the side of the hole-that will work in "hard" rocks (see Appendix B). The use of conventional side-wall coring is contingent on "soft" formation conditions, as the side-wall sampling tool "shoots" the core barrel into the formation. We have both technologies available and hole conditions will determine which technology, if either, can be used for side-wall core recovery.
4. As alluded to in Section I (and described in more detail below), a separate continuous coring phase will be conducted 2 years after the end of the rotary drilling phase. We propose to drill four continuous core holes, each ~250 m in length, as "laterals" from the main borehole. The locations of these coreholes will be carefully selected on the basis of the results obtained in the initial rotary drilling phase and the subsequent two-year period of fault zone monitoring.

Sampling of fluids for geochemical measurements will be obtained in several ways:

1. During both the initial drilling phase and the final continuous coring phase of operations, gases dissolved in the drilling mud will be analyzed on a continuous basis utilizing extraction and analysis techniques developed during drilling of the KTB-pilot hole and main borehole (see discussion in Section IV). Several different tracers (such as fluorescene) are being considered for use in the mud system to make it easier to identify when inflow occur.
2. After drilling and casing of the rotary drilled hole to TD, a series of 10 Drill Stem Tests (DST's) are planned through perforations in the cemented casing. These tests are described in more detail in the next Section. Each DST should provide relatively large volumes of formation fluid which can be used for subsequent analysis by the various PI's.
3. After each core hole is drilled, fluid samples will be extracted from the core holes after each is drilled.
4. Fluid samples will be extracted from core samples.

Taken all together, these two multiple sampling strategies should provide ample rock and fluid samples for the principal investigators to use in the studies described in Section IV.

Downhole Measurements. Downhole measurements are critical to understanding overall fault zone properties and behavior and a multiple measurement strategy is planned to assure their success. Additional details on the downhole measurements plan are provided in Section IV and in Appendix B.

1. First, hole conditions permitting, a comprehensive suite of geophysical logs will be run prior to casing each section of the borehole. The logs will be acquired commercially using state-of-the-art technology currently used in the petroleum industry. In situ temperature measurements will be made repeatedly by USGS personnel during the project, both before and after the hole is cased.
2. Because physical property measurements are very important to this project and it is possible that unstable hole conditions might constrain the geophysical logging program, a Logging While Drilling (LWD) system will be used in the lowermost section of the vertical hole and throughout the entire interval through fault zone.
3. After the hole is cased and cemented, a detailed Vertical Seismic Profile (VSP) and surface seismic experiment will be carried out as part of site characterization studies. This will allow the observations made on the core and in the borehole to be "scaled up" and extrapolated away from the borehole.
4. To assess pore pressure, permeability and stress, a comprehensive suite of packer tests will be made after the casing is cemented and perforated, as illustrated in Figure 10. The packer tests will be made by the PI's, who have extensive experience with such tests. The techniques that will be used are relatively standard well-test and hydrofrac techniques that are well-established for determination of the least principal stress.

The pore pressure and stress measurements alone will allow us to test directly several of the hypotheses proposed to explain the weakness of the fault zone. For example, if the Rice (1992) and Byerlee (1990) hypotheses are correct, then both the pore pressure and the magnitude of the least horizontal principal stress should be distinctly higher inside the fault zone than in the adjacent country rock (Figure 10).

Figure 10. Schematic illustration of how a profile of pore pressure and least principal stress measurements can be used to distinguish between the two hypotheses illustrated in Figure 2. (click for more information)

A series of pore pressure and least principal stress measurements can be made (most easily through perforations in the cemented casing, see below) at various positions with respect to the active trace of the fault to directly test these hypotheses. The inset in Figure 10 shows theoretical calculations of pore pressure within the San Andreas fault zone from Rice (1992), in which he assumes that the permeability of the fault zone is much greater than that of the surrounding country rock and decreases exponentially with increasing effective confining pressure. His calculations indicate that if the mean effective normal stress ( = normal stress minus pore pressure) at a maximum seismogenic depth of ~14 km is only 100-200 bars (10-20 MPa) as required by the upper bound strength implied by both the heat flow and directional constraints, then the pore pressure at 3.5 km would be about 900 bars (90 MPa). This is approximately equal to the vertical stress outside the fault zone (see also Figure 2a) and 55 MPa in excess of hydrostatic fluid pressure. Alternatively, if pore pressure does not increase in the fault zone, we would expect that laboratory measurements of rock strength will reveal anomalously low frictional strength. While performing these measurements at 3.0-4.0 km will not reveal whether such weakening processes are operative at much greater depth, these measurements will represent an important first step towards testing these and other hypotheses pertaining to the mechanical behavior of the San Andreas fault.

Fault Zone Monitoring and Site Characterization. As discussed above, the opportunity to continuously monitor an active, seismogenic fault at depth and make near-field seismic observations will answer many questions about precursory processes and earthquake rupture nucleation and propagation. While such measurements will continue for a period of at least 20 years after the drilling has been completed, we propose to conduct an initial phase of fault zone monitoring during the two-year period between the rotary drilling and continuous coring phases of the project. There are two principal reasons for doing this. The first is to locate, with extreme precision, the position of the fault patches generating microearthquakes as well as creep within the fault zone. This information will be critical in deciding where to conduct the continuous coring operations during the second phase of drilling. The second purpose for the initial fault zone monitoring efforts is to have seismometers in place within the fault zone during detailed site characterization studies. As explained in Section IV, several different types and scales of seismic imaging experiments are planned for the 2-year period between drilling phases, which will benefit appreciably from having seismometers located within the fault zone. The instrumentation system schematically shown in Figure 5 was designed to accomplish both of these objectives.

Project Timetable and Hole Design

Figure 11. Generalized time table for Parkfield drilling project. (click for more information)

Pre-Drilling Site Characterization. A variety of site characterization studies near the proposed drill site have been going on for the past several years. For example, the data shown in Figure 9 were collected as part of these studies. The next significant set of studies at the drill site will begin in October 1998, prior to the work described in this proposal. There will be two principal experiments. First, a 7-km long high resolution seismic reflection profile will be shot across the primary drill site and fault zone. Funding for this project has already been guaranteed by the USGS (Earthquake Hazards Program and Venture Capital Fund) for 1998-99 and will be carried out by a team principally coming from the USGS and Stanford. Twelve portable reflection seismographs will be borrowed to install 700 channels of seismometers along a dense, fault-crossing profile (10 m spacing). As the principal costs of this experiment will be borne by the USGS, this experiment is not discussed at length in this proposal. The main goals of this experiment are to refine the picture of near-surface geology in the drill site area, as shown in Figure 8, and to assure that the drill site is not located directly above small-scale secondary faults which might needlessly complicate drilling at shallow depth. The second pre-drilling site characterization study will be to deploy a number of additional seismographs in the region surrounding the drill site for earthquake monitoring. This experiment is described at greater length below.

1999/Year 1. Rotary drilling, downhole measurements and casing of the hole are scheduled to commence in late 1999 and end in early 2000. The recently-developed capability of the petroleum industry to drill "multi-laterals"-satellite wells which drilled from a single "parent" well-has enabled us to make a significant change in our proposed drilling strategy since preparation of Z&H'96. In that proposal, the most operationally challenging part of the project was the necessity to continuously core an inclined hole across the entire fault zone. As the fault zone is likely to be severely crushed and altered to gouge (as well as potentially overpressured), continuous coring of a directionally-drilled hole over such an appreciable distance was going to be a formidable challenge, especially as we needed to maintain sufficient hole diameter to conduct the necessary downhole measurements and deploy fault zone monitoring instrumentation after casing the hole.

Figure 12 presents a simplified view of the rotary-drilled "main hole" penetrating the entire fault zone. After drilling and logging it will be cased and cemented. Rotary drilling, geophysical logging, casing and cementing of such deviated holes is routine in the petroleum industry, even in poorly consolidated and overpressured formations. Thus, by using

Figure 12. Schematic diagram of the proposed drilling project at Parkfield. (click for more information)

a rotary drilling strategy to penetrate the entire fault, even if the rock is seriously disaggregated and pore pressures are quite high, it should be possible to drill through the fault zone. After appreciable study of the results from the rotary drilling phase, four, continuously-cored "multi-laterals" will be drilled off of the main hole at carefully selected locations. Additional technical aspects of the drilling plan are summarized in Section V and Appendix A.

As shown in Figure 12, the main hole will be rotary drilled vertically to a depth of ~ 2.2 km and then deviated through the fault zone at a ~50° inclination from the vertical to a final depth of 4.0 km. The trajectory shown in Figure 12 was designed to satisfy the following geological and geophysical constraints (see Figure 9):

1. To move the surface position of the hole far enough to the west so that it will avoid a fault trending sub-parallel to-and southwest of-the San Andreas fault zone and be well outside of the low-resistivity anomaly imaged by Unsworth et al. (1997).
2. To get very close to the microearthquake hypocenters, as located both by Michellini and McEvilly (1991) and by Eberhart-Phillips and Michael (1993).
3. To pass all the way through the "geophysically anomalous" fault zone as well as through the vertical projection of the surface trace, terminating drilling in Franciscan rocks on the northeast side of the fault (see Figure 8b).

As the hole is being drilled, we will perform two earthquake calibration experiments. Drilling will be stopped for 1 day (after a bit trip at 1.5 km depth and after setting the 13 3/8" casing at 2.0 km depth) and a seismometer will be run into the well. Shots will then be set off at the nearby permanent seismograph stations as well as at the temporary stations to be deployed as part of the site characterization process (see below). This will effectively reverse the path of seismic waves coming from earthquakes near the hole at depth. By "re-locating" the bottom of the hole seismologically, we will be able to greatly improve the hypocenter locations shown in Figure 9 and make minor adjustments to the hole trajectory if needed. While our goal is not necessarily to intersect the very small fault patches that are producing these earthquakes, we do want to be as close to these seismogenic fault patches as possible.

2000/Year 2. Fault zone monitoring begins in the year 2000 and goes on for 2 years. Measurements on core and cuttings will begin as well as analyses of borehole geophysical data.

2001/Year 3. Fault zone monitoring and measurements on core and cuttings continue. In mid 2001 a comprehensive suite of site characterization studies is carried out.

2002/Year 4. Data from site characterization and fault zone monitoring are analyzed. A comprehensive analysis of all available data will be used to pick intervals for continuous coring. As shown in Figure 12 (heavy black lines), continuous coring within the fault zone will be carried out at four depth intervals through "windows" cut in the casing.

The seismicity rate in the area is such that after 2 years of fault zone monitoring there should be a sufficient number of shallow earthquakes near the drill hole to accurately locate the active fault trace(s) using the clamped seismometer array in the borehole. This information, when combined with the geologic data from spot cores and cuttings, geophysical logs, downhole measurements, fluid and gas chemistry data, pore pressure and in situ stress measurements and the results of site characterization studies will enable the science team to determine the optimal intervals for continuous coring. The advantages of delaying coring until after so much data has been collected and analyzed is clear. Each core will be interpretable in terms of its proximity to active fault traces, the composition and physical properties of the fault zone, pore pressure and stress, etc. The comprehensive scientific measurement program planned for the exhumed core as well as the on-site core-handling procedures are described in detail in Section IV. Following core retrieval, the core holes will be lined with uncemented perforated casing and used for monitoring fluid pressure at depth.

2003/Year 5. When coring activities have been completed at the site, a permanent monitoring string will be deployed in the hole so that the hole can be utilized as a continuous fault zone observatory well into the future. Intensive measurements on core samples are underway.

2004/Year 6. Data analysis and measurements on core are completed. Fault zone monitoring continues.