Certified_Final_Tech_Specs_Volume_2I_-_GBR_Appendix_I.pdf

W911WN-13-R-0002

East Branch Lake East Branch Clarion River Elk County, Pennsylvania Dam Safety Modifications/Cutoff Wall

Certified Final Document Volume 2I – Geotechnical Baseline Report, Appendix I April 2013

Appendix I

Instrumentation Data

I – 1

INSTRUMENTATION EVALUATION

EAST BRANCH DAM

1. SUMMARY The project is instrumented to monitor surface water flows, piezometric levels, and horizontal and vertical movements at the dam crest and grouted void. This instrumentation includes eight weirs, 86 piezometers, five alignment pins/settlement monuments, and two inclinometers. Plates 1-7 show the plans, locations and details of the instrumentation at East Branch Dam.

Prior to the Ninth Periodic Inspection (2007), East Branch Dam was evaluated under the Screening Portfolio Risk Analysis (SPRA) program. East Branch Dam received a Dam Safety Action Class (DSAC) II, Urgent – Unsafe or Potentially Unsafe. As a result of this classification, interim risk reduction measures (IRRMs) were identified and implemented to address potential failure modes, including modifying (automating) and adding additional instrumentation. These IRRMs have directly affected the readings and operations of the instrumentation. Prior to the implementation of IRRMs, some instrumentation data was showing changes in conditions, as stated in the previous periodic inspection report.

Inclinometer Summary

Two inclinometers were installed in the fall of 2010 to monitor the area of the grouted void. One inclinometer is located just upstream of the crest and the other is located just downstream of the crest of the dam at Stations 08+59 and 08+15, respectively. Since the initial reading in March 2011, no relative movement has occurred in either inclinometer.

Weir Summary Weir 8 was replaced by weir 8a and Weirs 1 and 6 were automated by US Geological Survey (USGS). Weirs 1, 4, 6, 8a and 9 show no changes in pool-related seepage rates. Correlation plots of weir gage readings versus pool levels for the period of 2000 through 2007 indicate that flows at Weirs 5, 7 and 10 are noticeably higher than for the period of 1994 through 1999. Correlation plots of weir gage readings versus pool levels for the period of 2008 through 2012 indicate that flows at Weir 5 have slightly increased; Weir 7 flow rates decreased at lower pool levels and have remained at these lower rates with the new interim summer pool elevations; and Weir10 generally continues to increase as described in the previous periodic inspection report.

Piezometer Summary

Seventeen piezometers were installed in 2008 at various locations across the project site. Two piezometers were installed in 2010 near the 1957 grouted void.

The majority of piezometers showed no change in pore pressure

I – 2 response to pool levels prior to the implementation of the reduced seasonal pool levels. With the reduced pool level, piezometers are generally showing three behaviors: a gradual reduction in levels over the past four years; a reduction in the peak readings during summer months; or no change in levels at all.

The piezometers in the right abutment that show levels near the reservoir are in the vicinity of the cutoff trench, upstream of the cutoff trench in the foundation overburden near the top of rock, and around the location of the grouted void. The piezometers screened in the foundation overburden downstream of the cutoff trench are consistently dry. Piezometers such as 77-EBD-25, 90-EBD-36, 90- EBD-42, and 90-EBD-45 all read higher than pool levels. Two piezometers in particular, 77-EBD-25 and 90-EBD-45, have been rising since the introduction of the interim reduced pool levels.

Piezometer levels in upstream portion of the valley section of the dam show embankment and foundation pore pressures with a 20-30% response with pool fluctuations. The piezometers downstream in the impervious core show very little to no response as well as those downstream along the valley bottom and the downstream toe. It is difficult to correlate these with pool fluctuations. With the lowered pool elevations, the trend still occurs but at times fluctuation is more muted than previously noted.

Piezometer levels in the left abutment have not changed much since the implementation of the restricted reservoir levels.

In the previous periodic inspection report readings at a few of the piezometers (90-EBD-32, 90-EBD-36, 90-EBD-37, 99-EBD-59, 99- EBD-61 and 99-EBD-62) indicated a trend of higher water levels (2 feet higher) while readings at 97-EBD-54 and 97-EBD-56 indicated a trend toward lower water levels (2 feet lower). In 2009, most of those piezometers indicated no change in reading levels. However, piezometers 90-EBD-32 and 90-EBD-36 showed a reduction of 5 feet. Piezometer 99-EBD-59 levels have slightly reduced by 2 to 3 feet. This reduction started in 2007 prior to the interim reduced water levels.

Alignment and Settlement Monitoring

Alignment and settlement readings show minor variation in alignment (horizontal position only) with no apparent trend. The variation in the survey is smaller than the accuracy requirements of EM 1110-2-1009, Structural Deformation Surveying. Some of the variation may be due to different methods used to survey the monuments since 2005.

Due to the varying survey methods utilized and the datum change, comparisons were made pre- and post- 2005. Survey of alignment pins in reference to the line along monuments M-1 and M-2 indicated movement of 0.47 inches downstream and 0.23 inches upstream, with no discerned trend.

I – 3

Less changes are reflected in settlement as compared to horizontal alignment readings (done at the same survey points), which match previous readings (pre-2008) closely in terms of total displacement and the displacement profile. Therefore, and considering a lack of trend, it seems unlikely that the unusual alignment measurements represent embankment movements; the unusual measurements are likely due to accuracy limitations.

Note that monitoring of the alignment and settlement pins may not reflect the loss of embankment or foundation material through piping.

The cavity that formed in 1957 did not cause any movement in the surface of the embankment.

General Plates 1-7 are included in this appendix for reference. Exhibit H-1, showing the precipitation with corresponding pool for the period 2007 through 2012, is provided for comparison with the instrumentation data. The majority of time history plots in this appendix are also for the 2007 through 2012 time period, with some exceptions. All elevations provided within the appendix are in

NAVD 88.

I – 4

2.

INCLINOMETE

RS

Two inclinometers were installed in the fall of 2010. The orientation of the casing was not installed with the four grooves perpendicular and parallel to the dam axis as recommended. This makes future evaluation difficult since the total magnitude will be not displayed during the readings. Initially, the cable pulley used in lowering, raising, and orientating the inclinometer during data collection could not be attached to the inclinometer casing or to the outside protective casing per design. The set up was modified so that it could be attached to the inside of the protective casing. Due to the pulley setup, the last reading is obtained at the 4-foot marker instead of the 2-foot marker on the cable.

Initial readings taken between March 2011 and June 2011 were obtained incorrectly. A proper initial reading was taken on 29 June 2011. Data obtained after June 2011 used correct methods. Readings obtained from March until June 2011 were modified to reference correct axes as a means to compare to readings taken post June 2011.

All data presented are referenced from June 2011. Readings are taken on a quarterly basis.

Inclinometer

10-EBD-I1

Exhibit H-2

Inclinometer 10-EBD-I1 was installed at Station 8+15, 30 feet downstream of the dam centerline at a total depth of 90 feet. Less than 0.1 inches of cumulative movement has been measured through May 2012. No significant movement is visible in the cumulative displacement plots. The minor movement is smaller than the accuracy limitations.

Inclinometer

10-EBD-I2

Exhibit H-3

Inclinometer 10-EBD-I2 was installed at Station 8+58.5, 11 feet upstream of the dam centerline at a total depth of 130 feet. Less than

0.1 inches of cumulative movement has been measured through May 2012. No significant movement is visible in the cumulative displacement plots. The minor movement is smaller than the accuracy limitations.

I – 5

3. WEIRS. There are eight weirs monitoring flows at the project. One of the IRRMs implemented includes improvements to monitoring and surveillance of the dam and its instrumentation. Weir 8a was installed in 2009 to replace Weir 8. In July 2011, USGS was contracted to automate Weirs 1 and 6 by gages and a remote reading acquisition system. Readings are taken every 15 minutes and uploaded to the USGS website.

Weirs 4, 5, and 7 measure flows from springs along the valley wall downstream from the left abutment of the dam. Weir 1 is located at the outlet of the spillway drainage gallery, and measures seepage from the drainage system behind the side walls and floor of the spillway. Weir 8a (replacement for Weir 8) measures seepage flowing down the hillside to the right of the stilling basin sidewall.

Weir 9 is located in the embankment toe drain to measure flow from the left abutment. Weir 6, located in a gage house at the downstream toe near the right abutment, measures collected seepage at the lowest point of the downstream toe. Weir 10 was installed to monitor a persistent wet area along the left valley wall, downslope of existing Weirs 4, 5, and 7. The weir measures the persistent flow from the hillside that accumulates along the service road.

Reading and evaluation frequencies were increased at all weirs in 2009 as a part of the monitoring and surveillance IRRM. Both automated and manual readings are now taken. Weir 1 and Weir 6 are automated and read on 15 minute intervals. All weirs, including Weir 1 and Weir 6, are read manually and evaluated at least twice per week except Weir 6 which is read once a day. Weirs 8 and 9 are read daily when the reservoir level is at or above elevation 1660. Weirs 1, 4, 5, 7, and 10 are read and evaluated daily when the reservoir level is at or above elevation 1670. The reading frequency of Weir 6 increases to twice a day when the reservoir level is at or above elevation 1670. In addition to taking weir readings, project personnel also visually inspect all weirs for discoloration and turbidity daily with Weir 6 being checked twice during each 8-hour shift. Any unusual variations are reported immediately to the Dam Safety Section.

Several weirs have historically exhibited relatively high flow rates, including Weir 1 (typically 300 to 700 gpm) and Weir 6 (typically 50 to 100 gpm). Exhibit H-9G shows the response of Weir 6 with pool but also the dramatic spikes over a short period of time associated with large rainfall events as well as snow melt and hillside drainage.

The exhibit also shows that with the interim operating pool (water control plan) in place, the lower reservoir levels (elevation 1623 to 1650) has decreased the typical flow for Weir 6 (20 to 75 gpm). This is also occurring in Weir 1 with flow typically 10 to 300 gpm.

I – 6

As noted in the previous inspection report, numerous occasions of cloudy water at Weir 10 have been reported. It was determined that this description was being applied to the condition of the water in the pool behind the weir, which has a considerable amount of soft sediment and is inhabited by frogs. Project personnel believe that the cloudiness in the pool is generally attributable to the movement of the frogs.

The same remarks were being made again in 2011 at Weir 10 and also noted by the staff regarding the pump house pond, just downstream of Weir 6. Investigation in conditions and flow rates at Weir 10 indicated that the cloudy nature of the water is due to the movement of frogs. In addition, the water at pump house pond was visited by Joe Premozic, engineer, and Rose Reilly, biologist, in late October 2011. Samples taken during the visit indicated that the iron-stained appearance on the eastern edge and grayish-green appearance over the rest of the pond surface may be a source of measured turbidity. Both of these have been present for many years.

As previously recommended, project personnel should note the water condition (clear, cloudy, or muddy) to describe the water entering the Weir 10 pool from the upstream ditch rather than the water in the pond to allow a more accurate characterization of the associated seepage. This characterization should be reported in the reading comments. Additional information is included in a memorandum included in Appendix J, Special Reports.

Weir 1

Exhibits

H-4A thru H-4C

Weir 1, located at the downstream end of the spillway channel, measures flow that varies directly with fluctuations in the pool. The flow at the outlet of the spillway drainage conduit gallery is collected from vertical 6-inch semicircular corrugated metal pipe drains placed against the excavated rock surface behind the spillway sidewall lining. A 4-foot diameter corrugated metal drainage conduit gallery at the base of each wall carries the flow to beneath the spillway bucket where the combined flows discharge over Weir 1. Near the upper end of the spillway, surface runoff collects at the base of an 8-foot high step in the spillway channel and drains through two 6-inch diameter holes into the drainage galleries. Flow typically rises to approximately 300 to 600 GPM during large rainfall events. Flow follows previously established relationships to the pool and, with the exceptions of 25 January 2010, 28 February 2011, 21 March 2011, and 4 April 2011, the flow has been clear. The few dates of cloudy flow coincide with high flows (300-500 gpm).

Since the automation of Weirs 1 and 6 in July 2011, weekly reports have been created comparing the flows to the pool elevation, precipitation and any other potential influences. It appears that Weir 1 is heavily influenced by the drainage from the hillside above the left spillway wall and tends to lag a few days following precipitation. Overall, Weir 1 correlates well with the reservoir level.

I – 7

During the periodic inspection, the two drain holes within the floor of the spillway channel just upstream of the sill at approximately Spillway Station 8+30 were cleared by project personnel and the standing water drained into the conduit galleries. A large spike in flow of 800 gpm was seen in Weir 1 readings during this period.

Project personnel have been notified to clean out as necessary and to inform District personnel upon completion of cleanout.

Flow briefly increased to 1329 gpm in June 1989 as noted in the 1992 periodic inspection. The flow was attributed to runoff from the upper end of the spillway.

Weirs 4, 5, 7

H-5A thru H-5C, H-6A thru H-6C, H-7A thru H-7C

Weirs 4, 5, and 7 all reflect reservoir seepage through the left abutment. Weirs 4 and 7 are located at springs while Weir 5 is located downstream in the access road ditch to measure collected flow from the left abutment ground surface. Weirs 4 and 7 were both repaired in August 1985 and have functioned well since. Correlation plots were developed for each of the weirs. The correlation plots for Weirs 4, 5 and 7 exhibits a bilinear relationship: weir flows increase at a relatively steep rate for the lower range in pool levels, above which, the weir flows increase at a slower rate. It is interpreted that there are geologic features (e.g., fractured bedrock) that have sufficient capacity at the lower pool levels to allow seepage to increase more freely. Once the capacity of the geologic features is reached (e.g., the seeping water encounters the top of the fractured bedrock), seepage is limited by the size of the features and the seepage can only increase at a much slower rate. To a lesser extent, this relationship is apparent at Weir 1.

Weirs 5, 7, and 10

H-6A thru H-6C, H-7A thru H-7C, H-8A thru H-8C

Correlation plots (weir gage reading versus pool) indicate a trend toward higher flows over time for a number of the left abutment weirs (i.e., Weirs 5, 7 and 10). These conditions have been observed in the instrumentation for some years, but have received renewed and heightened scrutiny as a result of the potential failure modes analysis (PFMA) and risk assessment of the dam conducted between January 2008 and December 2009. Annual reviews of correlation plots for all weirs are conducted and compared to data from previous years.

Weir 4

H-5A thru H-5C

Weir 4 was reportedly installed at the location of a natural spring that was observed prior to construction. With the IRRM reduced pool level, Weir 4 typically exhibits a flow rate of about 30 to 65 gpm during summer pool levels. Flow reduces when the pool drops below elevation 1635. The interim winter pool elevation is approximately 1623, so Weir 4 routinely stops in the winter until March when the pool rises toward summer elevation 1650. The flow increases rapidly with pool to 60 to 65 GPM. The weir flow – pool level relationship has remained consistent with that observed during the period of 1994 to 2007. It is possible that the flow paths (e.g., jointed rock) that direct seepage to Weir 4 were already “clean” prior to construction and that, accordingly, there is no erodible material (at least under the

I – 8 current range of reservoir levels) remaining within the seepage paths to Weir 4. It is interesting to note that Weir 4 appears to reach a maximum flow rate at about pool El. 1660, with little increase with additional pool loading.

Weir 5

H-6A thru H-6C

Weir 5 has shown an increase in flow on the order of 10 to 20 percent over the last 12+ years. The weir flow versus pool elevation relationship show the most change over the 2000 to 2007 period as compared to that observed during the period of 1994 to 1999: the weir flows were generally higher between 2000 and 2007 (as correlated with pool) than previously observed. Since the interim pool levels, the weir versus pool elevation relationship has returned to values close to those during the 1994 to 1999 period with no signs of further increasing flow rates.

The increase in flow versus pool relationship over the period of 2000 to 2007 likely indicated that there was less soil/rock material providing resistance to seepage along the existing seepage paths to Weir 5. This led to the conclusion that internal erosion was occurring in the dam embankment, foundation or abutment (or some combination of these). Note that the observed relative increases in flow have been observed at Weir 7 as well, but not at nearby Weir 4 (at the location of a spring that pre-dates construction and reservoir filling). It is possible that the flow paths (e.g., jointed rock) that direct seepage to Weir 4 were already “clean” prior to construction and that, accordingly, there is no erodible material (at least under the current range of reservoir levels) remaining within the seepage paths to Weir 4. Therefore, since readings suggest that no internal erosion was occurring within the portion of the embankment tributary to Weir 4, and given their nearness and similarity of behavior, it seemed more likely that the relative increases in flow at Weirs 5 and 7 was related to erosion within the foundation or abutment and not within the embankment. The previous instrumentation evaluation stated that if it was found to be true, then the expected rate of flow increase would diminish over time. The data recorded since 2007 in conjunction with the interim operating elevations appears to verify this relationship as the flow has not shown as much additional increase.

This is also verified in the comparison of the reservoir versus outflow plot. It appears that the outflow is slightly higher for pool elevations between 1642 and 1647.

With the IRRM reduced operating pool levels, Weir 5 typically exhibits a flow rate of about 80 gpm to 140 gpm when the pool level is above elevation 1640. The flow at Weir 5 approaches zero when the pool drops below about approximately elevation 1632. Cloudy flow was observed at Weir 5 in the fall of 2008, 2009, and 2011 and to lesser extent 2010, accompanied by relatively low flows. No instances of reported cloudy flow were accompanied by relatively high flows as mentioned in the previous (2007) instrumentation evaluation.

I – 9

Weir 7

H-7A thru H-7C

With the IRRM reduced pool level, Weir 7 typically exhibits a flow rate of about 10 gpm to 35 gpm during the summer months and at pool elevations above 1640. The flow essentially stops from August through March or when the pool drops below approximate elevation 1635. Prior to the IRRMs, typical flow rates ranged from 25 gpm to 70 gpm, except at times when the pool dropped below elevation 1647. Cloudy flow was observed at Weir 7 on 21 September 2009, in April 2010 and a few times in spring 2011. At all other times, the flow was described as clear. The weir flow – pool level relationship does not appear to have changed as compared to that observed during the period of 2000 to 2007: the weir flows at lower pool elevations are slightly lower than those during the previous performance period.

However, the weir flows were found to be generally higher (as correlated with pool) than what was observed over the 1994 to 1999 performance period. See the discussion of Weir 5 flows for an evaluation of the relative increases in flow at Weir 7.

Weir 10

H-8A thru H-8C

Weir 10 has been read since May 1994 and confirms flows are related to pool located along the base of the left valley wall adjacent to the service road, located at approximate Sta. 17+25, 1048 feet downstream of the axis. With IRRM reduced pool levels, flow rates vary with pool elevations from no flow up to approximately 5 to 30 gpm at the new interim summer pool of 1650. Typical flows prior to the interim summer pool were 40 to 50 gpm at normal summer pool (1670).

The weir flow versus pool elevation relationship shows an increase in the weir readings over the period of 2008 through 2012 despite the lower pool elevation which is shown in Exhibit H-8B. The increase at Weir 10 has been as much as 50 percent for reservoir levels between El 1650 and El 1670. Almost all readings were taken below a pool elevation of 1650 and show a trend of slightly increasing gage readings. The weir flow versus pool level relationship changed during the previous periodic inspection period of performance as compared to that observed during the period of 1994 to 1999: the weir flows were generally higher (as correlated with pool) than previously observed. This trend seems to be ongoing.

Cloudy flows were observed at Weir 10 in March through August 2008, October and November 2009, April through October 2010, and March through November 2011, with no typical correlation in flow.

Similar observations were made in May to July 2001, April 2002, April 2004, and in the early months of 2006 and 2007, typically accompanied by high flows. During field inspection made prior to the 2007 periodic inspection, it was determined that the source of the cloudiness in the pool behind the weir (at least for a large number of the observations) was the presence of frogs living in the pool.

Observations regarding the clarity of the Weir 10 flows should be based on the water entering the weir pool.

I – 10

During a teleconference in December 2010, project personnel were concerned regarding the cloudy nature of Weir 10 pond and the possibility of it relating to the recent exploratory drilling. No obvious accumulation of sediment or muddy discharge was observed and no changes in weir flow rates were apparent. The project personnel were asked to make a distinction in their remarks between clear/cloudy/muddy observations of the pond and the inflow. During Water Management’s visit to the site in September 2011, a water sample was collected from the pond at Weir 10 which had a distinct green coloration. The test results indicated it as algae. An additional site visit was made the same month and conditions of Weir 10 were observed. Weir 10 had a cloudy appearance; however the very small flow entering the pond was clear and did not appear to be moving any sediment.

Weir 6

H-9A thru H-9E

Weir 6 measures flow from a pond at the downstream toe. At this location, the weir is influenced by surface runoff as well as foundation or embankment seepage. Part of the flow at this weir fluctuates directly with changes in pool elevation. This can be seen on plots of time vs. flow as a base, or minimum, flow on which are superimposed transient peaks due to surface runoff. This base flow is apparently due to seepage from the reservoir. The seepage related flow historically has been 100 to 120 GPM at summer pool and 20 to 50 GPM at winter pool. With the IRRM reduced pool level, flow rates are approximately 50 – 100 gpm during summer pool and 20 to 50 gpm during winter pool. Measurements between 2003 and 2007 match closely with trends observed in the 1994 through 1999 data.

This is consistent with the data evaluated between 2008 and 2012, although on the lower end of the range due to the pool restrictions.

Cloudy flow was observed at Weir 6 in November 2011, accompanied by a sharp peak in flows. At all other times, the flow was described as clear.

Since the automation of Weirs 1 and 6 in July 2011, weekly reports have been created comparing the flows to the pool elevation, precipitation and any other potential influences. Weir 6 reacts quickly to the rising and lowering of the pool, but especially to precipitation or melting snow. Overall, Weirs 6 correlates well with the reservoir level.

On 5 November 2010, project personnel observed a transient change in flow measured at Weir 6. The second and third shifts of the 24-hour monitoring cycle, which is one of the IRRMs, first observed and reported the change in flow. Over a period of four hours, the flow at the weir dropped from consistent readings of 0.18 gpm to below the V-notch weir: a reduction in flow of approximately 15.4 gpm. Flow remained below the v-notch throughout the evening and early morning, approximately 13 hours, at which time readings started increasing. On 6 November 2010 district office engineers traveled to

I – 11 the site and observed the embankment slopes and abutments. Select piezometers and weirs were read and observed with no unusual seepage or movements observed. All other instrumentation readings were within the historical range for the current pool elevation.

Levels in the ponds at and below weir 6 were visibly lower than normal, but no leakage was evident. After an additional 14 hours, on 6 November 2010 at 2200 hours, readings became stable and returned to expected values. After much evaluation, it was determined that the recent drilling program had no effect on the seepage. It is believed that a dynamic change in foundation seepage conditions, the closing and opening of seepage pathways, caused the transient drop in flow.

Further detail can be found in the memorandum for record (MFR) included in Appendix J.

Weir 9

H-10A thru H-10B

Weir 9 was placed in the embankment toe drain near the bottom of the left valley wall in an attempt to separate left abutment flow from the measurements of Weir 6. The flow is unrelated to pool but fluctuates widely in response to precipitation and date of reading as noted in the February 2009 major increase to approximately150 gpm and March 2011 major increase to approximately 300gpm. The background flow rate is 1.0 GPM or less during dry weather periods.

Weir 8a (replaced Weir 8)

H-11A thru H-11B

Weir 8 measured flow from the right abutment near the tunnel outlet.

Flow was consistent over the life of the instrument. The approach channel to the weir was plagued with debris and talus due to its location above the right training wall at the base of the steep slope.

In addition, the approach channel was damaged such that water that would have been tributary to the weir flowed over the training wall upstream of the weir, thereby bypassing the weir. Weir 8 was relocated in Spring 2009, under a contract separate from the geotechnical exploration, to a location adjoining the tunnel end wall, where it is more easily accessible. The relocated weir is denoted as Weir 8a. Measurements do not show a relationship between pool level and seepage. Flow was normally less than 10 gpm at its former location with higher transient peaks due to surface runoff. Since its relocation, the weir receives seepage from a smaller area and therefore flow is reduced to2 to 6 gpm. Reviews of the readings indicate no new problems or concerns and are generally consistent with similar data from previous years.

I – 12

3. PIEZOMETERS

Summary

During the first year of the evaluation period (2008), the majority of piezometers showed no change in pore pressure response to pool levels. Since the implementation of the restricted interim pool levels, numerous piezometers have been affected by the reduced reservoir levels. Piezometers are generally showing three behaviors: a gradual reduction in levels over the past four years; a reduction in the readings during summer months; or no change in levels at all.

A fair amount of piezometers, mainly located in the right and left abutments, read above the pool level especially with the interim reservoir levels – interim summer reservoir elevation is 1650 and interim winter reservoir elevation is 1623.

Exceptions:

Piezometer Condition/Trend

77-EBD-25

90-EBD-39

90-EBD-45

Possible long-term shift towards higher readings (no discerned correlation with pool levels).

90-EBD-32

Readings rise 2 feet while pool drops from summer reservoir level to winter reservoir level.

This is evident over its lifetime.

72-EBD-1 Historically reads dry. Since interim reduced reservoir levels, reading 1-2 feet above tip.

a. Background

A total of 111 piezometers have been installed in the dam in 1957, 1972, 1977, 1990, 1997, 1999, 2008 and 2010. Piezometers are located in the embankment, foundation, and abutments. Piezometers 1 through 18 were installed in 1957 at the conclusion of the emergency grouting in two lines at approximately Sta. 8+30 and Sta.

9+30. All eighteen of the 1957 piezometers have been abandoned and grouted – No. 18 in 1987 and the remaining 17 in 1992.

Piezometers 72-EBD-1 through 10 (including 72-EBD-2A and 72- EBD-2B) and 72-EBR-1 through 8 were installed in 1972 in a line across the dam at Sta. 11+05, which is in the vicinity of the original river channel, as well as along the downstream toe and along the spillway.

Piezometers 77-EBD-11 through 28 were installed in the left abutment and along two lines across the dam at Sta. 8+30 and Sta.

14+00. Piezometer 77-EBD-29 was installed along the outflow channel at Sta. 12+60.

Piezometers 90-EBD-30 through 90-EBD-50 were installed in lines at approximate Sta. 7+30, Sta. 8+25, Sta. 9+15, and Sta. 11+15.

Seven piezometers were installed in the left abutment in late 1997

I – 13

(Piezometers 97-EBD-51 through 97-EBD-57).

Seven piezometers (99-EBD-58 through 99-EBD-64) were installed at various points between approximate Sta. 9+30 to Sta. 11+05 to replace a number of the older piezometers.

Nineteen new piezometers were installed following the previous periodic inspection in 2007. Seventeen piezometers 08-EBD-64 though 08-EBD-81 were installed in 2008 in areas near the contact between the dam and the right hillside on the downstream face of the central and left portions of the dam. These areas were selected after three-dimensional mapping of seepage paths through and beneath the right portion of the dam using geophysical methods was conducted.

The piezometers installed in the area of the grouted void, 08-EBD-65 through 08-EBD-70, were unsampled; however, sonic core borings were completed offset few feet upstream from each piezometer. Two additional piezometers 10-EBD-82 and 10-EBD-83 were installed in 2010 simultaneously with the inclinometers to provide additional monitoring of the area near the 1957 grouted void, a major risk area within the dam.

All piezometers are the open-standpipe type, except for 72-EBD-2B, which is read by a pressure gage on the standpipe. Piezometer 72- EBR-3 became inoperative in October 1994 and was abandoned.

Piezometer 90-EBD-44 apparently filled with surface water in February 1996 and is considered inoperative. Readings ceased at 90- EBD-44 after December 1999. Only 86 piezometers remain in use after the abandonment of the eighteen piezometers installed in 1957;

abandonment of six piezometers (72-EBR-3 through 72-BER-8) installed in 1972; and the discontinuing of piezometer 90-EBD-44 in 2000.

I – 14

The frequency of piezometers readings have increased since the adapted interim risk reduction measures. Improvements to monitoring and surveillance at the dam and instrumentation have provided additional evaluation and better understanding of piezometers and their behavior. Prior to initiation of the IRRMs in 2008, piezometers were read once per month until the reservoir level achieved elevation 1675. Readings were made twice per week at elevation 1675 or above. During the interim water control plan, all piezometers are read once per month except those at Stations 7+30, 8+30, 9+30, and the toe which are read twice a month. Reading frequencies increase to once per week depending on stationing and pool elevation: when the pool reaches El.1655 for piezometers located near Station 8+30; when pool level reaches 1660 for piezometers at Stations 7+30 and 9+30 and the toe; and when pool level reaches El. 1670 for all remaining piezometers. Readings are taken daily starting at pool elevation 1670 for piezometers at Stations 7+30, 8+30, 9+30 and at the toe. All remaining piezometers are read daily at pool elevation 1675 and above. A table of reading frequencies is provided below.

Frequency of Readings Reading Read Locations 1 1 1 2 1 2 1

Type By Per Per Per Per Per Per Per Year 2 Months Month Month Week Week Day

Piezometers Project All X (1) 1670 1675

(3) Sta 8+30

X (1) 1655 1670

(3) Sta 7+30 X (1) 1660 1670

(3) Sta 9+30 X (1) 1660 1670

(3) Toe X (1) 1660 1670

Notes:

X – Indicates minimum reading frequency

(1) Project personnel to adjust reading date as necessary for safety concerns while maintaining frequency at reasonable intervals. Not to wait until last opportunity if bad weather is anticipated.

I – 15

To establish consistent visual monitoring, an IRRM was established to have the project staffed 24 hours a day. This is achieved by three 8 hour shifts. Project personnel read certain piezometers every shift at key locations, such as the grouted void, left abutment, and downstream toe, that pose the biggest risk of seepage and potentially will indicate seepage first. The piezometers read during each shift are provided in the table below. In addition to piezometers, the pool and outflow gages are also recorded.

Piezometers read every shift

72-EBD-01 79-EBD-13 99-EBD-64 90-EBD-38 10-EBD-82

72-EBD-02 79-EBD-14 90-EBD-50 10-EBD-83

72-EBD-2b 79-EBD-15

72-EBD-03 79-EBD-16

79-EBD-25

79-EBD-26

An additional nine open standpipe piezometers are to be installed over the summer of 2012 on the downstream face of the dam, near the crest, and at the toe of the dam. Select open standpipe piezometers will be automated with vibrating wire pressure transducers to gather information prior to the start of construction of the cutoff wall planned for FY 2013. These automated piezometers will take very frequent readings and be evaluated daily during construction to help monitor water pressures in the dam. It is planned to keep the automated instruments in place after construction to aid in monitoring over the life of the dam.

I – 16

As a part of the Geotechnical Exploration conducted in 2008 and 2009, contractor, D’Appolonia, tested some piezometers (see table below) to help understand and determine their functionality. The plots produced for this evaluation period show spikes in readings during October 2011 and a four month lag it took to return term to expected values This spike is apparent for all 21 piezometer tested and not included in the evaluation of the behavior of the piezometers.

Piezometers Tested by D’Appolonia

Piezometer Stationing Offset Sensing Material 72-EBD-1 9+70.69 R415.28 foundation rock

77-EBD-25 8+35.02 R46.25 embankment impervious fill 77-EBD-26 8+35.02 R46.15 foundation overburden 90-EBD-30 7+33.97 L49.31 embankment random fill 90-EBD-32 7+38.63 R10.46 embankment impervious fill 90-EBD-36 8+21.53 R9.78 embankment impervious fill 90-EBD-37 8+28.16 R10.45 embankment impervious fill 90-EBD-39 8+30.15 R79.19 embankment random fill 90-EBD-40 9+18.59 L62.75 embankment impervious fill 90-EBD-42 9+06.24 R10.16 embankment impervious fill 90-EBD-43 9+11.78 R10.21 embankment impervious fill 90-EBD-45 9+14.86 R60.65 embankment random fill 90-EBD-46 9+19.29 R60.26 embankment impervious fill 90-EBD-47 9+25.54 R54.75 foundation rock 90-EBD-48 9+20.06 R101.26 embankment random fill 90-EBD-50 11+14.06 R10.51 embankment impervious fill 99-EBD-59 11+02.79 R41.85 foundation overburden 99-EBD-61 10+99.67 R166.98 rock drain 99-EBD-63 10+95.34 R294.92 rock drain

Two rounds of tests were performed: The first round occurred on 23 and 24 October. Each piezometer was filled to the top of the riser with water and purged with compressed nitrogen introduced through tubing to the bottom of the piezometer. The post-purging water level and sounded depth were obtained and recorded.

The second round occurred on 6 and 7 November. Each piezometer was then purged and subjected to a rising and/or falling head permeability test. Purging involved either using compressed nitrogen if they exhibited higher water levels or flushing with water if they exhibited lower water levels. Most piezometers purged with nitrogen were monitored for water level recovery (rising head). The remaining piezometers were monitored for dropping water levels (falling head).

See Appendix I for the data tables regarding these tests.

I – 17

In May 2012, USACE district engineers conducted field falling head tests on 19 piezometers. See the table below for a list.

Piezometers Tested by USACE

Piezometer Stationing Offset Sensing Material 72-EBD-8 22+09.78 R93.96 foundation rock 72-EBD-9 21+58.85 R669.29 embankment impervious fill

72-EBD-10 20+87.88 R909.10 foundation overburden 77-EBD-17 14+03.48 L9.73 embankment random fill 77-EBD-21 14+03.28 R221.81 embankment impervious fill 90-EBD-35 8+26.06 L29.13 embankment impervious fill 97-EBD-56 23+34.84 L89.13 embankment impervious fill 99-EBD-58 11+12.55 R41.93 embankment random fill 99-EBD-60 11+10.70 R166.95 embankment impervious fill 99-EBD-62 11+03.16 R293.13 embankment impervious fill 08-EBD-66 9+83.55 R11.57 embankment impervious fill 08-EBD-67 10+71.15 R11.64 embankment random fill 08-EBD-68 11+57.72 R11.36 embankment impervious fill 08-EBD-69 12+59.63 R11.89 foundation rock 08-EBD-75 12+59.05 R133.10 embankment random fill 08-EBD-77 12+60.77 R232.53 embankment impervious fill 08-EBD-79 17+59.17 R132.29 foundation overburden

Testing involved filling the riser with 5 to 10 gallons of water and recording the height of the water level as it dropped over a period of time. The amount of water introduced into the piezometer riser varied depending upon the ability of water to rise. The water levels in piezometer X, X, X, and X could not riser to taking readings so no test was performed.

A memorandum and test data sheets are provided in Appendix I.

Piezometer locations in plan and sections are shown on Plates 1-7.

Plots of piezometer levels at various stations are included as exhibits.

I – 18

b. Right Abutment

The 1957 emergency grouting took place near the right abutment which is where a considerable number of the piezometers are located.

The cavity in the impervious core was approximately centered at Sta.

8+30. Most right abutment piezometers are centered on lines at approximately Sta. 7+30, Sta. 8+30, and Sta. 9+30 with sensing zones located in the embankment and foundation. Three additional piezometers, one in 2008 at Station 9+80 and two in 2010 at Stations 8+60 and 8+68, were installed within the embankment impervious fill to supplement existing piezometers and establish sensing zones at particular depths. Another three piezometers were installed in 2008 near Station 6+50 upstream and downstream of the centerline of the dam. The sensing zones for these piezometers were located in the foundation rock of the right abutment which tends to be highly permeable. They show a high seepage gradient between the impervious core and the foundation in the cavity area. Estimated piezometric surfaces and equipotentials sketch can be seen in the 1992 Periodic Inspection Report Exhibit H-13. Exhibits H-12 through H-28 of this report shows estimated piezometric surfaces for the right abutment stations (Sta. 6+50, 7+30, 8+30, and 9+30).

The piezometers between Station 7+50 and 9+50 reveal high piezometric levels in the embankment and reflect the pervious nature of the foundation. In general, the piezometers show influence by the pool in the impervious core upstream of the zone of consolidation grouting and in the foundation.

(1) Station 6+50

Exhibit H-12

Piezometers 08-BED-71 through 08-EBD-73 were installed at the right abutment in the foundation rock in 2008 to further evaluate the phreatic surface in this area.

Over the three years following installation of these piezometers, 08- EBD-71 has shown a direct and tight correlation with the pool. This behavior is expected due to its location 100 feet upstream of the dam centerline and within the very broken and jointed foundation rock at the right abutment. This is a highly permeable zone and poses a significant avenue for seepage around and into the embankment in conjunction with groundwater.

Piezometers 08-EBD-72 and 08-EBD-73 have a slight response to the pool elevation. Also as expected, piezometer 08-EBD-72 reflects readings lower than the pool due to its location 15.5 feet downstream of the dam centerline. Piezometer 08-EBD-73, located 58.8 feet downstream of the dam centerline, reads below both the pool and piezometer 08-EBD-72 elevation. The lack of response in these piezometers implies that the grout curtain installed during construction is effective at interrupting seepage paths through fractures in the rock. However, the actual effectiveness of the grout curtain cannot be stated with certainty from this data alone.

During installation in November 2008, the sand pack and bentonite pellets placed in piezometers 08-EBD-72 and -73 washed out during

I – 19 installation. The contractor suggested in their report that longer term monitoring of these piezometers is necessary to bettered understand the seasonal variations within the right abutment seepage regime as well as the combined effects of hillside seepage and reservoir pool level on piezometric levels and gradients through the right abutment.

The water levels in 08-EBD-72 and 73 rose during the 2011 to 2012 geotechnical exploration and testing program when boring E-39 was drilled and pressure tested. A total of 26 pressure tests were completed in the boring. It appears that the drilling possibly introduced or widened an existing joint(s) supplying water to those two piezometers. After completion of testing/imaging in boring E- 39, the borehole was backfilled in stages from, starting in mid-March, ending on March 23, which resulted in the somewhat precipitous drop. The grout place in the boring was 693% above theoretical volume. A portion of the grout was washed out due to audible running water in the bore hole. There is a possibility that grout-backfilling of E-39 probably resulted in filling-in some of the joints (that originally fed the water levels in P-72 and P-73) and potentially permanent change in the piezometer level.

(2) Station 7+30

H-13 thru H-15

Piezometers 90-EBD-30 through 90-EBD-34 were installed near this section to measure the influence of natural groundwater in the right abutment. Piezometers were installed at Sta. 7+30 to be closer to the right valley wall, the suspected source of the influence.

Random fill piezometer 90-EBD-30 shows a seasonal fluctuation at levels above the pool. However, in early 2009 the inner pvc broke eleven feet down from the top of the protective casing, therefore no readings were taken between January 2009 and March 2011. Project personnel secured the upper broken section to the lower intact pvc pipe to maintain a functioning piezometer. When comparing the limited readings for the period between 2008 and 2012 to that of the previous performance period, values appear to be within the appropriate range for this piezometer. It is difficult to determine over the short period if the readings that were taken show a slight response to pool changes as it has in the past.

Foundation piezometer 90-EBD-31 reveals a direct relationship to the pool elevation; albeit, sometimes at elevations higher than the pool level. The piezometer follows pool fairly consistently except when the pool drops below an elevation of 1640. At this level, the piezometer stops following pool and maintains reading elevation 1640. There is one instance in April 2010 where the piezometer ready “dry” at the tip elevation of 1637.88.

Piezometer 90-EBD-32 is located in the impervious core with its tip at elevation 1680, 10 feet above the normal summer pool. Prior to the implementation of the interim water control plan with summer pool of 1650 (20 feet lower than the normal summer pool), the piezometer showed a seasonal fluctuation opposite to the pool

I – 20 fluctuations and at levels well above the pool. The piezometer readings fluctuate slightly around elevations 1680 to 1683 despite the lowered interim pool level. Pool elevation appears to have no influence on pressures read at the piezometer.

Random fill Piezometer 90-EBD-33 and foundation Piezometer 90- EBD-34 both reflect no relationship to the pool. Both piezometers have tip elevations substantially higher than pool. Neither obtains readings and both remain “dry.” Piezometers 90-EBD-33 and 90- EBD-34 continue to read at elevations 1657.3 and 1644, respectively at their tip elevation.

Groundwater influence in the right abutment is substantial, as indicated by the behavior of piezometers at Sta. 7+30.

(3) Piezometer

08-EBD-65

Exhibit H-16

Piezometer 08-EBD-65 was installed within the random impervious fill at Station 7+83 and located on the outskirts of the area of the grouted void. It was installed after the implementation of the interim operation pool levels during a geotechnical exploration contract. The bore hole used to house the open standpipe was not sampled and some grout loss occurred during piezometer installation. A sampled sonic core boring, 08-EBD-65L, was performed approximately 6 feet upstream. No material was encountered that would suggest a venue for grout loss. The report submitted after the drilling program completed assumed that the additional grout takes were being lost into the coarser zones near the top of the borings.

The piezometer is located 11.5 feet downstream of the centerline of the dam with a tip elevation of 1619, at a depth of 90 feet from the ground surface. It shows a very slight response to changes in pool levels. During summer pool levels it reads between elevations 1625 and 1628, except on 31 March 20111 when the piezometer spiked to an elevation of 1635. When the pool lowers to winter levels, the piezometer reads around an elevation of 1622.

(4) Station 8+30

Exhibits

H-16 thru 22

This station is the location of the 1957 grouted cavity and it is heavily instrumented with piezometers. There are currently six 1977 piezometers, five 1990 piezometers, and two 2010 piezometers near Sta. 8+30. The nine 1957 piezometers (Nos. 1 through 9), which provided data through 1992, were replaced by the more recent installations. The old instruments had questionable reliability and were abandoned and grouted in 1992.

The geotechnical exploration in 2008 and 2009 encountered zones of wet, loose rock fragments adjacent to competent grout in the area of the 1957 grouted cavity. The report stated that these zones are close to the inferred surface of the construction access road at the right abutment. The report’s finding indicates that “the grouted zone and the consolidation grouting might be creating a barrier effect wherein the convergence of the seepage from the right abutment/hillside and through the dam results in unusually elevated seepage horizons or phreatic levels in this confined area.”

I – 21

Offset 30 feet US Piezometers 77-EBD-23, 77-EBD-24, and 90-EBD-35 are all located on the upstream face of the dam. These piezometers are not read during the winter months, typically November through March, when the riprap on the upstream face is covered in snow and ice making access dangerous for project personnel.

Piezometers 90-EBD-35 and 77-EBD-23, both located upstream in the impervious fill, appear to respond well to pool fluctuations prior to the new interim pool elevations. Since the implementation of the interim water control plan, piezometer 77-EBD-23 continues to appear to closely follow the pool changes. However, data from winter pool elevations are not obtained due to the dangerous conditions. It is difficult to determine how the piezometer will respond during those months. In addition, readings in piezometer 77- EBD-23 were “dry” in October and November 2011.

Piezometer 90-EBD-35 appeared to respond closely to pool levels prior to the interim water control plan. Since the pool level dropped below 1660, the fluctuation with the pool has become more subtle. In years prior to 2008, the readings were essentially at the same elevation as the pool and now it fluctuates between elevations 1651 and 1656. One exception occurred on 14 August 2010 when the piezometer read an elevation of 1662. In May 2012, a field falling head test was conducted at piezometer 90-EBD-35. Initially the rate of head difference was much faster than expected.