Hydro Review Magazine
Vol. XXIV, No. 3, May 2005
Choosing the Best Method for Gate Repair
By Thomas L. Kahl
Water control gates at aging hydroelectric projects may not provide safe operations. However replacing gates in operating facilities presents many unique challenges, such as how to de water a working area. Three case studies show how hydro project owners found economical, effective rehabilitation options.
At many aging hydroelectric projects, safety concerns and economic risks associated with water control gates are common problems. While the rehabilitation of existing gates shares many technical aspects with new gate construction, rehabilitation often presents unique site-specific challenges. For example, existing gates typically are being used as a primary water retention structure, so any work must be accomplished without jeopardizing worker or facility safety. Also, the site-specific conditions and configuration of hydra facilities sometimes impose difficult and costly access and construction constraints. Knowing the aspects to consider when rehabilitating existing gates can help hydro project owners undertake this work safely and cost-effectively.
Determining gate condition
The first step when considering gate rehabilitation is to determine if the gate presents an unacceptable risk, looking at both actual gate condition and the consequences of failure. Gate failure includes both uncontrolled flow releases and the failure to release flows when needed. Obviously, gates that pose higher consequences as a result of failure need more rigorous scrutiny.
The most fundamental consideration – structural competency – can be determined by analyzing complete load paths for the pressure load-bearing structural members and the operating system. Actual component condition primarily is determined by measuring the dimensions of the in–place gate members and visually observing condition deficiency, such as missing members, deformation, corrosion, and interferences.
Next, the condition of the mechanical and electrical components should be measured through an on-site inspection and discussions with the station operators and maintenance personnel. Typically, the most critical gate operating conditions are during opening and closing.
In addition, the condition of permanently embedded gate guides — particularly the seal mating surfaces — should be assessed, Overlooking a deteriorated gate guide can invalidate an otherwise successful rehab project. However, underwater gate guide repairs are costly, particularly if they require a cofferdam or divers, so typically they should be performed only when needed to ensure proper gate function.
Finally, the current function of any gate should be considered. For example, the present gate operating system may not be appropriate for older hydro plants that have fewer on-site personnel, such as one that has been converted to full remote automation. Also, increase in a dam’s downstream hazard potential can alter a gate’s risk consideration.
Planning the project
Perhaps the most significant challenge of gate rehabilitation is the means of dewatering. Dewatering can be a non-issue when rehabilitating temporary bulkhead or intake gates that have up stream stoplog provisions. But it can be extremely significant for deep sluice or large spillway gates that lack bulkhead or stoplog provisions.
One of the primary considerations in a gate rehab project is whether the gate and/or operating system should be repaired or replaced. A gate that is experiencing leakage, surface corrosion, or structural inadequacy of a limited number of members may be more economically repaired. For example, insufficient strength of the strut arms on large tainter gates typically is more easily addressed by field welding additional cover plates or lateral bracing to the existing strut arms. Conversely, gates with extensive general deterioration or mechanical and electrical components that do not meet safety requirements may be better candidates for replacement. In addition, the need to limit the time that a gate is out of service can make replacement the optimal alternative.
Another factor to consider is whether to use a gate design from the owner and consultant or from a gate vendor. There are advantages and disadvantages to each scenario. Site conditions, the level of design ability, and owner preferences can affect the approach chosen. Some general guidelines:
- If only selected components in a larger system will be repaired or replaced, frequently it is more technically effective and cost-effective to use owned consultant design. If the project owner can unambiguously define and comprehensively convey the extent of the repair work and the performance requirements, vendor design can work better.
- Larger and less-complicated gates may be more cost-effectively procured from local fabricators by using owner/ consultant designs. Conversely, smaller gates, particularly if they are more complicated, may be more cost-effectively procured from a vendor who can incorporate components and details that have performed successfully on similar gates. The author’s rule of thumb is to limit the cost of consultant design to no more than 10 percent of the total gate cost.
- Gate rehabilitation that involves extensive modifications to existing structures typically can be more efficiently completed by the owner/consultant.
Performing the rehabilitation work
Because often it is difficult or even impossible to conclusively define the condition of all gate components before work begins, project management must ac knowledge this additional risk. To control costs, owners must clearly define the contractual scope of work and share potential risks that cannot be fully de fined when the project is bid. For ex ample, in-place sluice gates must be dewatered and the gate bodies lifted from the guides before the condition of the gate guide sealing surfaces can be con- elusively determined. Therefore, it is prudent to bid prices for one or more repair scenarios, then implement the repairs that best match the conditions encountered.
Once the project is complete, comprehensive record drawings and documentation should be com- piled and archived. This is particularly important for portions of the work that are difficult to observe, such as submerged and embedded components. This documentation can greatly improve the reliability of future condition assessments and reduce the ambiguity and costs of future work.
Case studies
The large variety of types, functions, and site-specific conditions of existing water-control gates results in a variety of potential rehabilitation alternatives that defy a single formulaic approach. However, different types of gates frequently share common similarities. The following examples comprehensively explore the evaluation, planning, and implementation phase of three rehabilitation projects for the common types of deep sluice, intake bulkhead, and spill- way tainter gates.
Sluice gates at Middlesex 2
In the 1990s, Green Mountain Power Corporation began to be concerned about the condition of the sluice gates at its 3.2
MW Middlesex 2 hydro station on the Winooski River in Vermont. The two 9- foot-square steel gates with screw stem operators
are located at the spillway’s base in a natural “V” shaped rock gorge 45 feet be low normal reservoir level, Figure 1 shows
a cross section through one gate. These gates, installed in 1936, play an important role in maintaining safe water levels.
Work performed in August 1999 to maintain reliable operation resulted in an accidental sediment release that highlighted the need for more permanent rehabilitation. Sediment was released when the headpond was lowered to reduce the gates’ submergence depth. Summer rainstorm river flows washed the sediment downstream within a few weeks, but Green Mountain received a substantial fine from the state. In addition, the contractor and divers per forming the repair indicated both gates were in poor structural condition. Green Mountain also knew that both gates had substantial leakage, causing winter icing problems. Based on these factors, no further formal gate condition evaluation was needed to justify replacing both gates.
As Figure 1 shows, these gates had no dewatering provisions, probably because when this spillway was built in 1929, designers envisioned that the headpond could be drained during low river flows. But it is now impossible to obtain permits to simply open the gates and completely drain the small headpond. Replacing the gates would require some method of cofferdaming.
During initial planning in the fall of 2001, Green Mountain considered three options:
- Install a cofferdam in a shallow area about 0.5 mile upstream of the spillway;
- Adapt an abandoned timber crib dam used as a cofferdam for the original spillway construction; or
- Install a braced sheet pile coffer dam on the upstream face of the gate openings.
The first two options would mean dewatering the gorge, presenting significant permitting difficulties. And the deep water and steep rock gorge profile upstream of the spillway made the third option technically difficult and expensive, with an estimated cost of $150,000 to $200,000.
Green Mountain hired Kleinschmidt as the design engineer and Fairbanks Mills of St. Johnsbury, Vt., as the contractor. The three companies conducted a joint inspection in December 2001 and formulated a new cofferdam alternative that involved installing a temporary bulkhead in the dry sluice tunnel downstream from a gate, installing a new gate, then using the new gate to dewater the tunnel for work needed on the sluice tunnel concrete. This approach lowered costs compared to an upstream cofferdam and eliminated the need to lower the headpond elevation. It also did not interfere with normal station operation.
The inspection showed that the existing gate operators and threaded stem were in good condition, so they were reused. However, Green Mountain decided to upgrade the gate operators with new remote control and position indication instrumentation. Excellent detailed record drawings were available showing complete dimensions and details of the existing gate, stem extension, and gate guides, so Green Mountain decided to solicit vendor-designed replacement gates. The specifications included:
- A design pressure water level of Elevation 150 feet, the water level for the 100-year flood.
- The gates must conform to the requirements of the U.S. Army Corps of Engineers’ manuals EM 1l10-2-2105 “Design of Hydraulic Steel Structures” and EM 1110-2-2701 “Vertical Lift Gates,” using either the American Institute of Steel Construction (AISC) Load and Resistance Factor Design Method (LRFD), 3rd Edition, 2001, or Allowable Stress Design (ASD), 9th Edition, 1989. For the load and resistance factor method, Green Mountain specified a reliability factor of 0.85 and load factors of 1.4 for the differential hydrostatic forces and 1.2 for the gate deadweight. For the allowable stress design method, this gate was classified as Type A (inaccessible), which requires that the allowable stress used in the design be 0.75 times that normally allowed by AISC.
- The gate stern, including gate connection, would have a minimum structural capacity of at least 238 kips (one kip = 1,000 pounds), which was the maximum capacity of the existing stem. A review of the existing operators deter mined that they could develop a maxi mum stalled motor thrust of about 3,100 kips in tension, so the operators were to be retrofit with a thrust-limiting device.
- A lateral stem extension ice loading at the water level of 5,000 pounds.
- A pseudostatic horizontal seismic acceleration of 0.1 g (gravity).
- Gate sealing details that would provide minimal long-term leakage.
- Mandatory approval by Green Mountain of the proposed gate design and all details.
Green Mountain solicited competitive bids in February 2002 to maximize the time for gate fabrication and allow the new gates to be fabricated at a lower cost in the early spring when specialty steel fabricator workloads are historically low.
Cross Machine of Berlin, NH, was awarded a $35,000 contract in March 2002 to supply both gates. Green Mountain, Kleinschmidt, and Fairbanks Mills reviewed the initial gate body design in April and May. The original gates were fabricated of ASTM A-7 steel with a ½- inch-thick skin plate and ten deep I beams. For the replacement, Cross Ma chine selected a higher-strength ASTM A-36 skin plate reinforced by ASTM A992 material wide-flange beams. The new 40-foot-long stem extension was fabricated of ASTM A 500 Grade B HSS 8-inch by 8-inch by 5/8-inch members that greatly improved compressive buck ling resistance compared to the original I beam stern extension.
Because of the difficult installation conditions at the site, Fairbanks Mills reviewed the gate design for constructability concerns. This review resulted in several design modifications, such as pin- type gate stem-to-body connections, that simplified the underwater installation. These deep sluice gates typically are closed except during occasional high river flow, so considerable attention was focused on the gate seal design details. Solid bulb fluorocarbon-coated J seals were used on the vertical sides, with double stem type seals on the top and bottom edges. Kleinschmidt also reviewed the compatibility and physical clearances of the vendor’s new gate with the existing guides to reduce the chance of costly field revisions during underwater installation.
In addition, Green Mountain gave particular attention to selecting a quality paint system to extend the service life of the gates. The utility chose an epoxy mastic coating having excellent performance characteristics, such as a Salt Fog Test rating exceeding 5,000 hours.
In June 2002, Fairbanks Mills in stalled the Gate No. 1 bulkhead and re moved the sluice gate. This allowed the guides to be completely inspected for the first time since their installation in 1933. Based on videotape taken by Commercial Divers, Kleinschmidt finalized the gate guide repair drawings, which included replacing deteriorated gate guide anchor bolts with new stainless steel adhesive anchors and cleaning the guide seal mating surfaces by grinding. The guide surfaces that mate against the gate- mounted rubber seals had only minimal pitting, so provisions for smoothing pit ting were not needed. Fairbanks Mills completed the underwater guide repair in July 2002. Green Mountain, Kleinschmidt, and Fairbanks Mills conducted a joint shop inspection of the gates later that month, and the first new gate was installed in August.
Fairbanks Mills performed the same steps for both gates:
- Install a temporary bulkhead in the dry sluice tunnel downstream of a gate.
- Raise the sluice gate to water the tunnel upstream of the temporary bulk head.
- Remove the existing gate stem, stem extension, and gate.
- Inspect and repair the underwater gate guides.
- Install the new gate.
- Install the new stem extensions and stem and fully connect them to the re used existing screw stem operator.
- Close the new gate to dewater the sluice tunnel.
- Remove the bulkhead in the dry.
- Repair the sluice tunnel concrete, which was found to have significant deterioration, in the dry.
During gate stem disassembly, Fair banks Mills discovered that four of the eight bolts attaching the threaded rod to the stem extension of Gate No. 1 were either completely sheared or severely distorted. This condition is undetectable by any method except complete gate removal and disassembly. If the utility had not decided to replace these gates, the failure of all the bolts would have resulted in the stem to gate connection severing when personnel attempted to raise the gates during flood conditions.
After the first gate was fully operational, Fairbanks Mills removed and reinstalled the bulkhead behind the second gate in early September. The contractor removed the second gate, repaired the underwater guides as discussed for the first gate, and replaced the second gate.
Fairbanks Mills repaired the tunnel concrete downstream of both gates by resurfacing eroded areas with 4,000 pounds per square inch (psi) air-en trained concrete anchored by No. 5 rebar dowels -on 30-in centers. The entire project was completed by the end of October 2003 at a cost of about $350,000. Both new gates had zero leakage.
Intake gates at Vergennes
The dual-stern rack-and-pinion operating system for the two 8-foot-wide by 10- foot-high wooden bulkhead intake gates at Green Mountain’s 2.4-MWVergennes project was more than 80 years old in 1992. In addition, the wooden gate bodies were more than 40 years old. The project has a shallow intake, with 16 feet from the sill to normal pond water level. The intake gate guides were replaced with new steel guides during a 1992 dewatering of the south channel, but the 1992 repairs were only intended to be interim. The gates were scheduled to be re placed as part of a general station rehabilitation in the late 1990s, but project economics indefinitely postponed these re habilitation plans.
In December 2001, one of the wooden gate stems on Gate No. 1 broke while the gate was being lifted. Wood decay had reduced
the stem’s cross section below the limit that could carry the gate reactions. Green Mountain used a come-along attached to
cables looped through a 1-foot-square filler gate in the main gate body to lift and temporarily secure Gate No. 1. This
filler gate is opened to fill the penstock and equalize the pressure across the headgate before the main gate is lifted
after penstock rewatering. Due to the poor condition of the gate structures and obsolescence of the operators, Green
Mountain decided to replace both.
Because Green Mountain uses these shallow low-head gates only a few times a year for turbine maintenance dewatering, the utility decided to install simple slide gates operated by a standard monorail-sup ported electric hoist. Figure 2 shows a cross section of the replacement gates.
To reduce costs, Green Mountain decided to use in-house maintenance personnel to install the new gates. The in take deck is congested, and the monorail supports and gate access required some reconfiguring of access stairs and plat forms. Also, monorail supports needed to be placed where the lightly reinforced concrete intake is structurally adequate. Green Mountain confirmed this structural adequacy through a combination of site inspection and a review of original 1920 intake rebar drawings.
The intake’s existing structure would dictate the monorail design, and its load capacity had to correspond with the reactions needed for operating the gate. Consequently, Green Mountain hired Kleinschmidt to prepare a comprehensive final design package that integrated all the civil, structural, mechanical, and electrical components, including the gates, monorail with steel support system, and new access platforms and stairs. Kleinschmidt determined that a less expensive sealing system of neoprene-backed ultra high-molecular-weight polymer bearing and sealing strips mounted around the perimeters would be adequate for these low-head seating head gates. The gates typically will be lowered under equal hydrostatic pressure with the turbine cylinder gates closed. However, Green Mountain decided to fill the spaces upstream of the skinplate between the gate horizontal beams with concrete to provide ballast so that the gates would close by gravity under an emergency condition of full turbine discharge.
Crane access at this congested site is limited by the river to the north, a busy state highway to the east, a three-story building to the south, and the penstock crossing a ravine to the south. Therefore, to reduce crane size, Green Mountain determined that its crew would add concrete ballast required for gravity closure to the steel gates after the gate weldments were placed on the intake deck.
Although the monorail system was designed to raise each gate under full unequal static pressure, Green Mountain requested smaller penstock filler gates in the new gates to equalize hydrostatic pressure before normal gate raising. An unusual circumstance at this site is that the turbine cylinder gates have so much leakage, a second filler gate was necessary in each headgate to provide enough water flow into the penstock to compensate for the turbine leakage and equalize the gate’s static water pressure.
Kleinschmidt finalized the design over the winter of 2002, and Green Mountain solicited bids for the supply of the gates, monorail system, and new access plat forms in March 2002. The next month, a local steel fabricator, Reliance Steel of Colchester, Vt., was awarded a $35,000 contract.
From May to July, Reliance fabricated the gates, monorail, and new stair and platform. Green Mountain’s electrical crew changed the intake’s service line in early July, and the hydro maintenance crew installed the new gate and operator frame in about a week in late July and early August 2002. First, they erected the new monorail system, which they then used to help install the new intake gates. Because the portable crane used to unload the gates was about 130 feet from the intake, the new gate concrete ballast was pumped to the intake deck and placed into the gates. The monorail system tem was then used to raise the gates and place them in the slots the next day.
Since that time, the gate system has performed well and seals with leakage of less than 2 gallons per minute, which Green Mountain considered acceptable.
Tainter gates at Clark Falls
The three 24-foot-wide by 23.5-foot- high 1938 steel tainter gates at Central Vermont Public Service Corporation’s 3 MW Clark Falls hydro station in northern Vermont were refurbished as part of a general spillway rehabilitation project in 1988. These spillway gates control the water levels in Arrowhead Reservoir.
Additional problems began to develop in 1999, when small pieces of the gates’ 11-year-old wooden sills began to wear, increasing bottom leakage. While not an immediate safety concern, this was a potential operational concern because winter sill leakage can result in ice buildup, requiring manual ice removal before the gate can be raised during high winter river flows. Light beige paint applied in 1988 exacerbated ice buildup by reducing solar heat absorption compared to the previous black color. In addition, the light color showed the rust bleeding from between the riveted faying surfaces(1) between the downstream skin plate face and horizontal stiffeners.
The most troublesome problem, which began to develop in 2000, was corrosion of the embedded vertical side seal face plates. The holes that developed allowed water to enter the guide heater cavities and leak around the vertical side gate guides. Also, these gate guides were experiencing long-term corrosion that roughened the surface and presented the potential for damaging the gates’ vertical side rubber J seals.
Because these problems were minor compared to the overall good gate condition, Central Vermont and Kleinschmidt decided rehabilitation, rather than re placement, was the optimal approach.
Kleinschmidt evaluated several alter natives for repairing the wooden bottom sill and side-seal faceplates in the summer of 2002. Because these are large gates without permanent bulkhead provisions, the first options concentrated on less costly techniques that would not require a cofferdam (the 1988 tainter gate cofferdam costs were $100,000). The ideas concentrated on variations of lag bolting a new steel plate on the downstream face of the current wooden sill. The initial conceptual side seal re pair option consisted of wedging a 3/16- inch-thick stainless steel guide faceplate overlay upstream, between the gates’ neoprene J seals and the existing guide carbon steel face, that was flush with the vertical abutment side walls. The up stream face of this plate would then be welded underwater, and the downstream face would be welded in the dry to the existing embedded guides. But Central Vermont decided to rehabilitate these three gates in the dry behind cofferdams so that the gates could be more permanently repaired and the upstream skin plates repainted.
Kleinschmidt prepared project design drawings and bid drawings in the late fall of 2002. These documents incorporated
comprehensive record drawings and a project maintenance manual summarizing the 1988 repairs, materials, and cofferdams.
Lessons that were learned and clearly documented from the 1988 work were used to reduce the cofferdaming risks to the
bidding contractors and subsequent cost to Central Vermont.
Figure 3 shows the design of the new sill where the wooden sill is replaced with a steel weldment that has a replace able neoprene bottom seal. The refurbished vertical side seal detail features a new 304 stainless steel faceplate placed over the existing vertical embedded guide faceplate, with heater tubes welded to the backside. The gate guide cavities then were filled with closed cell foam to concentrate the heat in the guide seal area and improve the efficiency of the gates’ side heaters.
Alpine Construction of Stillwater, N.Y., received a $400,000 contract in February 2003. In April 2003, Alpine presented its proposed braced steel bulkhead cofferdam design, which was reviewed, approved, and included in the final Federal Energy Regulatory Com mission project approval received in May 2003. Alpine began cofferdaming the first gate in early July 2003 and completed the work in October 2003.
Just before construction, Central Vermont operational personnel expressed concern that the rubber in the Figure 3 sill detail would be difficult to replace, so the rubber was deleted for the first two gates. While the first gate initially sealed adequately, the second gate sealed poorly. Therefore, the seal was reinstated for the third gate, and Central Vermont retrofitted the second gate with a rubber bottom seal in the summer of 2003.
Tom Kahl, P.E. is senior engineer and manager of the Hydro Engineering Department of Kleinschmidt in Pittsfield, Maine. Mr. Kahl may be reached at Kleinschmidt , 141 Main Street, P.O. Box 650, Pittsfield, ME 04901; (1) 207-487-3328; E-mail: Tom.Kahl@KleinschmidtUSA.com.
Note
(1) Faying surfaces are the steel surfaces that are in contact with each other and are held tight by the rivets.
