Close 

Strengthening Gate Arms at Rocky Reach Dam

The spillway gates at Rocky Reach Dam are similar in size, design, age and environment to those used at Folsom Dam. To avoid a similar failure, owner Chelan County Public Utility District took proactive measures that resulted in an ongoing program of gate arm monitoring and strengthening.

By William Christman, Gene Yow and Justin Fletcher

The spillway gates at Rocky Reach Dam, on the Columbia River in Washington, are similar in size, design, age and environment to those used at Folsom Dam, on the American River in California. Thus, the spillway gate failure experienced at Folsom Dam in July 1995 caught the attention of Rocky Reach Dam owner Chelan County Public Utility District.

The failure resulted in nearly 40% of Folsom Lake draining through the broken gate. The freshwater that reached San Francisco Bay altered the migration habits of some salmon and striped bass. A year-long investigation by the Bureau of Reclamation attributed this failure to a design flaw. According to the report, gate pivot bearings seized up when the gate was lifted and lateral loading stressed support beams to their yield point and the structure buckled.

Hoping to avoid a similar tragedy at Rocky Reach Dam, Chelan County PUD developed a standard testing and maintenance program that includes direct observation tests of gate operation. This is part of a strategy arrived at in 1996 to use the lessons learned from the Folsom Dam failure to help avoid a similar occurrence. The district has been repairing the Rocky Reach Dam spillway gates, based on the test results, for nearly eight years.

So far, the arms of eight of the 12 gates at Rocky Reach Dam have been strengthened, and trunnion bearings have been replaced on two gates. The district plans to strengthen the arms of the remaining four gates in 2012 and 2013 and to continue to monitor trunnion bearing friction by means of gate arm deflection measurements so that bearing replacement can be planned and carried out as needed.

Rocky Reach Dam, on the Columbia River in Washington, is equipped with 12 spillway gates that are similar in size, design, age and environment to those used at Folsom Dam, which experienced a spillway gate failure in July 1995.

Initial testing leads to hammer noise

The 12 tainter (radial) spillway gates at Rocky Reach Dam are 50 feet wide by 58 feet high. They were designed and fabricated in the late 1950s and put into service in 1961. The original trunnion bearings were greaseless bronze bushings with a graphite-based lubricant (Lubrite G1). After more than 40 years of high-quality operation, it seemed reasonable that the condition and performance of the bearings might have been declining.

During initial testing of the gates in 2004, one trunnion bearing on Gate 12 at Rocky Reach Dam "chattered." Those working in the area reported hearing a noise that "sounded like someone hitting the gate with a hammer." District personnel immediately launched an investigation, and the gate was removed from service. This gate was seldom used and consequently one of the "least exercised" gates within the spillway.

Systematic testing of all 12 gates began in the summer, during a period when river flows were low enough to reduce the chances of a spill and convective solar heating was at or near maximum levels, thereby resulting in maximum-normal bearing resistance. During this testing, personnel correlated strain gage readings with measurements of arm deflection in an effort to simplify future gate monitoring.

Testing results showed that the bearings' performance had deteriorated to varying degrees, with some bearings still in relatively good condition. Friction coefficients ranged from about 0.10 to 0.18. The design-basis friction was assumed to be nearer the 0.10 value. Numeric structural modeling of the gates, performed by Mead & Hunt, also showed that the gate design was an "elegant" one, or in other words there was not any significant excess structural capacity. Consequently, the gate arms did not have as much margin in terms of strength as would be considered desirable in a contemporary design. Unity factors for the main arm beams ranged from 0.79 to 1.17, with an average of 1.02, which was considered to not provide an acceptable margin, especially considering the criticality of reliable gate operation.

To address these issues, the district considered two alternatives for remediation of the gates: trunnion bearing replacement and arm strengthening. As alternative solutions were evaluated, it became clear that bearing replacement would be the less costly approach, but the benefit to be gained from bearing replacement was uncertain, since even with restored low bearing friction, the excess capacity of the gate arms did not provide much margin for general unknowns in loading or performance operations.

Efforts focus on Gate 8

Gate 8 was found to have the highest trunnion friction and highest calculated unity factors. Thus, in 2005, the district replaced the trunnion bearings in this gate. The bronze bushings were replaced with composite bushings, which were substantially less expensive, were far easier to handle and install, and had a much shorter lead time.

As part of the bearing replacement, beginning with the first gate that was remediated, the district examined the alignment of the trunnion bearings and found that, at the time of original construction, they were aligned very well in an unloaded condition. But in a loaded condition, they were out of alignment by about 0.15 inch over the 39-inch length of the bearings. The misalignment was significant in the overall bearing/gate performance. Consequently, this was remedied during bearing replacement. Follow-up testing indicated that the bearing realignment had a beneficial effect on stresses in the gate arms. The beneficial effect of the realignment was just as great as the effect of installing new bearings.

Extensive follow-up testing was conducted in an effort to ensure that the new bearings were performing well. During this testing, the bearings exhibited increasing friction through June and July. The trend, for which a sound conclusion was not discovered, was disconcerting. As a result, the district removed the gate from service and planned to replace the bearings yet again.

By the summer of 2006, the district had replaced the composite bearings in Gate 8 with greaseless bronze bushings, similar to the original bearings but with epoxy/Teflon lubricant (Lubrite G10) rather than graphite-based lubricant. Follow-up testing indicated that the friction coefficient was acceptable but that the same degree of improvement in the case of the other spillway gates would not provide the margin desired in terms of arm stresses.

Due to the "slenderness" of the original design, improvements in bearing friction alone, even to the original bearing friction coefficient, did not produce measured arm stresses low enough to satisfy Chelan County PUD's gate resilience demands. The gate arm stresses measured after bearing replacement during hot summer periods left too little safety margin to be viewed as acceptable.

For this reason, the district changed direction again and prepared a design for arm strengthening.

Testing performed to determine condition of the 12 spillway gates at Rocky Reach Dam includes correlation of strain gage readings with measurements of arm deflection.

Strengthening gate arms

Options available for strengthening the gate arms included welding plates across the flange edges, welding plates to the outside of each flange, or adding standard steel WT (wide-flange, tee-shaped) sections on each side of the main arm beams. The chosen approach, adding WT sections, had a number of benefits. Large, long plates are more difficult to handle while avoiding bending them. Welding the stem of a tee to the flange of another beam minimizes the creation of spaces that cannot be inspected. And use of WTs minimizes interference with the bracing members of the gate arm.

To allow for future aging of bearings, the arm strengthening design assumed a trunnion bearing friction coefficient of 0.3, whereas the actual coefficients measured were about 0.15 or less. Designing for a coefficient of 0.3 rather than an optimized value of about 0.15 required an addition of about 10% more steel. Consequently, the district's experience is that significantly more strength, and therefore resiliency, is obtained for a relatively small additional cost.

Strain gage testing showed that, although the design and fabrication were "identical" for the 12 gates, the stresses in each arm were unique due to sweep and camber in the steel beams and trunnion bearing performance. The WTs for each arm beam were chosen based on the worst stresses in comparable beams from among the 12 gates. Thus, the same size WTs were added to all the lower arm beams, and the size was selected based on the lower arm beam with the highest stresses from among the 24 lower arm beams. This was done to maintain a uniform design among the gates.

The district selected Gate 7 as the gate next in line for remedial work and strengthened the arms in 2007 to provide a more robust gate structure, as ongoing monitoring of the gate showed its reserve strength was too low for the district to feel comfortable. Follow-up testing showed that the design worked as planned to reduce stresses in the gate arms. Stresses in the gate arms are acceptable for trunnion bearing friction coefficients up to 0.30.

Two gates a year

Subsequent to the work on Gate 7, the district has worked on two gates per year. Most of this work has involved arm strengthening. However, in August 2008, relatively high arm deflections on Gate 9 were discovered, which indicated friction coefficients as high as 0.25. This led to replacement of the trunnion bearings on this gate in 2009. The original bearings were replaced with greaseless bronze bushings with epoxy/Teflon lubricant (Lubron AQ100). These performed similarly to the Lubrite bearings and, again, improved the bearing alignment provided as much as it benefited the change to new bearings. Consequently, the district reinforced its understanding that strengthening the gate arms and ensuring good trunnion bearing alignment were both important factors in obtaining well-understood and long-lived gate resiliency.

Measurements of gate arm deflection under high temperature conditions - ambient temperatures near 100 degrees Fahrenheit and full sunlight - have shown higher bearing friction at those times. One possible explanation that has been considered is loss of adequate clearance between the bearing and pin due to expansion of the bronze bushing. The cause of this changed performance, which correlates with temperature, is unknown but appears to be an important factor in overall gate performance as verified by the gate arm structural stress monitoring program.

Strengthening work for the gate arms on the 12 spillway gates at Rocky Reach Dam is being performed at a rate of two gates per year.

What was learned?

Chelan County PUD has learned several important lessons over the course of this work on the spillway gates at Rocky Reach Dam.

The moment induced in the gate arms by trunnion bearing friction is sometimes treated as a short-term, transient load. This is not necessarily appropriate. If the gate is closed under water load, the moment is maintained until the gate is moved again.

Graphite-based lubricants can produce a galvanic reaction with steel - even stainless steel - given a little bit of water. Water with electrolytes exacerbates the problem. Correcting alignment of the trunnion bearings improved stresses in the Rocky Reach Dam gate arms about the same amount as replacing the bearings.

Some older gates were designed with very little extra margin for safety, and investigating potential deficiencies is a worthwhile endeavor.

Replacement of trunnion bearings in one gate cost about $260,000. Strengthening the arms of two gates has cost an average of $607,000. For other dam owners considering gate rehabilitation, costs will vary based on size of the gates and bearings.

With regard to why the trunnion bearings appear to exhibit a higher friction coefficient at high temperatures, the district is still seeking sound reasoning for this behavior. But the ongoing gate arm structural monitoring is verifying that the rehabilitated tainter gates at Rocky Reach Dam are performing in a high-quality, robust and safe manner as a result of the rehabilitation program.


Bill Christman is engineering services department manager, Gene Yow is principal civil engineer and Justin Fletcher is project engineer for Chelan County Public Utility District.

More HR Current Issue Articles
More HR Archives Issue Articles

To access this Article, go to:
http://www.hydroworld.com/content/hydro/en/articles/hr/print/volume-31/issue-2/article/strengthening-gate-arms-at-rocky-reach-dam.html