After receiving a telephone call from a representative of a utility, George traveled to visit one of the utility’s hydroelectric facilities. His goal was to assess the damage to an intake gate hoist house that occurred after an emergency shutdown of all the units in the powerhouse. During this situation, water had flooded the hoist house to a height of more than 1 meter.
Luckily, the damage in the hoist house was fairly minor. All the electrical equipment was waterproof and thus had not been damaged. The primary damage consisted of a broken latch on the steel door of the hoist house. During this flooding event, the water level in the concrete-walled hoist house had risen past the air vent, compressing air and consequently blasting open the door.
Before this flooding event, the utility had a “standard” hydroelectric intake design. It consisted of an upstream-sealing gate with a wire rope hoist mounted on a steel frame. The frame spanned the gate well in a room just below the intake deck. The galvanized steel floor grating was only a few centimeters above the reservoir full supply level. The hoist room was accessed down a set of stairs and through the steel door. The hoist house roof included a removable concrete cover that allowed access for a large mobile crane to lift the hoist and gate out for servicing.
The intake air vent was located just below the hoist house floor. This would allow the hoist house to be heated without the warm air escaping through the air vent. Unfortunately, what utility personnel failed to realize during design of the station was that this air vent would be almost fully blocked by water at full reservoir level and would be completely blocked by the rising water should a load rejection occur.
Utility personnel did recognize that occasional minor flooding of the hoist house floor to a depth of a few centimeters could occur in the event of a full turbine load rejection coinciding with a full reservoir. For this reason, personnel waterproofed the electrical equipment.
However, the flooding that occurred during this particular event had been much more severe than expected. Fortunately, no personnel had been working in the hydro plant intake during the incident.
What had gone wrong?
A fault on the transmission line caused all three units in the powerhouse to shut down at the maximum governor close rate, to avoid excessive generator overspeed. At the time of the shutdown, the reservoir was full. The intake gate remained fully open.
A factor personnel had not taken into account when designing the intake was the full effect of inertia in the water column from the intake to the turbine. The normal waterhammer had been calculated at 40%.
The pressure line was drawn as a straight line from the full reservoir level at the trashracks to the vertical shaft centerline (at a level 40% of the turbine head) above the reservoir level – all very normal. This line passed a few centimeters above the hoist house floor, and utility personnel assumed it indicated the flood level within the hoist house.
However, with this design the intake gate well will act as a small surge tank to mitigate the effect of the waterhammer wave. And because the area of the gate well is much smaller than the penstock area, the water will rise rapidly within the gate well and flood into the “surge tank” hoist house.
Another factor is the open head pond water surge on a full load rejection. However, because this surge is small and slow and should not coincide with the rapid waterhammer surge, it usually can be neglected. The combined surge flood level would be too difficult to calculate without resorting to a detailed computer analysis. Instead, it is far simpler to just increase the level of the hoist house floor, and the utility specified this alteration for all future standard hydroelectric intake designs.
The utility’s new standard, to be applied to all future hydroelectric developments as a result of this incident, calls for the hoist house to be located at intake deck level, several meters above the reservoir level. The air vent will be lifted to just below deck level, where it would be fully effective at all times. Finally, the area of the gate well has been increased to avoid the possibility of gate catapulting on control failure.
Who would have thought of the gate well acting as a mini surge tank?
This utility has an effective safety program in place and prides itself on its low accident rate. All of its hydroelectric plant designs are scrutinized for safety, and redundant controls are used wherever possible to avoid operator error. Operating procedures are detailed, with error scenarios investigated in all new designs.
Despite all these precautions, this incident took everyone at the utility by surprise. We all can learn from such incidents.
– By James L. Gordon, B.Sc., hydropower consultant
We want to hear from you!
Do you have suggestions for magazine articles and departments you believe would be helpful to colleagues? Please send your ideas to the Editor, E-mail:[email protected].