TECHNICAL MEMORANDUM FINAL w Figures 8-16-16.pdf

COTTON, SHIRES AND ASSOCIATES, INC.

TECHNICAL MEMORANDUM

To: Michael Love & Associates, Inc.

Michael Love, Principal Engineer, From: Cotton, Shires and Associates, Inc.

David Schrier – Principal Geotechnical Engineer

Patrick Shires – Principal Geophysicist Dale Marcum – Principal Geologic Engineer John Wallace – Principal Engineering Geologist

Date: August, 2016

Re: Engineering Geologic Investigation Technical Memorandum Eagle Canyon Fish Passage Improvements in Battle Creek

1.0 INTRODUCTION

This technical memorandum presents the results of Cotton, Shires and Associates, Inc. (CSA) engineering geologic investigation for the Eagle Canyon Fish Passage Improvements project located on the North Fork of Battle Creek near Manton, California (see Figure 1). The study area, which consists of an Upper Barrier Site (UBS) and a Lower Barrier Site (LBS), is located about 1.5 miles west of Manton, California, and about 25 miles east-northeast of Red Bluff, California, near the confluence of the North Fork Battle Creek and Digger Creek. In the site vicinity, North Fork Battle Creek serves as the dividing line between Shasta County to the north, and Tehama County to the south. This memorandum summarizes the results of Task 3 work carried out under Work Authorization 2, entitled, “Eagle Canyon Fish Passage Improvements in Battle Creek”, dated June 3, 2015 between Michael Love and Associates, Inc. (MLA) and Cotton, Shires and Associates, Inc.

(CSA).

2.0 SCOPE OF WORK

The work performed for this investigation included:

- Background research and stereo-pair aerial photograph research and analysis;

- Engineering geologic and reconnaissance level mapping;

- Preparation of engineering geologic cross sections;

- Sampling and laboratory testing;

- Geologic characterization and kinematic analyses;

- Geophysical investigation using Refraction Microtremor (ReMI) methods (under Work

Authorization 3); and

- Preparation of this geologic investigation technical memorandum.

3.0 BACKGROUND INFORMATION

3.1 Project Information

The UBS is located just upstream of the confluence of North Fork Battle Creek (NFBC) and Digger Creek, while the LBS is located approximately 1/4-mile downstream of the UBS. The Eagle Canyon Diversion Dam, which is owned and operated by Pacific Gas & Electric Company (PG&E), is located approximately 750 feet downstream of the confluence. The construction date of the dam is believed to have been in the early 1900s (Hepler, 2010). The dam is part of PG&E’s Battle Creek Hydroelectric Project that was first licensed in 1919. The project consists of five diversion dams on the North Fork Battle Creek and three diversion dams on the South Fork Battle Creek, along with 30 canals totaling 54 miles in length that feed five high-head low-volume hydroelectric power plants. The Eagle Canyon Diversion Dam directs water into the Eagle Canyon Canal, which consists of a short tunnel segment near the dam, elevated steel Lennon-type flumes, and unlined canal segments. The Eagle Canyon Canal merges with the Inskip Canal, and the combined water is delivered to the Inskip Penstock and Powerhouse.

Due to declining salmonid populations in the Sacramento River system, a multi-agency agreement was made with PG&E in 1999 to restore and enhance about 48 miles of fish habitat.

As part of this program, the Eagle Canyon Diversion Dam was retrofitted with a fish ladder in 2011. The agreement also includes bypassing or removing the two boulder barriers located near the Eagle Canyon Diversion Dam that are the subject of this memorandum. The UBS is located about 800 feet upstream of the dam, and just upstream of the confluence with Digger Creek. The LBS is about 500 feet below the dam, or about ¼ mile downstream of the confluence. These barriers are composed of large boulders that form natural, and in the case of the LBS, partially man-induced blockages to the upstream migration of steelhead and salmon. In 2012, an engineering geologic investigation of the LBS was performed by the California Department of Fish and Game (Smelser, 2012). This study characterized the conditions of the barrier, and discussed a failed attempt in 1997 to split one of the boulders by drilling a hole and filling it with an expanding grout compound. The study concluded that removal of all of the rocks that form the LBS was not feasible. Instead it was recommended that the barrier be mitigated by removing two of the rocks (by blasting) as well as a portion of bedrock on the north side of the canyon. We observed a web based video, entitled “Battle Creek Rock Exploded To Allow Fish Passage”, dated October 17, 2012, that shows the blasting of a large boulder at the LBS. We do not know how much time was required to drill the boulder prior to the blast, but the technique appears to be an effective tool for reducing block sizes in the creek channel.

3.2 Topographic Setting

The Battle Creek watershed lies on the west slope of Mt. Lassen, and the creek is a cold water snowmelt and spring-fed tributary to the Sacramento River. In the site vicinity, the creek has carved a slot canyon (Eagle Canyon) through a plateau that slopes gently to the west. The canyon is relatively narrow with precipitous to locally overhanging slopes. At the UBS, the canyon is about 40 to 50 feet wide and about 50 feet deep, and at the LBS, the canyon is about 350 feet wide by 200 feet deep. At the LBS, the south side of the canyon intercepts groundwater, resulting in abundant springs and waterfalls, likely fed by snowmelt from Mt. Lassen.

3.3 Regional Geology

The subject site is located on the southern edge of the Cascade Range geomorphic province of California. The Cascade Range includes a chain of prominent volcanic peaks that extend from Northern California through Oregon and into Washington. The southern extent of the province is Mt. Lassen, located about 22 miles east of the site.

The northern Sacramento Valley has a diverse and complex geologic history. Convergence of the Pacific and North American plates has created tectonic stresses that caused the present-day northern Sacramento Valley to go through many changes. From the Mesozoic era through the mid-Cenozoic era, the present-day northern Sacramento Valley was inundated with Pacific Ocean waters, and the Pacific shoreline oscillated back and forth from the eastern side to the western side of the valley. From the mid-Cenozoic era to present, the Pacific shoreline migrated westward to its current position west of the California Coast Ranges, exposing the emergent valley as it looks today. The Sierran metamorphic and plutonic basement rocks extend to the west beneath the valley, and these are unconformably overlain by shales and sandstones of the Great Valley sequence. Beginning in the early-Miocene (about 23 million years ago), volcanism began and this process is still active today (e.g., Mount Lassen’s latest eruptions occurred from 1914 to 1917). On the east side of the valley, the Tuscan Formation (Pliocene age), overlies the Great Valley rocks, and contains layered ash, volcanic mudflows and some alluvium. In the site vicinity, The Tuscan Formation unit is generally covered by young volcanic flows extending from east to west. In the Manton area, the young basalts that underlie the broad plain are exposed on the walls of Eagle Canyon and have been mapped by Helly and Harwood (1985) as the Basalt of Eagle Canyon. A portion of this regional map that includes the area of interest has been reproduced as Figure 2. The rocks that underlie much of the broad plain in the Manton Area are described in the referenced literature as dark grey, vesicular olivine basalt.

3.4 Seismic Setting

The site is located in an area of moderate seismicity. The United States Geological Survey (USGS, 2008a) shows no historic faults (less than 150 years) in the site vicinity. The north-south trending Hat Creek fault (latest Pleistocene fault, less than 15,000 years) is located about 32 miles to the northeast, and the east-west trending Battle Creek Fault Zone (late Quaternary fault, less than 130,000 years) is located about 2 miles north of the site where the broad plain characteristic of the Manton area meets Shingletown Ridge.

Using the interactive website USGS deaggregation procedure (USGS, 2008b, see Appendix D), the peak ground acceleration with a 2% chance of being exceeded in 50 years (about a 2,500 year return period) is 0.40g. The modal distance and magnitude were 3.6km and M6.5, respectively.

A screening analysis outlined in the CGS Special Publication 117A (2008) was used to estimate an appropriate seismic coefficient for a displacement threshold of 5 cm (2 inches). In this method, keq = feq * MHA (keg=seismic coefficient, feq is a factor related to the seismicity, and MHA is the maximum horizontal acceleration at the site). Based on the modal distance and magnitude, the feq factor is about 0.48, which gives a keq of 0.19.

3.5 Geophysical Investigation

A geophysical investigation was performed by CSA using Refraction Microtremor (ReMi) shear wave survey lines as part of this investigation. Two survey lines were performed on both the UBS and LBS, and the data and results are included in Appendix A. Pertinent results from the survey are discussed in general terms in the main body of this technical memorandum; however, the reader is referred to Appendix A for a complete discussion of the methods and results.

3.6 Laboratory Testing

CSA performed laboratory testing on rock core samples and bulk soil materials from the UBS and the LBS. The rock testing included compressive strength and unit weight, while the soil testing consisted of Atterberg Limits, particle size (sieve analysis, hydrometer and percent minus #200), corrosivity, and residual and fully softened peak torsional shear strength. In general, the UBS rock testing yielded unit weights that varied from 130.9 pcf to 168.6 pcf, and had compressive strengths of 7,960 psi to 16,310 psi, while the LBS rock tests yielded unit weights of 155.1 pcf to

164.4 pcf, and compressive strengths of 14,880 psi to 16,320 psi. There are two low unit weight tests from the UBS (132.1 and 130.9 pcf) that were performed on vesicular cobbles that likely were transported downstream to the site, and are not representative of the in-place rock exposed on the canyon walls or the large boulders that comprise the in-stream barriers.

Our laboratory testing on the UBS soils indicates that the materials are silty sands and sandy silts with low plasticity (PI=7). The sand has low fines content (<2%), moderate to high fully softened peak and residual shear strengths (φ=320 and φr=310), and low corrosivity (chloride < 2 mg/kg, Sulfate = 54 mg/kg, Min Resistivity = 5095 Ohm-cm).

Our laboratory testing on the LBS soils indicates that the materials are non-plastic silty sands with moderate fines content (8% to 18.3%), moderate to high fully softened peak and residual shear strengths (φ=330 and φr=330), and low corrosivity (chloride = 2 mg/kg, Sulfate = 45 mg/kg, Min Resistivity = 16,366 Ohm-cm). The Laboratory test results are presented in Appendix B, and the locations of the samples are shown on Plates 1 and 4 for the UBS and LBS, respectively.

4.0 UPPER BARRIER

4.1 Preliminary Fish Bypass Concepts

The UBS is located on NFBC just upstream of the Digger Creek confluence, and includes many large boulders tens of feet in size that create an impassable barrier for fish. Prior to our investigation, preliminary fish bypass concepts were prepared by Michael Love & Associates, Inc.

and depicted in drawings dated November 2015. Photographs 1 and 2 (below) show the general character of the canyon in this area, as well as some of the locations of proposed improvements that are discussed below. The concepts are discussed briefly in this section, as the elements are referred to in our descriptions of the fieldwork. A more detailed discussion of the concepts, the potential hazards, and constructability issues are presented in a later section of this technical memorandum.

Photograph 1 – Battle Creek just upstream of the confluence with Digger Creek. Rocks with Control Points 7 and 8 are noted. West Canyon Wall on right, East Canyon Wall on left. View downstream (to south).

Photograph 2 – The confluence of Digger Creek and Battle Creek looking downstream. Control Points CP-2, 3 and 4 are noted. The proposed location of the fish ladder is over an apron of boulder alluvium, and closer to the west canyon wall slope debris.

CP-7

CP-8

CP-2

North Fork Battle Creek

Digger Creek

CP-3

CP-4

Approximate area of spillway, headworks walls

Approximate area of upstream tunnel portal

Approximate location of flume

Approximate location of Downstream Tunnel Portal

Area of Fishway (ladder)

4.1.1 Concept 1 – Flume and Fish Ladder - This flume and fish ladder concept is shown below in Photograph 3, and includes a concrete ramp fishway at the downstream end (in the vicinity of CP-2, see Photograph 2), a flume, spillway, and headwalls at the upstream end (in the vicinity of CP-7 and 8, see Photograph 1).

Photograph 3 – A photo-illustration of the Flume and Fish Ladder alternative (Concept 1).

North is to the left.

Digger Ck

NFBC

4.1.2 Concept 2 – Tunnel and Ladder – This concept has some of the same elements as Concept 1, such as a fishway ladder, the removal of boulders, a spillway, and headworks walls installed at the upstream end. However, instead of a flume, the fish would enter into a tunnel at the end of the fishway. The concept is shown below in Photograph 4, and the approximate locations of the upstream and downstream tunnel portals are shown in Photographs 1 and 2.

Photograph 4 – A photo-illustration of the Tunnel and Ladder alternative (Concept 2). North is to the left.

Digger Ck

4.1.3 Concept 3 – Regrade (with Larinier Superactive Baffle) – This concept is shown below in Photograph 5. The passage would be created by constructing a concrete box ramp (at a 14% grade) between two concrete walls, and installing a series of baffles (Larinier) to slow the water flow.

Photograph 5 – A photo-illustration of the alternative of regrading the central Battle Creek channel with a Larinier Superactive Baffle (Concept 3). North is to the left.

4.1.4 Concept 4 – Regrade (Natural) – This concept is shown below in Photograph 6. This option creates a channel at a 10% gradient with a natural rock surface instead of the baffle structure.

Photograph 6 – A photo-illustration of the alternative that includes regrading the creek bed (Concept 4). North is to the left.

4.2 Field Program

Fieldwork was performed at the UBS by John Wallace (engineering geologist) and Dale Marcum (geologic engineer) on December 1 and 2, 2015, and from February 29 through March 2 of 2016 with John Wallace, Dale Marcum, Patrick Shires (geophysicist) and Marcel Peliks (staff geologist).

The work included engineering geologic mapping, rock mass characterization, joint characterization, sample retrieval, Schmidt hammer testing, geophysical surveying, and rock slope evaluations. GMA Hydrology surveyed the site between July 13 and November 4, 2015 for the purpose of generating a topographic base map, which was used by CSA for geologic mapping and rock face logging.

The results of the engineering geologic mapping are presented on the Engineering Geologic Map (Plate 1), the West Wall Geologic Face Log (Plate 2), and the East Wall Geologic Face Log (Plate 3). Prior to entering the field, horizontal views of the west and east canyon walls were prepared using the LiDAR point cloud data. In addition, photographs were taken by CSA of the west and east canyon walls from the canyon rims, and the photographs were printed out prior to mapping.

Both the west and east canyon walls had areas obscured by vegetation where the point cloud views and the photographs were obscured. In these areas, we made sketches of the walls using measuring tape and laser hand tools. The methods used to map the face logs (point cloud, photographs or measurements) are noted along the base of the logs in Plates 2 and 3. Based on the methods used, we estimate the horizontal and vertical accuracy of the features we mapped is likely within about 5 feet.

Observations and mapping performed to characterize the rock mass and to gather the data for this evaluation were made from the top of the west and east rims, and along the base of the west and east walls. The bottom of the canyon was accessed by descending into Digger Creek, and also by descending a game trail on the west canyon rim that intersected the NFBC about 300 feet upstream of the study area. For future reference, we noted a black bear on this game trail during our December 2015 mapping. The east canyon rim was accessed by crossing the NFBC at the game trail, and hiking up another game trail on the east side.

4.3 Earth Materials

The earth materials in the vicinity of the UBS consist of bedrock materials of the Eagle Canyon Basalt and several non-bedrock units consisting of colluvial soil, creek bed alluvium, slope debris, and unstable bedrock blocks along the precipitous Battle Creek embankment walls. These materials are described as follows, and their spatial distribution is illustrated on Plate 1 (Engineering Geologic Map):

Colluvial Soil (Qc) – The young, resistant character of the Eagle Canyon Basalt results in very shallow soil development. Soil materials consist of thin (generally less than several feet), sandy, clayey silt with large basalt rock fragments. On Plate 1, these materials are identified as Qc over Qeb, indicating very shallow soil cover blanketing bedrock.

Alluvium (Qa) – Within the incised NFBC corridor, thin, small pockets of sand and gravel are locally deposited in protected areas and back-eddy locations. Due to the precipitous embankment sidewalls, the sudden nature of the flood flows, and ‘slot canyon’ character of the gorge, these deposits are limited in size, and likely temporary in their location.

Slope Debris/Talus (Qsd) – In isolated portions of the NFBC gorge, talus and debris shed from the precipitous upper slopes have accumulated as aprons along the lower portion of the canyon.

Slope debris that has been mobilized and deposited into the flow path of the creek has been eroded of all but the largest blocks, stranding these within the creek bed. Plate 1 illustrates some (but not all) of the larger blocks within the creek where slope failures have deposited large blocks into the NFBC and formed the upper fish barrier.

Rock Topple/Unstable Basalt Rock Face (Qbb) – Massive basalt exposures flank the UBS just upstream from the confluence with Digger Creek, and the vertical rock walls contain pervasive vertical cooling joints that are highly susceptible to rock topple. Additional discussions regarding possible mechanisms that initiate block toppling at the site are discussed in a later section.

Eagle Canyon Basalt (Qeb) – The NFBC gorge area is underlain by the Eagle Canyon Basalt, which is characterized by a thick-bedded to massive, slightly weathered to fresh, very hard and very strong basalt. Bedrock is exposed along the plateaus above the gorge where soil cover is sparse, and excellent exposures crop out in the canyon bottom. The Red Bluff Formation volcaniclastic deposit was not identified in the UBS, but crops out near the LBS at an elevation consistent with that of creek level at the UBS. Potential ramifications of this are discussed in a later section.

4.4 In-Place Rock Strength

The strength of the rock was estimated using three different methods:

1) Field estimates of rock strength using a rock hammer and a qualitative scale (Hoek and Brown, 1997) varied from strong (unconfined compressive strength between 7,250 and 14,500 psi) to extremely strong (unconfined compressive strength greater than 36,000 psi).

2) Schmidt Hammer tests were performed at various locations along the west canyon wall. The locations of the tests are presented on Plate 1, and the results are included in Appendix C. The unconfined compressive strength correlations for the tests along the west wall varied from about 22,000 psi to 50,000 psi, with an average of about 33,500 psi. We note that typical unconfined compressive strengths for basalt published by Attewell & Farmer (1976) range from about 15,000 psi to 45,000 psi, so these results place the strength of this rock near the upper end of that range.

We used a Proceq Model L/LR Schmidt Hammer and correlation between the rebound number and unconfined compressive strength from Deere and Miller, 1966, as published by Hoek and Bray, 1997 to estimate the unconfined compressive strength.

3) Unconfined compression laboratory tests (see Appendix B) were performed on two rock fragments taken from the ground surface close to the base of the west wall near the downstream portal of the proposed tunnel (Option 2). The results included unconfined compressive strengths of 7,960 psi, 11,510 psi and 16,540 psi. The 7,960 psi result was from a vesicular boulder that was likely transported downstream to this position, and we do not consider this sample representative of the rock exposed on the canyon walls.

In general, there appears to be a large difference between the laboratory unconfined compressive test results, and the field estimates using rock hammers and the Schmidt Hammer. We suspect the laboratory test results were lower due either to disturbance from coring in the lab, and/or the presence of micro fractures in the basalt.

4.5 Boulder Strength

The unconfined compressive strength of several of the large boulders was also estimated using qualitative methods and Schmidt Hammer testing. Field estimates of rock strength using a rock hammer varied from very strong (unconfined compressive strength between 14,500 and 36,250 psi) to extremely strong (unconfined compressive strength greater than 36,000 psi). Field estimates were also made using a Schmidt Hammer on six of the boulders at the UBS, and the results ranged from 25,000 psi to 75,000 psi, with an average of about 37,000 psi. The locations of the Schmidt Hammer tests are shown on Plate 1, and the results are included in Appendix C. As noted above, published unconfined tests for basalt range from about 15,000 psi to 45,000 psi, so although the highest Schmidt Hammer result of 75,000 psi appears unrealistically high, the boulders likely have unconfined compressive strengths in the upper end of the published range.

4.6 Rock Mass Characterization

The rock exposed in both the west and east canyon walls is generally non-vesicular basalt with a structure that varies from thin sub-horizontal flow structures (less than 1 foot thick) to massive.

The rock is generally fresh to slightly weathered, with fractures that vary from widely spaced (3 feet to 10 feet), to extremely widely spaced (greater than 10 feet). We characterized the rock mass using the Geologic Strength Index (GSI) developed by Hoek et al (2013). Visually, we classified the structure of the rock as “blocky” and the surface conditions as good to very good, resulting in a qualitative GSI for the rock mass of 60 to 90. We also used two numerical methods to estimate the GSI (Hoek et al, 2013) which uses the average number of joints per meter, and qualitative characterizations of the joints (Jcond89 from Bieniawski, 1989, and Jr/Ja from Barton, 1974). The GSI values from these quantitative methods were 78 and 91, respectively.

To estimate Mohr Coulomb strength parameters for the rock mass, we used the computer program RocLab (Rockscience, 2013). Using the lower GSI value from our visual assessment (60), an unconfined compressive strength of 16,000 psi, a disturbance factor (to characterize effects of blasting) of 0.0 (i.e., no blasting disturbance), and an ‘mi’ of 25 (typical value for basalt given by the computer program), the resulting strength envelope parameters for the rock mass (using a stress range appropriate for a 100-foot high slope) are Ø=65°, c = 176 psi. The computer plot from RocLab is included in Appendix C.

4.7 Rock Joint Characterization

As shown on Plates 2 and 3, although the rock mass is infrequently dissected by joints, the joints are moderately continuous (10 feet to 30 feet in length), to highly continuous (30 feet to 100 feet in length). The joints are strongly curvilinear and relatively smooth. Barton Choubey (1977) devised a method to estimate friction angles for joints that take into account the increased strength due to joint roughness. The Joint Roughness Coefficient (JRC) is a number (selected qualitatively) that represents the roughness of the joint surfaces (higher numbers correspond to rougher joints).

We characterized the joints as having JRC’s that vary from 2 to 12, with typical values of about 5 to 7. Using Barton and Choubey’s equation and a JRC=5, an unconfined strength of 8,000 psi along the joints, and a base friction angle (friction angle of the material without any roughness) for basalt of 31 degrees (low end of published values in the reference), the best fit representative joint friction angle for the canyon walls is 42 degrees. A spreadsheet showing these calculations is included in Appendix C. For conservatism, however, we recommend using 35 degrees, which is the upper value given in the reference for base friction angles for basalt.

The joint apertures varied from tight to up to 2 inches wide, and were generally open ¼ to ½ inch.

We estimated the Jcond89 value (Bieniawski, 1989) to be 16, which places the joints between the categories for Jcond89 = 10 (described as slickensided surfaces or gouge <5mm thick, and 1-5mm of separation) and Jcond89 = 20 (described as slightly rough surfaces, <1mm of separation, highly weathered walls). These are general descriptions from Bieniawski, and although the joints did not exhibit slickensides or contain clay gouge, the low score is the result of the continuous nature of the joints, the apertures, and the relative smoothness of the joint walls. We also evaluated the Jr (Joint roughness) and the Ja (Joint alteration number) after Barton (1974). We assigned the joints a Jr = 2.5, between the categories for Jr = 2 (smooth, undulating) and Jr = 3 (rough and irregular, undulating). We assigned the joints a joint alteration (Ja) = 1, corresponding to the description of “unaltered joint walls, surface staining only”.

We took attitude measurements (strike and dips) of the joints on the west and east canyon walls, and plotted the data (the normals or poles to the planes) on a lower hemisphere stereonet, included as Figure 3a. The analysis and plots were prepared using Dips by Rocscience (2003).

The shaded contours show concentrations of poles, and the mean attitudes of the three major joint sets (i.e., the joints that are commonly repeated) were determined using Dips, and these are shown in Figure 3b. The major joint sets are:

Set 1) “Northeast” N55E, 90 Set 2) “North-South” N10W, 88E Set 3) “Northwest” N62E, 90

For our purposes, the 88 degree dip for Set 2 will be treated as vertical (i.e., 90), and the major joint sets are also shown on Plate 1. Note that these are mean orientations, and the strikes of the joints in the field can vary by up to 15 to 20 degrees. We note that most rock masses are commonly dissected by anywhere from two to five major joint sets, so the number of major joint sets at the site (three) is relatively common. What is somewhat unusual in our experience is all three of the major joint sets have vertical to near-vertical orientations.

4.8 Kinematic Analyses

Kinematic analyses are commonly used in rock slope investigations to gain insights into the type of rock block failures that can be anticipated (e.g., block sliding, wedge sliding, toppling), and the general stability of the features. The great circles of the major joint sets are typically plotted on a lower hemisphere stereonet, along with the great circle of the plane representing the slope face.

If the dip vector of a joint plane plots outside the great circle that represents the attitude of the rock slope, then block sliding could occur. Likewise, if the intersection of two joint planes plots outside the great circle of the rock slope, then wedge sliding could occur. Consequently, we plotted the great circles of the three major joint sets on Figure 4a, along with the three major orientations of the west canyon wall in the site vicinity. As can be seen on this figure, all of the planes (joints and rock slope orientations) are vertical, and all intersect each other at a single point in the middle of the stereo net. This suggests that typical rock block failure modes, such as blocks on inclined planes or wedges that daylight in a slope face will not be predominate failure modes at the site. What these joints will form, however, are rock columns (rectangular or triangular shaped prisms).

Rock columns (made of near-vertical side walls) will likely be subject to block toppling, and as will be discussed in a later section, we observed this failure mode occurring at the UBS. The kinematic test for flexural rock toppling involves plotting the joint normals on a lower hemisphere stereonet, plotting the great circle of the slope face on the same stereonet, along with a safe zone equal to the friction angle of the joint interface, and an unsafe zone beyond the friction angle. Joint normals that plot outside of the friction angle may be subject to flexural toppling.

Although we know block toppling is occurring at the site, and the kinematic test for rock toppling is for flexural toppling, we performed a kinematic analysis for flexural toppling along the west canyon wall to see if we could gain additional insights. The results are shown on Figure 4b, where we used a joint friction angle of 35°, and where the unsafe zones for the three orientations of the west wall are shown in shaded red. In this figure, the Northeast joint set plots in the unsafe zone for the downstream portal area (where it is subparallel to the canyon wall), and the North-South joint plots in the unsafe zones for the middle and upstream areas of the west wall (where it is also subparallel to the strike of the canyon). Although flexural toppling does not appear to be an important failure mechanism at this site, this exercise suggests the joints may be releasing joints (i.e. back joints) in rock columns at these locations, and may have structurally controlled the orientation of the canyon in this area.

As shown on Plate 1, we placed the major joint sets in groups along the west and east canyon walls to estimate which sets may be forming rock columns for the different areas, and these are reproduced below in Photograph 7.

Photograph 7 – Possible column-forming major joint sets.

4.9 Block Toppling Discussion

A rock column with vertical sides and a horizontal base will be in contact with the rock slope via vertical joints, and the weight vector of the rock column will also be vertical. Therefore, the friction of the joint interfaces on the vertical joints should not have a significant effect on the stability of a detached column. Instead, the criteria that control the stability of block toppling in this case are the width of the base versus the height (i.e., tall narrow columns are more subject to block toppling than shorter wider columns), and destabilizing forces that act to tip the block past the center of gravity. Lateral forces that can start the toppling process can include frost (ice) jacking, undercutting, hydrostatic forces, seismic forces, creep of overlying soil, and stone jacking.

Stone jacking is a process where small rocks enter an open joint, and as the joint aperture widens due to temporary temperature changes, the stone falls further into the joint and acts as a cam to keep it open. This process is repeated seasonally until the joint opens to the point where the column becomes unstable and topples. At the subject site, we observed this process occurring along a joint on the east canyon rim (see Photograph 8, below).

Photograph 8 – Stone jacking occurring in the releasing joint of a rock column above the east rim of the canyon.

Another column destabilizing mechanism that we observed at the site was undercutting of the columns by river erosion (see Photograph 9).

Photograph 9 – Rock column (termed “The Wave”) actively being undercut by the river below the water line, with evidence of past scour preserved above. A prominent toppling releasing joint is at the rear of the block, and close near-horizontal fractures in the rock mass attest to rock stress.

Thus, of the list of potential destabilizing forces that could initiate block toppling, two were observed at the site, and we suspect hydrostatic, frost jacking, and seismic forces also play a role in block topple initiation.

4.10 West Canyon Wall Mapping

Mapping of the west canyon wall was performed by viewing the wall from the base of the slope, and from the east canyon rim. As described previously, a graphic log was prepared and is included as Plate 2. The following discussion of the west wall will be presented from upstream to downstream, and pertinent portions of Plate 2 have been copied and embedded into the text for convenience. The identified rock blocks were characterized by their potential to fail using the legend presented below

High – Failure could occur at any time, but likely within 10 years.

Moderate to high – Failure likely within 10 to 50 years Moderate – Failure possible in 50 to 100 years Low – Failure unlikely within 100 years

Stations 0+00 to 0+30 – This section is just upstream of the fish passage concepts 1 through 3, and at the upstream end of concept 4 (the natural grading option). The NFBC begins to bend to a north-south orientation in this area, and the water is forced up against the west side of the canyon, resulting in high scour velocities and erosion of the rock walls. The erosion has undercut a large rock column labeled “The Wave” or Block W1 (Photograph 9 and Figure A, below) that has begun to topple along a back joint within the north-south joint set. The joint is open 1 to 2 feet at the top, is in-filled with soil and vegetation, and is tight near the river. The downstream end of The Wave is about 1 to 15 feet upstream of the walls for Concepts 1, 2 and 3. We estimate the block has a moderate to high failure potential, and although the proposed concrete walls for some of the concepts are not in contact with the block, failure of the column could damage the improvements.

Figure A – Stations 0+00 to 00+30 West Wall Geologic Face Log.

Stations 0+30 to 0+80 - The headwalls for Concepts 1, 2 and 3, as well as the upstream tunnel portal for Concept 2 are located within this section of the west canyon wall (Figure B and Photographs 10 and 11, below). The locations where the various conceptual concrete walls would be in contact with the canyon wall contain a block with a moderate failure potential, labeled Block W2. Block W2 (moderate to high failure potential) is bounded on the upstream side by a large open joint, and there are a series of near-horizontal fractures within the block that may indicate the build-up of compressive stresses due to removal of support by river erosion and incipient toppling. The upstream tunnel portal for Concept 2 daylights in Block W6 (moderate to high failure potential due to horizontal fracturing) and is below the highly unstable Blocks W3, W4 and W5. These blocks are currently failing due to undercutting, have multiple open joints, and have a high potential to fail. These blocks are also rectangular prisms that have been formed by the northwest joint set along the back (releasing joint), by the northeast joint set on both sides, and by the northwest joint set at the front.

Figure B – Stations 00+30 to 00+80 West Wall Geologic Face Log.

Photograph 10 – The west canyon wall at the location of proposed upstream tunnel portal where loose blocks are above the portal area, and are causing stress fractures in the rock below.

Photograph 11 – Unstable blocks W3 – W6 on the west canyon wall.

Stations 00+80 to 1+50 - This portion of the canyon wall (Figure C, below) would be traversed along its base by the flume in Concept 1, and the downstream wall for Concept 3 would abut the canyon wall near Station 1+00. There are two relatively small blocks with a high potential to fail (Blocks W8 and W10), and a medium sized block (W7) that is partially detached and has a moderate to high failure potential. Much of this portion of the wall contains moulds, or cavities where previous rock column failures evacuated (likely from toppling), and these past failures may be the source of some of the large boulders in the NFBC that formed the UBS. Many of the moulds suggest that the rock columns were triangular and rectangular shaped prisms formed on the sides by the northwest and northeast joint sets, and at the front by the north-south joint set.

Figure C - Stations 00+80 to 1+50 West Wall Geologic Face Log.

Stations 1+50 to 2+20 - This section of the wall is near the confluence with Digger Creek where the NFBC (and the canyon wall) begins to bend toward the southwest. Many blocks are actively failing and have a high failure potential (Figure D, below). Many of the blocks are completely detached from the wall (e.g. W14, W18, W19, W 20 and W-21), and several are partially detached (W13, W15, W16 and W17). In general, the blocks are triangular and rectangular shaped prisms or wedges formed by the northwest and northeast joint sets. Due to the number and size of the blocks, this area was named the “Block Party”. The semi-circular indentation in the edge of the cliff (see Plate 1) is likely the result of previous large rock block failures. The channel below this area contains numerous boulders that are part of the UBS, and we suspect many came from the Block Party area. This area would be traversed along its base by the flume in Concept 1. The fishway ladder for Concept 1 would end near Station 1+90 and would be below this area, and the ladder for Concept 2 would end around Station 2+10 (i.e., would end near the edge of this zone).

Dense vegetation prevented good photographs of this area.

Figure D - Stations 1+50 to 2+20 West Wall Geologic Face Log.

Stations 2+20 to 2+80 - This area of the west canyon wall is oriented in the northeast-southwest direction, and is the most stable area we observed on the west canyon wall. As shown below in Figure E, the downstream portal for the tunnel in Concept 2 is located in this area. There are several relatively small blocks with a high failure potential, and a rock column on the rim of the canyon (see Plate 1), that may pose a hazard; however, the hazard is generally low to moderate.

Due to the dense vegetation in this area photographic coverage was poor, consequently the geologic face log was prepared using tape and hand-held laser measurements.

Figure E - Stations 2+20 to 2+80 West Wall Geologic Face Log.

4.11 East Canyon Wall Mapping

Mapping of the east canyon wall was performed by viewing the wall from the base of the slope, and from the west canyon rim. A graphic log of the wall was prepared and is included as Plate

3. The following discussion of the east wall will be presented from upstream to downstream, and pertinent portions of the Plate 3 have been copied and embedded into the text.

Stations 0+00 to 0+80 – The section of the east canyon wall (Figure F, below) contains the proposed headwalls for Concepts 1 through 3. Within the area where the proposed concrete walls would abut the canyon wall, there are two large rock columns (E2 and E4) that have open back joints of ½” to possibly several inches (i.e., are actively toppling), and one smaller block (E5) that appears to be toppling. These blocks have a moderate to high failure potential. In general, the columns are formed by the northwest and northeast joint sets.

Figure F - Stations 0+00 to 0+80 East Wall Geologic Face Log.

Stations 0+80 to 1+70 – This section of the east canyon wall (Figure G, below) includes the proposed location of the downstream concrete wall for Concept 3. The wall would be located near the edge of a series of toppling blocks that have open joints (up to 4 inches wide) and a high potential to fail (Block E7). The portion of the wall from Station 1+20 to 1+70 has no observed joints, and is relatively stable.

Figure G - Stations 0+80 to 1+70 East Wall Geologic Face Log.

Stations 1+70 to 2+40 – This portion of the east canyon wall (Figure H, and Photograph 12 below) contains many, very large and tall blocks that are in various stages of active toppling failure, and the area was labeled “Land of the Giants”. The joints are open anywhere from about 1 inch (Block E9), up to 2-4 feet wide (Blocks E8, E9 and E10). The area in front of the toppling columns has numerous large boulders that contribute to the UBS, and we suspect many came from this area of the canyon wall.

Figure H - Stations 1+70 to 2+40 East Wall Geologic Face Log.

Photograph 12 – Large toppling blocks in the ‘Land of the Giants’ area of the east canyon wall.

This photograph was taken from directly above the 20 to 30-foot tall blocks, and illustrates the large gaps at the backs of these blocks up to 4 feet in width.

4.12 Major Toppling Zones

As discussed in the preceding section, there is a major zone of block toppling occurring on the west wall (the “Block Party”) and another on the east wall (“Land of the Giants”), and both areas are near the confluence with Digger Creek. The west and east walls also contain other smaller, localized toppling zones with moderate to high potential for failure. We note that the canyon dramatically widens and deepens below the confluence with Digger Creek, likely due to increased scour potential from the combined hydraulic forces of the two creeks, and due to geologic factors discussed below. We note that the canyon is only approximately 40 feet in width upstream of the UBS, and approximately 3 to 5 times that width immediately downstream. Since these two major toppling zones appear to have generated many of the rock blocks that make up the UBS, the purpose of this section is to discuss possible mechanisms responsible for these concentrated areas of instability.

In the case of the Block Party, the west wall changes orientation at Digger Creek, and this has created more open space (more free face exposures, or freedom of movement) for blocks to move.

Thus, geometry and kinematics appear to be a contributing factor for the instability at the Block Party location. This is not the case, however, at the Land of the Giants, where the east wall is only slightly curved.

Another contributing factor could be the Red Bluff Formation. This is a volcaniclastic alluvial deposit that stratigraphically lies below the Eagle Canyon Basalt (which forms the walls of the UBS), and above the Basalt of Coleman Forebay, which forms the lower walls of the LBS. We observed likely outcrops of the Red Bluff Formation at the LBS, and it appears to be a permeable unit that is responsible for the abundant springs along the south wall. We also observed overhangs and unstable blocks of Eagle Canyon Basalt above the Red Bluff Formation at the LBS.

We note that at the LBS, the Red Bluff Formation crops out an elevation that is roughly consistent with the level of the creek at the UBS. We did not observe the Red Bluff Formation at the UBS, but our geophysical surveys reveal that there is an average of 10 to 15 feet of alluvium and large rock blocks choking the creek at the UBS, and the likely location of the Red Bluff Formation would be at the base of the alluvium, where scour at high flood flow would undercut the near-vertical cliff face. Consequently, we postulate that the dramatic widening of the canyon just downstream of the confluence of Battle and Digger Creeks, and the two major block toppling zones (i.e., the Block Party and the Land of the Giants), may be related to the location of this confluence, and their combined ability to erode into the weaker Red Bluff Formation volcaniclastic deposit and undercut the massively bedded Eagle Canyon Basalt. The Red Bluff Formation may also be a source of high water pressures that are pushing upwards on the base of the blocks and decreasing stability. Under this scenario, in a geologic sense, the UBS represents the active headward extent of canyon widening, where recent massive basalt blocks have been undercut, have toppled, and have choked the narrow gorge, but have not yet been reduced to smaller block sizes that can be transported downstream.

4.13 Mitigation Concepts, Constructability and Geologic Hazards

All of the mitigation concepts involve removing boulders in the channel, and one involves tunneling through the canyon wall. Some level of blasting will likely be required to break up the larger boulders. We observed a video of a boulder being blasted at the lower barrier (https://youtu.be/vdgYSdcRUAE), and that technique appears to have been very effective. Due to the large number of detached boulders on the walls of the canyon, safety and/or monitoring measures will also be needed to protect the workers from vibration induced rockfalls during blasting.

As mentioned previously, the canyon has numerous rock slope instability hazards that pose a risk to most of the concepts. For typical permanent repairs where the consequence of failure is high (i.e., steep slopes above roadways or habitable structures), a common rock slope mitigation method is to add support using tensioned rock bolts that are grouted in place and have a Double Corrosion Protection (DCP) system. However, there are several challenges associated with using DCP rock bolts at the UBS, including: 1) Safety – Due to the number of open joints, fractures and partially to fully detached rock blocks that characterize some of the hazardous areas, an experienced contractor will be needed to perform the work safely and ensure that construction, specifically drilling for the rock bolts, doesn’t trigger rock falls. We suspect that prior to drilling, the contractor will need to apply a layer of shotcrete reinforced with wire mesh to safely move the drill rigs around on the slope and drill, which will increase the cost. 2) Grouting – Achieving a fully grouted rock bolt may be difficult due to the numerous open fractures. A method to contain the grout that is sometimes successful is the use of grout socks. Another possibility is the use of rock bolts that do not require grout, such as bolts with resin cartridges or the use of mechanical rock bolts. These types of bolts are not double corrosion protected, are not suitable for protection of important structures, and are generally only used for temporary support.

However, a conversation with a rock bolt manufacturer (Williams Form Engineering) indicates these types of anchors have been used for permanent repairs in the past. 3) Environmental – This is related to the rock bolt grouting. Due to the numerous open fractures and potential for grout loss, there is a moderate to high potential for the grout to enter the creek, and provisions to monitor for and contain the grout will likely be needed.

Other factors that will affect the cost include: 1) the remoteness of the site; 2) difficult access to get equipment to the bottom of the canyon; and 3) lack of staging area and difficult maneuvering of equipment and resources in the bottom of the narrow canyon. Consequently, construction at the UBS will likely be significantly more costly than typical construction projects of similar size.

Also, depending on the effort expended to mitigate the various hazards, all of the options will likely require some level of long term maintenance to remain operational.

https://youtu.be/vdgYSdcRUAE

4.13.1 Concept 1 – Flume and Fishway Ladder - The elements of this option are shown in purple coloring on Figure 5 (attached) along with notes depicting the rock slope hazards that were identified during our mapping program and the layout of the geophysical ReMi survey lines. In general, this concept appears to be the most highly constrained by geologic hazards.

Constructability - As mentioned previously, the fishway ladder for this option begins near Control Point CP-2, and extends upslope toward the west canyon wall along the apron of alluvium (and possibly slope debris near the rock wall) that occupies this area. The concept drawings show the ladder is a formed concrete box structure with a slab foundation. The ReMi line RM-2 passes through the middle of the ladder, and the estimated average depth to fresh bedrock using the Refraction Micro-tremor data is about 9 feet. The cuts along the path of the fishway ladder may extend to more than 10 feet below the existing ground surface, and most of the deeper cuts will likely need to extend through large boulders that protrude above the surface and possibly into the underlying bedrock. Once the boulders have been split and removed, the subgrade below the ladder foundation will likely be composed of large boulders within a matrix of sand and cobbles, possibly with voids, and will likely be subject to settlement over time. Drilling piers through this material into bedrock to support the ladder would likely be very difficult and costly given the range of block sizes, rock hardness, and the cohesionless nature of the materials.

Possible mitigation methods for a shallow foundation could include extending the excavation into bedrock, or grouting and/or heavier reinforcement in the slab foundation to protect the structure against future settlement.

The flume portion of this concept is mostly in cut, which would likely have similar foundation conditions as the fishway ladder. Perhaps the best option to support the flume, from a constructability viewpoint, would be to attach the flume to the wall of the canyon with braced steel supports. However, as discussed in previous sections, the wall is subject to rockfall hazards.

Since the flume hugs the west wall of the canyon, the alluvium may be shallow, and piers embedded in bedrock may be an option. The upstream 30 feet or so of the flume is shown on the drawings to be elevated about 4 feet.