NECESSITY IS THE MOTHER OF INVENTION M. Leslie Boyd, P.E., Freese and Nichols, Inc Mike Lowe, P.E., Lower Colorado River Authority Dustin Mortensen, P.E., Freese and Nichols, Inc. I. INTRODUCTION When Buchanan Dam was constructed in 1938, there was no tainter gate dewatering system provided. Thus, routine maintenance occurred primarily when low reservoir levels exposed the entire gate face. At such times, gates could be exercised, painted and repaired. The only problem was that you had more than thirty gates to deal with and even small maintenance issues became complex because you never knew when the lake would suddenly rise. This caused the premature and unwanted termination of gate painting and chain replacement on at least two occasions. The termination of chain replacement efforts resulted in the hiring of divers to complete the unfinished work underwater and at a greatly increased cost. This made it very clear that a dependable and easy to install dewatering system was no longer a luxury but, much more so, a necessity. Also, the tainter gates had been evaluated and recommended for gate strengthening. There was no way the Owner could embark on such a major undertaking without a viable gate dewatering system. The Lower Colorado River Authority (LCRA) is the above referenced Owner of Buchanan Dam. Once the LCRA made the decision to authorize the development and design of dewatering systems, they also concluded that the system should be deployable in an emergency situation. That is, the system should be deployable in flowing water in the event of a gate failure or a gate malfunction. This decision was based on the value of the water which would be lost if the gate failure emptied the lake down to the gate sill elevation. They concluded that this occurrence is not unreasonable, having witnessed the problems subsequent to the Folsom Dam gate failure a few years ago. Since no system for dewatering a tainter gate in emergency flowing water existed to our knowledge, we embarked upon a very unique and challenging assignment. This is why you have to love engineering. A. History II. BACKGROUND Buchanan Dam was constructed approximately seventy years ago by the LCRA. Again, the original design did not include any tainter gate dewatering system, very likely because the designers deemed it reasonable to draw the lake down in order to perform maintenance on the gates or expected to do maintenance when the reservoir was low. In that period of time that was not an unreasonable expectation because there were very few people in the area and small water demands. However, today, with almost a million people dependent upon LCRA s water supply, the idea of wasting water for maintenance endeavors is essentially unthinkable.
B. Project Description Buchanan Dam is the uppermost of six dams located on the Colorado River in Texas. The system of six dams and reservoirs is collectively known as the Highland Lakes. Buchanan Dam is located near Burnet, Texas and approximately 100 river miles upstream of Austin, Texas. The dam is one of the largest multiple arch dams in the United States. It has a length of two miles and a maximum height of approximately 145 feet. The dam has three gated spillways which contain a total of thirty-seven tainter gates. The LCRA has been able to exercise/maintain these gates only during flood events or during periods of extremely low reservoir elevations. The gates have performed very well over the past seven decades but have been identified as needing some minor structural modifications to their framework to bring them into compliance with present day standards of performance. Thirty of the gates are referred to as the small gates and they are 33 feet wide and 15 feet high. There are 16 such gates in the northernmost spillway and 14 such gates in the middle spillway. The other seven gates are housed in the spillway just north of the powerhouse and are referred to as the large gates. They are 40 feet wide and 25 feet high. The large gates are also scheduled for a dewatering system but will not be discussed further within this paper. Buchanan reservoir is operated at two different normal pool levels depending on the season of the year. The two elevations (msl) are 1020 for winter/spring operation and 1018 for summer/fall operation (hurricane season). The crest elevation for the small gates is at 1005. The loss of stored water from elevation 1020 or 1018 down to elevation 1005 is 301,000 acre feet and 257,000 acre feet, respectively. LCRA recognizes this potential loss of volume due to a gate fault as unacceptable and thus stood firmly behind their desire to have a dewatering system capable of emergency deployment as well as an operator friendly routine maintenance system. LCRA contracted with Freese and Nichols, Inc. (FNI) in June 2009 to develop a dewatering system for all three spillways at Buchanan Dam. A. Background III. PROJECT APPROACH A preliminary evaluation of spillway gate dewatering options was performed in June 2007 by FNI. This report had studied typical stop logs, sheetpile bulkheads, horizontal floating bulkheads, and barn-door type sliding bulkheads for the small gates. Figure 1 taken from that report and included herein, also as Figure 1, presents the 2007 recommended typical stop log concept. The figure illustrates the geometric relationship wherein the pier sets forward of the concrete crest and a sealing apron is proposed for construction in order to provide a bottom sealing surface for the stop logs. Figure 1 also illustrates that the existing pier would need to be extended upstream to provide room for a stop log slot. The existing piers are only four feet wide and it was opined that cutting stop log slots in such a narrow pier without the ability to reinforce the existing concrete was too risky. Thus, the pier was recommended for an upstream extension. Each of the other options studied in the 2007 Report also required the construction of a concrete apron immediately upstream of each gate crest to provide a bottom sealing surface for their systems. None of the studied systems in the 2007 Report for the small gates were capable of being installed in an emergency situation of flowing water. However,
the information contained in this earlier report provided some good information for now attacking the design of a system capable of being installed in flowing water. Figure 1. Previous Report Concept
Subsequent to the 2007 Report, gate rating curves were developed as part of a hoist study wherein LCRA was determining whether they could react quickly enough to a rain bomb which might occur quickly and very near the dam. The study evaluated reaction time and the necessary number of hoists required for the 37 gates at Buchanan Dam. However, the part of the study which was determined relevant to our new dewatering design was the gate discharge ratings which were developed. The ratings were developed using computational fluid dynamic (CFD) models. The ratings which included shallow approach similar to the aprons proposed in the 2007 study indicated an unacceptable level of gate discharge reduction. Thus, we were challenged to approach any new dewatering design with either lower aprons or none at all. The LCRA has dewatering systems at some of their other dams along the Colorado River and they relayed some of their problem issues and criticism of the dewatering systems. These included needing divers for routine installation of stop logs and the difficulties associated with latching and un-latching lifting beams under water. On more than one occasion, the lifting beam and its attached stop log had been partially dropped on one end and required significant time and expense to rectify and retrieve the system components. The LCRA thus conveyed their desire for a system simple enough to install without divers or cranes, and with an improved lifting beam to be more user friendly and safe. B. Dewatering System Based upon our acknowledgement of the above described matters and from a cursory review of several documents in the technical literature mentioning emergency dewatering systems, we developed the following list of desirable/necessary features for a dewatering system for the small gates at Buchanan Dam: Must be deployable in flowing water Should be easy to install Should have a simple lifting beam system Should be deployable at any reservoir elevation, i.e. at or below normal pool Should not require an approach apron Should not significantly reduce discharge capacity of the gated spillways Must be readily deployable by LCRA without divers or cranes or transport by others Should not induce any significant hydraulic loads resulting in abnormal hoist function The preliminary dewatering system developed by the FNI team is shown on Figure 2. The following key points are presented: The curved track is mounted to the pier face and negates the need for any upstream apron The bottom stop log will seal against an existing concrete surface on the crest The logs will be hinged and traverse the curved track using two wheels on each end Each stop log will be its own lifting beam Each log will be lifted by the upper axle located at each end of the log
Any log can support the entire assemblage The entire system will be assembled one log at a time in the bay to be dewatered A dogging system will be used to allow the stacking procedure and free the hoist for transporting each log bay-to-bay The bottom log is special, all other logs are interchangeable All horizontal travel (bay-to-bay) will be with only one log All heavy lifts will be made near each pier to minimize cantilevered loading The top log will never be subjected to flows over its top surface Figure 2. Present Concept
Horizontal joints will be sealed with flap seals to prevent abnormal forces during emergency deployment Each end of each log will be sealed with vertical flap seals Being a tube sealed on the ends, each log through the simple addition of threaded inlets can allow water ballast to be added in the event more weight becomes necessary to seat the system The FNI team had confidence in the design developed above but advised LCRA that only a model test could prove its effectiveness in flowing water and provide data on any abnormal hydraulic forces or phenomenon. LCRA concurred and the modeling skills of Utah Water Research Laboratory (UWRL) were engaged. A. The Model IV. HYDRAULIC MODEL STUDY FNI developed preliminary concept drawings to a level of detail where UWRL could construct the hydraulic model components necessary for the study. The UWRL model team was led by Mike Johnson, P.E. The following is a very brief summary of the extensive efforts provided by UWRL on this investigation and is not intended to be a complete or allinclusive presentation. UWRL selected a model scale of 1:16.5 and determined that a Froude model was appropriate for the items needing study. UWRL constructed a gate and spillway model in the flume of their research facility in Logan, Utah. Figure 3 is a photograph of the spillway model and flume. The key issues presented for study were: Could the system be deployed in flowing water What pier nose modification provided acceptable hydraulic flow What effect would the tracks mounted on the vertical pier face have on discharge capacity What, if any, hydraulic phenomenon or forces were occurring on the stop logs What loads would be imparted to the lifting mechanisms (cables) UWRL had the stop logs machined from aluminum and duplicated the hinging and wheel system very thoroughly. Weights were also controlled to the proper model ratio by machining out the cavity within each log until the appropriate weight was obtained. A typical log is shown in Figure 4. The rubber seals were simulated using thin rubber strips. Load cells were placed on each lifting cable to measure the loads. The entire system was monitored with computers to provide real time output which was stored for future reference.
Figure 3. Spillway Flume Model Figure 4. Typical Log Model
B. The Testing The testing first proceeded with the various pier nose shapes presented for study. A half round pier nose and a blunt nose with ellipses were evaluated. The new pier shapes were found to produce very close to the same discharge. The FNI design team then chose the blunt nose with ellipses as being the one for future investigation. This shape provided more flat sealing surface area for the proposed vertical side seals. Both nose shapes were evaluated with the surface mounted track installed. It was determined that the reduction in discharge was only 1 to 2 percent. UWRL then performed a series of tests installing the stop logs at various reservoir elevations in both still and flowing water. LCRA and FNI witnessed some of the final testing of the system during a visit to the laboratory in October 2009. During this visit, a partially blocked gate was simulated (Figure 5) and the system was observed as performing very well in several emergency simulation deployments. Figure 5. Partially Blocked Gate
C. The Results The UWRL model study was a resounding success and indicated the proposed system was viable for Buchanan Dam s small gates. A complete Hydraulic Model Study Report has been prepared by UWRL and affords a much more detailed look at the system. However for the purpose of this paper, the significant findings are: The curved track system is viable in both still and flowing water. The reduction in flow capacity caused by the projecting tracks is minimal. There is no noted hydraulic phenomenon causing adverse conditions. A small jump was noted occurring on the downstream apron for just a few moments before the bottom log fully seated. The maximum load on each lifting cable was approximately 37,000 pounds and this corresponds essentially with the dead weight of the stop log system. V. CONCLUSION We believe the preliminary dewatering system developed for the 33 feet by 15 feet gates at Buchanan dam present a viable dewatering system both for routine installation and for deployment in flowing water emergency situations. We believe the system has key features which support its ease of use: first, with the innovative approach to the lifting beam inclusion into each stop log, and second with the articulating logs on a curved track. Further, this may well be one of the first dewatering systems for a tainter gate spillway suitable for emergency deployment.