Operators at a 1,259 MW Alabama reactor swapped to a backup hydraulic pump for a routine test, and the pressure surge it sent down the line snapped three bolts off a shutoff valve and shut the unit down twice

Hydraulic fluid was already spreading across the grating before anyone in the control room knew what had happened.

Somewhere above it, on a flanged joint the size of a dinner plate, the fasteners had let go.

The pressure reading on the control circuit fell.

The reactor was at full power at 1353 Central time on the twelfth of May in 2025 when the turbine tripped on loss of control fluid and the reactor scrammed.

Then the governing valves slammed shut.

Steam to the turbine at a boiling water reactor in northern Alabama cut off in an instant.

Then the whole thing happened again, before the crew had finished cleaning up after it the first time.

Why losing a fluid supply stops a reactor

In a boiling water reactor the turbine sits in the direct steam path out of the reactor vessel. There is no separate loop between them.

What holds the main governing valves open against heavy spring force is a bath of pressurized control fluid.

That arrangement is deliberate. Lose the fluid and the springs win, the valves close, and steam stops reaching the turbine without anyone deciding anything.

It is a fail safe, and it worked exactly as designed.

The reactor protection system watches that circuit. When pressure drops below its setpoint the scram fires automatically, and the control rods go in.

None of that is a malfunction. The malfunction was upstream of it, in a fitting nobody was watching.

The fitting that held the whole circuit

The component is a shutoff valve carried on a bolted bracket inside the turbine building, part of a circuit that runs at continuous high pressure whenever the plant is making power.

It is not a large thing. It is a valve body, a flange and a handful of fasteners.

Every one of those fasteners sees the pressure of the circuit plus whatever the turbine’s own vibration puts into the bracket, and it sees them every single day, for decades.

Bolts do not usually fail from one bad moment. They fail from millions of small ones, and then from one that is slightly larger than the rest.

When a fastener cracks that way it gives no outward sign until it has already parted.

By then the joint it was holding is a leak path.

What the operator told the regulator

The reactor was at full power at 1353 Central time on the twelfth of May in 2025 when the turbine tripped on loss of control fluid and the reactor scrammed.

Operators stabilized the plant. At 1407, with the unit shut down but still hot, a second automatic trip fired.

That one came from a brief drop in reactor water level while crews were manually opening and closing a main steam relief valve. Level was held afterward on a feedwater pump.

The event notification filed that evening called the first trip not complex, and said every system responded normally except the turbine bypass valves.

The later event report named the probable initiator, and it is the part worth reading twice. Operators had put an alternate control fluid pump into service for testing, which sent a pressure surge of roughly 54 psig through the circuit.

That surge is the most likely reason three bolts broke.

What the bypass valves could not do

Bypass valves are what separates this from a routine turbine trip.

Their job is to divert steam straight to the condenser when the main turbine goes offline, so pressure does not spike back toward the vessel.

They also run on the same control fluid. So the leak that shut the turbine took them with it, and the plant fell back on relief valves and steam line drains instead.

That is the sequence the crew was working through when the second trip came, and it is documented in the regulator’s own event record.

What is still open is the cause evaluation. The utility has said a supplement will follow once it is finished, so the fatigue question has no answer yet.

Small parts have taken reactors offline before, from swarms of jellyfish in a cooling intake to a handful of fasteners on one bracket.

What three bolts say about a fleet built on small parts

No release occurred. Nobody was hurt. The automatic systems did the thing they exist to do, twice.

The uncomfortable part is that the trigger was a planned test. Swapping pumps is routine work, and the circuit answered it by breaking.

Every valve, seal and fastener in a plant of this age has been through the same cycles, and inspection intervals are written around what engineers expect to find, not around what a surge can do to a fastener already carrying a crack.

Meanwhile grid operators are leaning harder on firm generation, and the emergency orders keeping two gas and oil fired Pennsylvania units running past their retirement date show how thin the margin has gotten.

The unit came back. The written record of what broke runs to a few pages.

A 1,259 megawatt machine was stopped twice in a quarter of an hour by three fasteners on a shutoff valve, and that is the number worth carrying forward.