ENGINEERING INSPECTION REPORT
Bend Hydroelectric Project Timber Crib Dam Spillway Inspection
December 10, 2013
Roger Raeburn, P.E – Chief Dam Safety Engineer
Nathan Higa, P.E. – Senior Engineer
On October 1, 2013 a change to the leakage pattern in the timber crib dam spillway structure at the Bend Hydroelectric Project (Project) was reported by the Bend plant operator at Needle Bay 11. The new leakage pattern appeared very similar in nature to events experienced in 2008 and 2009 which resulted in remedial action. The cause of the leakage was attributed to failure of deteriorated facing boards on the upstream side of the dam in the vicinity of the original needle bay.
The increase in leakage through the timber crib dam, combined with the seasonal decrease in releases from Wickiup Reservoir into the Deschutes River system, as a result of cessation of the irrigation season, resulted in the inability to maintain the water surface elevation behind the dam at the normal operating level. The result is the reduction in the water surface elevation of Mirror Pond by approximately two feet. There is no public safety issue associated with the recent change to the condition of the dam.
This leakage event and the associated change to the operating condition prompted an engineering inspection. This report provides documentation of an engineering inspection conducted for the water retaining structures associated with the Project, with an emphasis on the timber crib dam spillway.
The objective of the inspection was to ascertain the condition of the water retaining structures, and in particular the timber crib’s structural members, rockfill, and foundation connection to the extent possible, given the limited access to the timber crib’s downstream toe, face and interior due to a limited ability to drawdown the reservoir behind the dam. To fully isolate the timber crib would have required placing fill in the river, which would require additional permitting. The factors associated with the timber crib’s condition are considered to be the most influential in the development of judgments and conclusions regarding the Project’s overall condition and suitability for continued safe operation.
The Bend Hydroelectric Project (Project) is located within the city of Bend, Deschutes County, Oregon on the Deschutes River, at river mile 160. The impounded water creates Mirror Pond, which is surrounded, in part, by Drake Park. The Project operates on a non-expiring water right of 1325 cfs for power production with a priority date of 1905.
The Project was constructed and placed in service in 1916. The Project’s water retaining structures have remained essentially unchanged since that time. The generating assets and water retaining structures are approaching 100 years of operation and their condition reflects their age. While still operable if upgrades are performed, the Project is nearing the end of its expected service life.
The Bend Project is currently on a tri-annual inspection schedule by the Oregon Water Resource Department (OWRD) Dam Safety Section. The last inspection by OWRD was conducted July 13, 2012. Recommendations contained in the OWRD’s October 16, 2012 report (Attachment 1) included:
- Continue with good maintenance and operations, including security and vegetation control.
- Evaluate Deschutes River flow, and accompany OWRD on an inspection of the spillway structure at very low water. The timing of this inspection would be coordinated by OWRD and the local watermaster for the Deschutes River.
The OWRD inspection reports include an assessment of the condition of a project’s water retaining structures. The rating system utilizes five categories: 5 – Very good, 4 – Adequate, 3 – Maintenance or minor repair needed, 2 – Serious repair needed and 1- Urgent dam safety issue. The 2012 report identified no category 1 or 2 items, two project features rated category 3, and all remaining features rated category 4 or 5. The two category 3 features were:
- Wooden boards at the toe of the timber crib dam section.
- Gate leakage through the regulating outlet sluiceway. The next regularly scheduled inspection is due in 2015.
The water retaining structures associated with the Bend Hydroelectric Project comprise several components as depicted in Figures No. 1 and 2 below.
Timber Crib Dam Spillway – The timber crib dam spillway (timber crib) is comprised of square log timbers placed side by side and stacked to form open box sections that were then filled with large basalt rock. The design of the timber crib initially included 21 crib sections (bays), of which 14 were configured as “needle bays” incorporating vertical wood needles that allowed for the controlled passage of water directly through the timber crib by the adjustment of the needles. The timber crib section was later modified with the addition of a concrete sill on top of the crib that allowed for water to spill over the top of the dam and isolated the needles, rendering them non-functional. Following the timber crib modification, the release of water was accomplished over the top of the timber crib and controlled with wooden stoplogs that remain in service today.
Thirteen of the original needle bays remain exposed, with the 14th upstream-most partially covered with riprap from the abutment. The seven other bays were configured without needles, but do include stoplogs and are capable of spilling water over the timber crib.
Ice/Debris Sluiceway – The ice/debris sluiceway is located at the downstream abutment of the timber crib. It is constructed of grouted basalt rock and incorporates wood stoplogs that can be pulled to allow the passage of water, ice, and debris. The ice/debris sluiceway has not been used for a considerable time.
Non-Overflow Dam Section – The non-overflow dam section is located between the sluiceway and the powerhouse. It is comprised of a buttressed concrete wall approximately 12 feet high and founded on basalt rock. This section of the dam is considered a non-overflow section, with no provision for the passage of water.
Powerhouse – The powerhouse consists of a concrete foundation with a brick masonry superstructure and truss roof. It contains three James Leffel horizontal Francis turbine generator units operating at a gross head of 14 feet with a combined capacity of 1,110 kW.
Regulating Sluiceway Outlets – Two low-level sluiceway outlets are located to the right of the powerhouse. The original structure was designed with wooden stoplogs to control the release of water. The original stoplogs have been modified with the addition of two 4-foot wide by 5-foot high slide gates operated by an electrically powered manually operated gate actuator.
In response to the report of increased leakage through timber crib Bay 11, an engineer from Hydro Resources familiar with the 2008 and 2009 leakage mitigation work conducted an initial site visit to the Project on Friday, October 4, 2013. The site visit was conducted to assess the potential for the development of any dam safety concerns, observe the new leakage event, and assist with the preparation of plans for an engineering investigation/inspection of the timber crib.
At the time of the site visit the reservoir level staff gage reading was 16.3-foot, which is within 0.2-foot of the normal maximum level of 16.5-foot. The spillway was traversed back and forth, with photographs taken of each needle bay. Needle Bays 1 through 4 had no noticeable leakage flow. Needle Bay 5 had spillway overflow falling directly into the outlet, but it did not appear to be leaking. Needle Bay 6 had minor leakage flow. Needle Bay 7 and 8 had spillway overflow falling directly into the outlet, but did not appear to be leaking. Needle Bays 9 and 10 had a very small amount of leakage flow. The leakage flow through Needle Bay 11 was significant, with sufficient water being released to result in flow discharging through the timber crib above the needle bay. It appeared that some of the timber facing boards on the downstream toe of the dam had been dislodged due to the leakage flow. A 12-inch diameter vortex was noted on the upstream side of the dam directly in front of Needle Bay 11. Needle Bay 12 (upgraded with sheet pile in 2009) had no discernible leakage flow visible. Needle Bay 13 (upgraded with sheet pile in 2008) had a minor amount of leakage flow. Needle Bay 14 is partially covered with riprap materials from abutment erosion.
The performance of the dam since its construction has been satisfactory and the lack of any significant visual evidence indicating deformation of the timber crib structure makes it likely that the internal components are in satisfactory condition. After the initial site visit, it was concluded that confirmation of assumptions relative to the condition of the timber crib’s internal components would be performed. Actions identified include:
- Perform a drawdown of the reservoir (Mirror Pond) to minimize leakage through the timber crib.
- Conduct a visual structural inspection of the downstream face, needle bays, internal components and the dam’s foundation.
- Develop a monitoring and surveillance program for the timber crib
- Install horizontal and vertical alignment monuments on the timber crib,
- Conduct a baseline alignment survey of the timber crib,
- Establish an alignment survey schedule to monitor for any changes,
- Develop a visual inspection check list for water retaining structures,
- Initiate a monthly inspection of all water retaining structures by the Bend Operator,
- Perform a visual inspection of all water retaining structures by a PacifiCorp Energy Dam Safety Engineer on a bi-annual frequency.
Drawdown & Refill
The drawdown process began at approximately 7:00 AM on October 28, 2013 at a rate of two inches per hour. Water released from Wickiup Reservoir was reportedly 25 cubic feet per second during the drawdown period, with flows at the Project estimated to be 500 cubic feet per second. On October 31, 2013 (the day of the inspection), the level of the reservoir as measured on the temporary staff gauge was 9.56 feet, approximately 7 feet below the normal full operating level at 16.5 feet (Figure 3 & 4). Refill of Mirror Pond began at approximately 7:00 AM on November 1, 2013, at a rate of four inches per hour. The reservoir was restored to the starting elevation by the evening of November 2, 2013. All ramping was performed during daylight hours by manual operation of regulating outlet sluice gate aperture at one-hour intervals. The operator responsible for ramping procedures made conservative changes to the regulating outlet sluice gate aperture, and was thus able to maintain compliance with the prescribed ramp rates. All drawdown activities were performed in accordance with a plan approved by the Oregon Department of Environmental Quality (ODEQ) and in coordination with the Oregon Department of Fish and Wildlife (ODFW).
Engineering Investigation & Inspection
Design information and construction record drawings for the timber crib are limited. In preparation for the detailed engineering inspection, PacifiCorp Energy engineering personnel conducted investigative work into the history of the dam’s construction and modification. While the exact timing of various modifications to the timber crib remain obscure, it was clear that the available drawings do represent what is currently in place. Figure 5 represents a depiction of the timber crib, with descriptors for the various elements of the structure used in this report.
The investigation conducted into the dam’s history also revealed additional information about the dam’s history that was previously not readily available. The present concrete spillway sill was constructed in approximately 1919, earlier than previously thought. Bits and pieces of evidence were identified that would indicate the upstream and downstream facing boards on the dam have been replaced more than once in the dam’s history. No record of the last replacement of downstream facing boards was found, and the most recent replacement of upstream facing boards was completed in the early 1990’s on a selected basis. The access walkway was most recently replaced in 1975.
No record of leakage through the timber crib of the type experienced in 2008, 2009 or October 2013 was found in the records (Figure 6 & 7). It would appear from available records that previous work to address leakage was implemented to address general leakage issues through the replacement of upstream facing boards with various materials.
The work completed to address the leakage in 2008 and 2009 consisted of the installation of
sheet piles across the upstream face of the dam, immediately in front of the leaking needle bay. A drawdown of approximately 3 feet was required to allow the cutting of an upstream ledger board, allowing the installation of a wide flange steel section at each end of the sheet pile section, tying them to the dam.
Inspection of Needle Bays 6, 12 and 13 confirmed that the installation of the sheet pile was a viable method for controlling the leakage through the needle bays. The effectiveness of the sheet pile in controlling the leakage varied in each case, from virtually no leakage to leakage judged to be less than one cubic foot per second (Figures 8 & 9).
Inspection of Timber Crib Spillway
A detailed inspection of the timber crib was conducted on October 31 and November 1, 2013, by this report’s authors. The Hydro North safety administrator assisted with the logistics of access and provided a safety watch during the engineering inspection activities. The OWRD Dam Safety Engineer participated in the inspection for a couple of hours. An overall visual assessment of the timber crib was completed as part of the inspection with specific attention given to the 14 needle bays. The needle bays are considered to be the design section most vulnerable to damage/deterioration, and associated directly with the last three significant leakage events.
In addition to the visual inspection and documentation of conditions with photographs, methods involving the use of an impact drill/driver, 18-inch straight blade screw driver and 3-pound hammer were used to assess the condition of wooden members and connectors. Holes were drilled through wooden facing boards and into timber crib members using the drill/driver and a 6-inch long 0.375-inch diameter lathe bit. The condition of the steel drift pins and foundation bolts used to connect the 12-inch by 12-inch timber crib members, and the 60d nails used to attach the facing boards were checked using the 3-pound hammer. When accessible, the ends of drift pins and the heads of foundation bolt heads were struck with the hammer in an effort to judge their condition as well as their effectiveness in continuing to provide a credible connection. The hammer was also used to drive selected exposed 60d nails into the timber crib members and through facing boards. The ease with which the nails were driven was judged and the extent to which they would pull separated boards together and hold was observed. The 18-inch screwdriver was used as a probe and pry bar to assist with the evaluation of a wood member’s condition and to a limited extent, the condition of the rock fill.
Inspectors accessed the concrete spillway sill by climbing through the space created by flipping the hinged 2-inch x 12-inch walkway board located on the upstream edge of each 12-foot crib bay. Access was gained to the toe of the timber crib using a 20-foot fiberglass extension ladder that was extended from the spillway sill to the floor of the needle bays. Representative photographs were taken of each needle bay to document the leakage observed, condition of exposed timber crib members and the ceiling, side and floor facing boards. The presence of the concrete foundation sill was confirmed by touch as the water beyond the needle bay floor was most commonly too turbid to see it. The depth of the water beyond the toe of the concrete foundation sill was determined by stepping off the sill to the river bottom directly downstream of the sill. The extent of undercutting at the downstream toe was judged using the engineer’s boot and feeling with the toe.
For Needle Bays 1 through 3, one bay was accessed at a time, moving the ladder to each individual bay and wading to each side of the needle bay to access the adjacent full timber cribs. As the inspectors became more comfortable with the access and footing in and around the toe of the timber crib, the ladder was moved to multiple bays and a tether line was used to allow access to one bay to either side of the bay in which the ladder was secured.
Access Walkway – The wooden access walkway structure is founded on and anchored to the concrete spillway sill. It is constructed of pressure treated members of varying sizes. Construction drawings of the walkway structure are available. With the exception of a couple downstream vertical 4-inch x 4-inch center supports that are missing or unsecured at the bottom connection point, the walkway support structure members are in good condition and sound. The parallel 3-inch x 8-inch members that form the lower cords for the walkway structure at each end of the 12-foot bay are founded on an original 12-inch x 12-inch timber that lays between the concrete spillway sill slab. In a number of locations the 12-inch x 12-inch timber member has deteriorated as much as 4 inches, causing a loss of support for the 3-inch x 8-inch members and the front vertical 8-inch x 8-inch timber member. There were also a few locations where the downstream anchorage of the 3-inch x 8-inch member to the concrete sill was loose, damaged or missing. Concrete Spillway Sill Slab – The concrete caps were confirmed to be as shown on the design drawings, approximately 10-foot long, 6-foot wide and 20 inches thick. Other than the sill slab in Bay 21, they appear to be level and are in good condition, with a few minor pockmarks noted where aggregate most likely was ejected. All the sill slabs showed some erosion/spalling of the downstream nappe with exposure of aggregate. It appears that the upstream 12 inches of the present sill slab was placed at a later time than the main sill slab section. The construction joint between this upstream 12-inch wide section and the larger downstream sill slab varied in width. The sill slab in Bay 21 has failed (Figure 10).
Figure 10 – Bay B21 Failed concrete spillway sill.
There was no obvious deflection or settlement noted that could be attributed to the failure. Based on the condition of the cracks that extend the entire length of the slab, the failure is not recent. A condition found to be common throughout the timber crib was visible separation between the bottom of the concrete sill slab and what remains of the original wooden spillway sill boards upon which it appears the sill slabs were placed. This is evidenced by 4 inches of the 60d nails anchoring the original spillway sill boards to the timber crib structure being visible. (Figure 11)
Figure 11 – Bay B21 Failed concrete spillway sill.
Needle Bays –The nomenclature used to describe the bays on the dam is described as follows. The 12-foot timber crib bays are numbered B1 to B21 from Downstream (D/S) Upstream (U/S). The first two characters of a bay description refer to the 12-foot wide bay. The second two characters refer to the needle bay (6-foot wide bay section) located in the respective timber crib bay. (i.e. B9N6 – Needle bay N6 located in 12-foot timber crib bay B9). If only the first two characters are shown, this refers to a timber crib bay that does not contain a needle bay.
The condition of various components within the needle bays varied considerably from bay to bay. A brief summary of each needle bay’s condition is provided in the following table:
Bend Timber Crib Spillway Inspection Summary Results
Downstream Facing Boards – The boards on the downstream face of the timber crib are fir. This is in contrast to the crib timbers being pine. The condition of the downstream facing boards varies significantly, with some in fair condition and experiencing less than 10 percent section loss and some surface weathering, to locations where the boards are absent, exposing the crib timbers and rockfill material. The material in all the remaining facing boards was found to be solid. There was significant erosion and weathering, but drilling of the boards identified no areas could be described as soft or rotten. A common condition for the boards was cracking and splits in the end 25 percent of the boards. Boards near the tailwater did not show any significantly different signs of deterioration (Figure 12).
Figure 12 – Bay B2 D/S face of timber crib looking from Bay B6N4 showing condition variability of D/S facing boards.
Crib Timbers – The crib timbers were found to be composed of pine. This is in contrast to the facing boards composed of fir. The condition of the crib timbers was generally consistent throughout the timber crib section where it was exposed and available for inspection. The exceptions to these observations were the timbers that had been exposed to the leakage events in 2008 and 2009, with those timbers exhibiting the greatest amount of section loss. Section loss was also noted to be accelerated at locations where the crib timbers were exposed due to the loss of a facing board (Figure 13). The condition of the remaining portion of the crib timbers was found to be solid and sound. The wood shavings provided by the drilling confirmed that wood material to be free of rot, and damp to cores. Where the 60d nails were driven into the timbers, the effort required was similar to what would be expected to drive a nail into green timber. There were no signs of any treatment products in the wood. It was noted that the knots in the timbers were much more resistant to erosion, leaving the timbers to appear as round sections (logs) in many instances. Splices to horizontal members within the crib were also noted.
Figure 13 – Bay B4N3 Erosion of crib timber due to lack of cover by D/S facing board.
Connectors – The timbers in the crib were connected using drift pins at the interfaces where they crossed at 90 degree angles. Splices were composed of fitted ends with two bolts. The foundation timbers were set directly on a bed of concrete and fixed with anchor bolts. The facing boards on the interior of the needle bays and the D/S face of the timber crib were attached using 60d nails. All metal connectors were found to be in good condition. The drift pins showed very little corrosion loss, and in every case where a pin was available, and struck with the 3-lb hammer, it proved to ring solid and did not move. All foundation anchor bolts that were noted appeared to be in good condition. Bolt bodies, washers and nut exhibited little corrosion damage. Exposed threads were commonly still visible. All anchor bolts sounded with the 3-lb hammer rung true and did not move (Figure 14). The 60d nails were generally in good condition. A few nails that were exposed in a location that experienced a continuous flow ofwater did show some signs of section loss. Nails that were a minimum 50 percent embedded in a wood member could be bent without damage to the wood or significant loosening of the nail. All exposed nails that were driven into wood members held tightly.
Figure 14 – Bay B2 Absence of rockfill (typical).
Rockfill – The rockfill is comprised of basalt rock, relatively well graded for D50 sizes from 3 inches to 15 inches. The rock shows no signs of deterioration or damage from erosion. Consistently across the entire timber crib section, the rockfill has settled, leaving a space of 3 to 6 inches between the top of the rockfill and the next horizontal level of the timber crib (layer defined by the 12-in x 12-in timbers). In a number of cases where the rockfill is no longer confined in the crib by a needle bay facing board, the gradation of the material has caused it to remain in place. Observations of the lower and D/S-most 2-3 feet of the crib made it appear that rockfill was never placed in this zone, or if it was, it is missing. Where there are missing D/S facing boards the rockfill is generally still in place (Figure 15 & 16).
Figure 15 – Bay B2 Absence of rockfill (typical).
Figure 16 – Bay B10 Rockfill settlement.
Abutments – There are no design or construction drawings for either abutment. The left abutment of the timber crib section is a former Newport St. bridge pier (Figure 17). It is massive concrete and appears to be in excellent condition. Access to the U/S or D/S toe of this abutment was not available.
Figure 17 – Left abutment from U/S
The right abutment (Figure 18) is a grouted masonry structure most likely constructed at the same time as the timber crib. Based on inspection of foundation of the adjacent stoplogged sluiceway constructed in the same manner with the same materials it appears to be founded on basalt at the river bottom. The walls were plumb and the top was level to the eye. There were no signs of distress in the grouted joints.
Figure 18 – Right abutment on D/S side
There is no public safety issue associated with the recent change to the condition of the dam. The timber crib dam spillway section is in good condition considering it is approximately 100 years old. The primary components consisting of the crib timbers and the basalt rockfill have experienced varying degrees of deterioration. Based on the visual observations made during this inspection they are judged to be in satisfactory condition. No conclusion can be made for Needle Bay B16N11 at this time, however it is anticipated that its condition is similar to what was observed in Bay B18N12 and B19N13.
The observations provide better knowledge than previously available regarding the mode of the severe leakage events experienced in 2008, 2009 and most recently in October 2013. The needle bay sections of the timber crib are the most vulnerable to deterioration, in particular the ceiling boards. As the internal components of the needle bays age and deteriorate, their ability to provide adequate support/restraint for the rockfill material is gradually lost. The loss or failure of a needle bay’s side facing board does not appear to be as critical as the ceiling boards. A ceiling facing board failure results in a release of the rockfill into the bay. The associated loss of support to the upstream facing boards due to the release of the rockfill results in greater stress loads in those upstream facing boards. As individual upstream facing boards fail, the loads are shifted to adjacent boards, increasing the stress level in them, causing them to deform further, and eventually fail. As this cycle of aging and deterioration continues, larger pathways for leakage are created leading to the loss of more rockfill (and its associated support), resulting in an increased number of upstream facing board failures.
While there has been settlement of the rockfill in the timber crib sections that do not support a needle bay, the rockfill remains sufficiently confined to prevent the complete loss of the upstream facing board support that leads to overstressing. In these bays the increase in leakage is most likely due to the erosion of joints/seams between the upstream facing boards and the slow progression/growth of holes between the upstream facing boards.
The observations made during this inspection confirm assumptions supporting the conclusion that the dam is not a threat to the environment or public. There is no reason to suspect that a failure of the dam will occur in the near term. Leakage through the timber crib can be expected to increase on a global basis. Absent action taken to address the degradation of the upstream facing boards with their replacement or the installation of a bulk head system (equivalent or equal to the sheet pile), leakage events similar to the type experienced in 2008, 2009 and most recently in October 2013 can be expected to continue. Increased leakage does not present a threat to the stability of the dam in the near term. In a longer time frame, the rate of leakage will continue to increase, and with it the deterioration of the timber crib’s structural components, eventually leading to their failure. It is anticipated this process will evolve slowly, and it is likely that leakage through the existing timber crib would become so pervasive that the dam would be unable to maintain normal water levels prior to the development of a structural failure mode. The progression of leakage through the upstream face of the timber crib leading to a loss of its ability to maintain the reservoir level is the most likely failure mode for the dam, and not a catastrophic failure.
Attachment 1: State of Oregon Water Resources Department – Inspection Summary Letter dated October 16, 2012.
Read More: Bend Inspection Report PDF