Foreign Object Retrieval in Nuclear Systems
By Walt Kesler, Power plant maintenance manager, nuclear-qualified. Reviewed by Amanda McCallister, editor.
Underwater Inspection | Commercial Diving Contractors Alabama The call came in at 02:00. A 40-pound plate of 316 stainless had torn loose from the trash rack and was sitting on the bottom of the forebay, six meters down, directly in front of a circulating water pump. The plant was at 85% power and the cost of a forced trip runs into the millions. Foreign object retrieval in a nuclear plant is not like fishing a wrench out of a tank. You've got radiological controls, emergency cooling implications, and a regulator who wants to know what you did, in writing, before you start. We've pulled cable ties and chemistry bottles and one time a full-sized boot out of those bays. The boot isn't a good story. The dive plan is.
Assess Before You Splash
The first rule is always look before you touch. We run the plant's CCTV over the area, and if the water's turbid, we mount an Aqua-Vu HD camera on a pole and slide it down the trash rack guides. That gives us orientation, snag points, and clearance. Without that, you're suiting up a diver on a guess. Then we pull the flow charts. Water velocity in the forebay at our plant runs 0.3 to 0.5 meters per second, give or take, and we can throttle the circulating water pumps to bring that down. But reducing flow changes the condenser vacuum, and the operators watch that number like a hawk. So we coordinate every adjustment. And whatever you do, don't let the shift manager talk you into 'just poking it with a pole.' I've seen a pole become the second foreign object. That's a true story.
Pick the Tool for the Job
Most foreign objects in intake systems are not radioactive. They're process debris: paint chips, bolts, chunks of gasket. But we treat everything as potentially contaminated until proven otherwise. So tool choice depends on reach, size, and dose. For small objects, a long-reach grapple or a handled magnet works. For larger pieces we've used a 4-inch airlift dredge to clear sedimentation, and a hydraulic grab to lift a 75-kilogram bearing race that had somehow fallen off the intake screen lift gear. Those are 30-minute jobs. For the really tricky ones, we bring in an ROV. A Seabotix LBV-150 handles current up to about 3 knots and can work in zero visibility with its sonar. Last cycle we used one to fetch a 30-centimeter crescent wrench out of Unit 2's intake bay. The operator stayed clean and dry on the deck the whole time. But there are limits. When an object weighs more than the ROV can carry or it's wedged in between the screen guides, you need a human hand. That's when we call in the dive team.
The Dive Itself
Sending a diver into a nuclear intake bay is never the first choice, but when it's the right choice, we do it by the book. The NRC doesn't regulate recreational diving, but they absolutely care about plant safety. So we write a formal dive plan, review it with operations, and get approval on a document that names the controller, the tender, the diver, and the standby diver. The diver wears a Kirby Morgan helmet with hard-wired comms. We use a full tether with an air supply. The standby suit is always in the water—that's a commercial diving requirement and it's a good one. I don't care if the water looks like a swimming pool, that rule never bends. We monitor dose in real time. The diver wears a TLD and an electronic dosimeter. If the area is near a heat exchanger outlet, dose rates can be elevated. One retrieval had us sitting at about 0.5 millisieverts per hour, so we rotated divers every 20 minutes. Each of them did two dives. Total dose to the team was about a third of the limit. That's ALARA—as low as reasonably achievable—made real.
Verify and Document
Bringing the object to the surface is not the finish line. We take a post-dive video of the intake bay, run a sonar sweep, and then we physically check the trash racks and debris baskets by hand. That's how we found the mounting bolts that fell off with the plate. Then we write it up. Date, time, plant conditions, radiological dose, equipment used, crew names, object description, and chain of custody. If it happened at power, the report goes to operations and the regulatory affairs group. They will ask about root cause, so we keep the object in a labeled evidence bag until they make the call. The whole cycle—assess, retrieve, verify, document—takes anywhere from six hours to a full day. You can't rush it. A nuclear plant will wait for a dive team, but it won't wait for a careless one.
Frequently Asked Questions
Why can't we just shut down and drain the bay? Draining a forebay means a multipress to shut down that can take a week or more. In a multi-unit plant, you may be able to isolate one unit, but the cost is still in the millions. A dive team can often retrieve the object at power with no loss of generation. How deep are these retrievals? Depends on the plant. Our circulating water intake bays are about 8 meters at the bottom. Some plants, especially those on rivers, have intake screens set at 12 to 15 meters. Beyond 10 meters, you start doing saturation-style bottom time calculations. What qualifications do I need for this kind of work? Commercial diving certification is the baseline, plus site-specific training in radiological controls, tagging, and lockout/tagout. You also need to pass a dive physical that accounts for a nuclear environment. How do you know you got everything? Visual confirmation, sonar, and a magnet sweep are the usual steps. The real proof is running the screening system back up and checking for restrictions. If the screens run smooth, you're done.