When permanent deformation was found on a gate at Mercury’s Karapiro plant, Quest Integrity performed a number of assessments to determine whether it needed to be replaced, or if it still met criteria to be fit for service.
By Vitor Lopes Garcia and David Osuna
â— Peer Reviewed This article has been evaluated and edited in accordance with reviews conducted by two or more professionals who have relevant expertise. These peer reviewers judge manuscripts for technical accuracy, usefulness, and overall importance within the hydroelectric industry.
When personnel at the 96-MW Karapiro hydropower plant removed one of the four spillway gates from service in late 2014, they discovered the lower section of the gate had undergone permanent deformation and had suffered corrision on the internal and external gate surfaces. Concerned about the ability of the gate to operate reliably when returned to service, plant owner Mercury (formerly Mighty River Power) commissioned a fitness-for-service (FFS) assessment. This article covers the methods used to determine if the spillway gate was fit for continued service and the results of the assessment.
Background on the situation
Commissioned in 1947, the Karapiro hydro plant is the last in a series of nine power stations located on the Waikato River on the North Island of New Zealand. The facility houses three vertical Kaplan turbine-generator units and has a net head of 30 m. The dam is a 52-m-tall, 335-m-long concrete arch structure, with a spillway located on the right abutment.
The spillway has four gates, each of which is 6.1 m wide and rated to pass 3.77 cubic meters per second of water. The gates are of the Stoney roller design and were originally operated one-at-a-time by a gantry crane until 1979, when a dedicated winch system was installed, allowing the four gates to be opened simultaneously. In 1998, the staunching bars were replaced with music note seals and the gates were reinforced for seismic events.
In late 2014, Gate 2 was removed from service for installation of bird netting. After removing the gate, personnel with Mercury determined that the lower section had undergone permanent deformation. Corrosion was also reported on the gate’s internal and external surfaces, prompting an FFS assessment, which was undertaken by Quest Integrity.
Inspecting the gate
Two on-site visits were carried out by Quest Integrity engineers, with the aim of quantifying the level of superficial corrosion, using a Panametrics 36DL-Plus UT thickness meter. These measurements (see Table 1 on page 38) were taken at four locations between the spillway gate’s seven girders, designated A, B, C and D, with A being the lowest location and D being the top section of the gate (see Figure 1).
|This schematic shows the position of Spillway Gate No. 2’s seven horizontal girders in relation to the four locations selected for measurement.|
Location A showed the highest level of metal loss with respect to the nominal gate thickness of 15.875 mm. Meanwhile, location D showed the minimum metal loss, with a value very close to the skinplate nominal thickness. It was found that the magnitude of metal loss was related to the location of the skinplate section with respect to the spillway gate height. The bottom sections showed higher degrees of metal loss compared to the top sections.
Finite element analysis
A finite element analysis (FEA) was performed by Quest Integrity with the purpose of estimating the stresses at the spillway gate due to hydrostatic loading. Two different scenarios were considered:
– Deformation load: To determine the possible maximum loading that caused the current deflection in the gate. An as-designed model with no corrosion was used; and
– Fitness for service: To demonstrate the current gate is fit for service in accordance with the plastic collapse and local failure guidelines of the American Society of Mechanical Engineers.1 A deformed model with superficial corrosion was used.
The geometry of the spillway gate was generated based on engineering drawings provided by Mercury and photographic documentation taken during the site visits. The models were generated using the finite element modeling software Abaqus CAE 6.12-1. Two different models were generated:
- As-designed model with no corrosion
- Deformed model based on reported vertical deflection with superficial corrosion.
To reduce the computational cost of the finite element analysis, some non-load-bearing features of the spillway gate were not included as part of the modelled geometry. These included the diagonal reinforcement plate between beams A1 and A2 (see Figure 1); latching racks; and rope retainers, guides and anchor brackets.
Based on engineering drawings provided by Mercury, the skinplate material was specified as BS 4360:1968: Grade 43A, with a minimum tensile strength of 430 MPa, yield strength of 230 MPa, and elongation of 20%.
Based on these properties, an elastic-plastic stress-strain curve was generated based on the Ramberg-Osgood methodology found in the ASME FFS-1 standard. 1
Loads and boundary conditions
The applied hydrostatic load to the two models was based on a maximum operating level of 52.9 m, per the Hydraulic Structures Hydrological Data Book.2
Boundary conditions were applied to the side joists to represent their interaction with the spillway gate slots. Only rotation with respect to their own axes was allowed.
Stress results were obtained for the two different scenarios considered. The highest level of stresses occurred at the bottom of the gate and was due to hydrostatic loading. The locations that exhibit the highest levels of tensile stresses are at the top plates of the two bottom I-beams. The deformed model with superficial corrosion undergoes a higher level of stresses, especially at the center of the top plate of beam A3.
Mercury requested an additional assessment to evaluate the possible effect of the gate deformation and corrosion on its operation. Therefore, reaction forces at the side joists were also extracted, using both models, from the results for comparison purposes.
The direction and magnitude of the extracted reaction forces was determined for both models, and no difference was found between the as-designed model with no corrosion and the deformed model with superficial corrosion.
|Deflection was discovered at the bottom of Spillway Gate No. 2 at the Karapiro facility in late 2014.|
Deformation load estimation
The stress results were subsequently used to estimate the required load to cause the current spillway gate deformation. The process of estimating this deformation load was done in two steps by:
1. Calculating the strain that corresponds to the reported maximum vertical deflection of 23 mm at beam A3; then
2. Calculating the necessary load to reach the strain level from Step 1.
A strain of 0.13% was found to be required to cause a vertical deformation of 23 mm at the center of beam A3.
For the deformation load estimation, the maximum operating load was incrementally increased to reproduce the observed configuration. It was determined that for a vertical deflection of 23 mm, a load 2.4 times the maximum operating load is required.
To determine the acceptability of the spillway gate for protection against plastic collapse, a FFS assessment was carried out.
An elastic-plastic stress analysis was performed on the as-deformed model with corrosion damage. The acceptability of the spillway gate using an elastic-plastic analysis was determined by satisfying the following criteria: global collapse and local failure.
Global collapse criterion
To satisfy the global collapse criterion, the finite element model was subjected to a load factor of 3.6 and is required to reach the converged solution. The load factor was determined based on Equation 1, in which the remaining strength factor (RSF), defined as the ratio of the limit load of the damaged component to the limit load of the undamaged component, is 0.9 (see Figure 2).
|The von Mises distribution of the converged solution, with a load factor of 3.6.|
Load factor coefficient () = 4.0 x RSF
The as-deformed model of the spillway gate subjected to a load factor of 3.6 reached a converged solution. Therefore, the as-deformed model with corrosion damage satisfied the global criterion for protection against plastic collapse following the ASME FFS-1 guidelines.
Local failure criterion
In accordance with the local failure criterion guidelines found in the ASME FFS-1 standard, the model that included the permanent deformation and surface corrosion is required to satisfy the failure criterion against the applied loading condition with a load factor of 1.5. High-stressed locations were identified for the assessment.
For the component to satisfy the local failure criterion, the total equivalent plastic strain must be less than the limiting triaxial strain (ÎµL). Total equivalent plastic strain at the assessed locations was extracted from the results of the elastic-plastic analysis with a load factor of 1.5.
The limiting triaxial strain was calculated based on the guidelines found in the ASME FFS-1 standard, using the following equation:
ÎµL is the limiting triaxial strain;
ÎµLu is the uniaxial strain limit;
Î±sl is the material factor for multiaxial strain limit;
Î±1, Î±2 and Î±3 are the principle stresses;
Î±e is the equivalent (von Mises) stress.
The acceptability per the local failure criterion was assessed by calculating the ratio between equivalent plastic strain and triaxial strain limit.
Ratio values in the modeled gate that were less than one satisfy Equation 2 (i.e. equivalent plastic strain is less than triaxial strain limit). As all the areas showed a ratio of less than 1, the spillway gate satisfied the local failure criterion.
Additionally, it was determined that under the current levels of deformation and metal loss, the most critical location corresponds to the top plate at the center of beam A3.
Based on the results of the FFS assessment, it was concluded that spillway gate 2 at the Karapiro Power Station was fit for service as it met both the global collapse and local failure requirements.
This assessment showed that the lack of design information (e.g., corrosion allowance was not known) can be overcome by engineering assessments that are compliant with recognized standards, allowing for an informed decision.
Vitor Lopes Garcia is a consulting engineer with Quest Integrity. David Osuna was a consulting engineer with the company at the time the work discussed in this article was completed.
1Fitness-for-Service Standard, Appendix B, American Society of Mechanical Engineers, New York City.
2Waikato Hydro System: Hydraulic Structures Hydrological Data Book, Mercury Energy, Auckland, New Zealand.
Osuna, David, “Engineering Assessment of Deformed Spillway Gate,” Proceedings of HydroVision International 2016, PennWell Corp., Tulsa, Okla., 2016.
Several articles have been published previously about gates, including spillway gates. Below are two articles from Hydro Review that provide further reading on the subject. Find them at HydroWorld.com:
Anami, Keiko, et al, “Method for Identifying Dynamic Instability of Tainter Gates,” Hydro Review, Volume 35, No. 6, July 2016,.
Peterson, Josh, and Peter Haug, “Preparing for the Worst,” Hydro Review, Volume 35, No. 5, June 2016,.