Civil Works: Performance of Taum Sauk Upper Reservoir Dam During Refill

Experts describe the refill process, the instrumentation monitored at the upper reservoir, the inspections conducted during the refill program, and the performance of the dam during the refill program.

By Jared Deible, John Osterle, Charles Weatherford, Tom Hollenkamp, and Matt Frerking

The Taum Sauk Project consists of an upper reservoir and a lower reservoir connected by a vertical shaft, rock tunnel, and penstock. The powerhouse at the project has two pump-turbines with a total generation capacity of 450 MW. The original rockfill dike, constructed in 1963 to form the upper reservoir at the Taum Sauk Pump Storage Project, failed abruptly on Dec. 14, 2005. The original upper reservoir dam was removed and completely rebuilt from 2006 to 2010 as a 2.8 million cubic yard Roller Compacted Concrete (RCC) Dam in compliance with Federal Energy Regulatory Commission (FERC) regulations and Missouri environmental permitting regulations.

Paul C. Rizzo Associates, Inc. prepared a Dam Performance and Instrumentation Report in cooperation with AmerenUE to document and evaluate the performance of the Taum Sauk Upper Reservoir Dam and instrumentation during the refill of the reservoir.

The Taum Sauk Upper Reservoir was refilled in stages to assess the structure's performance at each step. Because it is a pumped storage project with no natural inflow, the reservoir level can be raised and lowered with the reversible pump turbines at the project. The reservoir levels, start dates, end dates, and approximate durations at each level are summarized in Table 1 (Pg. 30). It is noted that the floor of the upper reservoir is at approximately El. 1505 and the maximum operating level is El. 1597.

The primary objective of the Refill Program was to assess the behavior of the Dam under partial and full head by observing and measuring the following:

– Leakage and/or seepage into the Gallery, Adits, Main Access Tunnel, and Toe Ditches.

– Piezometric pressures in piezometers in the Gallery and downstream of the Gallery at the Adits, and at critical locations in the foundation throughout the Dam.

– Alignment changes (horizontal and vertical), if any, of monuments on the Crest Road.

– Deformation of the dam (downstream/upstream, along the dam axis direction, and vertical) as measured with Joint Meters installed at the Crest and Gallery at each Monolith Construction Joint.

Instrumentation and Inspections

What follows is a description of the dam safety instrumentation monitored during the refill program, the monitoring frequency, and a summary of the inspections performed. Instrumentation at the site includes 24 vibrating wire piezometers, 10 leakage flumes, and 23 surface monuments. Inspections at the site during the refill program included a complete inspection of the upper reservoir every 12 hours.


A total of 24 vibrating wire piezometers are located at 11 stations throughout the dam. The piezometers are designed to measure uplift at the RCC/Rock interface at several sections, and uplift along defects in the foundation in several locations.

Twenty of the 24 piezometers at the upper reservoir are connected to the Upper Reservoir Distributed Control System (DCS). The DCS enables the piezometers to be read automatically at specified intervals. Hourly readings were reported for the piezometers connected to the DCS during the refill program. The four piezometers not connected to the DCS were read manually four times per day (approximately every 6 hours) during the refill program.


A total of ten trapezoidal flumes were installed at five locations in the gallery at the upper reservoir. A pair of flumes is located at each Gallery outlet so that the total outflow from the drainage gallery is measured.

The flumes are located in the drainage gallery trench so that leakage from crest-to-gallery drains, foundation drains, and any other leakage into the gallery flows through the flumes. Flume readings were taken twice a day, once every 12 hours.

Surface Monuments

Twenty-three surface monuments were installed along the Crest of the Upper Reservoir Dam. Two monuments are located at each monolith joint, one on each side of the monolith joint. Two additional monuments are located at critical sections of the dam.

The survey system also includes benchmarks off the dam. The surface monuments are designed to measure any significant deformations of the dam. The monuments are surveyed with a robotic total station and a digital level.

A baseline survey was conducted before the beginning of the refill program, and surveys were conducted at Time Steps 1, 3, 5, and 7 during the refill program. The surveys taken during the refill program were compared to the baseline survey to estimate total displacements.

Joint Meters

Joint meters at the Taum Sauk Upper Reservoir consist of three pins set in the concrete around a construction joint. The gages allow displacements across the joint to be accurately measured and compared to previous readings to detect trends.

The distance between each of the three pins is measured with a caliper, and each joint meter allows measurement of movement in the vertical and horizontal directions.

Two joint meters are located at each monolith construction joint in the dam. Ten joint meters are located on the Crest of the Upper Reservoir Dam and ten joint meters are located at the same joints in the Gallery. The joint meters were installed to monitor differential displacements between Monoliths.


Visual inspections were conducted every 12 hours during the Upper Reservoir Refill Program by Rizzo and AmerenUE personnel. The following key items were included in the inspections:

– Leakage/Seepage in Gallery, Adits, Access Tunnel, and Toe Ditches, including an evaluation of the turbidity.

– Visual inspection of instrumentation pipe anchorages.

– Visual observation of Upper Reservoir Staff Gage.

Assessment of Dam Performance

The following sections discuss the performance of the Upper Reservoir Dam and Instrumentation during the refill program for each of the key parameters monitored.

In accordance with FERC guidelines, threshold levels and design basis values were established for instrumentation, where appropriate, prior to the start of the refill program.

Threshold levels are defined by FERC as the reading that indicates a significant departure from the normal range of readings and prompts an action. The design basis value is defined by FERC as the value that is used in the design analysis for the project.

Design basis values were established for piezometers based on stability analysis conducted for the project. Design basis values were not established for flumes, joint meters, or surface monuments because there was no design analysis involving these instruments.

Preliminary threshold levels were established for all instrumentation for the refill program based on expected values. Threshold levels were established based on expected values because historic data was not available for the new upper reservoir to establish a range of expected values. For piezometers, the threshold levels established corresponded to a drain efficiency of 60 percent.


Piezometers and Uplift Pressures

All piezometer readings were below threshold and design basis values during the refill program. Threshold levels were set based on a drain efficiency of 60 percent, and piezometer readings during the refill program indicate that actual uplift pressures on the dam were less than expected values.

An analysis of the refill program piezometer data indicates the drain efficiency varied from 79 percent to 100 percent for the uplift pressures at the RCC/Rock interface.

To verify that the piezometers at the upper reservoir were functioning correctly, the piezometer readings and reservoir elevations were plotted vs. time. The plots showed all piezometers responded to changes in reservoir elevation during the upper reservoir refill. This indicates that all piezometers were working correctly.

Movements and Structural

Visual Inspections

Visual inspections did not indicate any signs of structural distress during the refill program. No movement between monoliths was observed, and there was no cracking or any other sign of structural distress observed during the refill program inspections.

Joint Meters

All displacements between Monoliths measured by joint meters were less than the threshold levels established for the refill program.

The maximum measured displacement was 0.09 in. or approximately 2.3 millimeters (mm) in the horizontal (joint opening/closing) direction, and -0.38 in. or 9.7 mm in the upstream/downstream (joint shear) direction.

Joint meters did not show any trends that would indicate significant displacement occurred between monoliths during the refill program.

Survey System

Survey data indicate the dam performed as expected with respect to movements during the refill program. All measured displacements, both horizontal and vertical, were less than the threshold levels developed for the refill program.

A review of the survey data indicates no significant movement of the dam occurred during the refill program, and there were no trends indicating movement in the horizontal or vertical direction. The maximum measured displacement was 0.29 inches. This is within the accuracy that was estimated for the survey system.

Taum Sauk Upper Reservoir following the completion of the refill program.

Leakage and Seepage

Visual Inspections at Downstream Toe

Visual inspections did not indicate any major seepage or leakage at the toe of the dam during the refill program. Several wet spots were observed and one minor slough was noted a significant distance downstream from the toe. It is suspected that the wet spots and the associated minor slough were associated with a combination of sources including snow and ice melt, ponded water, and seepage through the floor of the dam.

Foundation Drain Flows

Leakage from foundation drains during the refill program was minimal and observed in approximately 15 of the 677 foundation drains installed during construction. The measured values for leakage from the foundation drains were below threshold levels established for the refill program. All drain flows were clear and no turbidity was observed.

Crest-to-Gallery Drain Flows

Leakage from crest-to-gallery drains was highly variable during the refill program, with some drains being completely dry and others leaking considerably, especially in Monolith Nos. 1, 2, and 4.

The major causes of the difference in the amount of leakage between monoliths was likely the difference in the design of the waterstops installed in these Monoliths and the presence of a horizontal cold joint in monoliths 1 and 2.

A cold joint was present in Monoliths 1 and 2 where a lift joint was left exposed for approximately one year, and flows from the crest to gallery drains in Monoliths 1 and 2 were significantly higher than in other monoliths.

The design of the vertical waterstops was changed early in the project, and the revised design allowed for more stable and restrained placement of the associated crack inducer board. With the early waterstops, there was a tendency for the crack inducer plate to move laterally, away from the water stop bulb. With the re-design, the crack inducer plate was fixed more rigidly and unable to move.

Leakage was also observed at several construction joints during the refill program.

The majority of construction joints where leakage was observed are located in Monolith 1and 2 where, as mentioned above, the older style of waterstop was used.


The flumes in the drainage trench capture all of the flow from the foundation drains, crest-to-gallery drains, and construction joints. They provide a measure of the cumulative leakage from the Reservoir for all sources, except for losses through the floor of the Reservoir.

The flume data showed that the cumulative leakage peaked at about 1,080 gpm on March 18, 2010, when the Upper Reservoir level reached El. 1597 for the first time.

More than 80 percent of the flow was passing through flume Nos. 1 and 2 in Monoliths 1 and 2. After the reservoir reached El. 1597 for the first time, the cumulative flow of all flumes and the individual flows from each flume decreased as the reservoir elevation was held constant.

At the completion of the refill program, the upper reservoir was dropped to approximately El. 1500. This allowed sealing of horizontal joints and vertical joints in the upstream face of the upper reservoir.

Over a period of several days, approximately 15,000 lineal feet of horizontal joints and 34 vertical joints were sealed. After sealing, the total leakage through the flumes with a full upper reservoir had decreased to approximately 300 gpm.

A comparison of the flow through each flume prior to and after sealing of the upstream face of the upper reservoir is provided in Figure 1 (Pg. 32). Each set of leakage readings in the plot in Figure 1 were taken when the upper reservoir was near full.


The performance of the upper reservoir during the refill program was monitored with dam safety instrumentation and inspections. The reservoir was filled through a series of time steps and the performance was assessed at each step.

The dam performed as intended at each time step. Instrumentation readings were generally consistent with expected values and no dam safety issues were identified during the visual inspections. The Taum Sauk Upper Reservoir was returned to operation successfully.

After completion of the Refill Program, the leakage at the upper reservoir remained at or below levels observed during the Refill Program through the summer of 2010. During the winter of 2011, the leakage at the upper reservoir increased above levels observed during the program. This increase in leakage was attributed to opening of joints as the temperature of the RCC decreased.

Jared Deible is a Senior Project Engineer for Paul C. Rizzo Associates. John Osterle is a Vice President of Paul C. Rizzo Associates. Charles Weatherford is a Resident Engineer for Paul C. Rizzo Associates. Tom Hollenkamp is Chief Dam Safety Engineer for AmerenUE. Matt Frerking is a Managing Supervisor for AmerenUE.

More Hydro Review Current Issue Articles
More Hydro Review Archives Issue Articles

To access this Article, go to: