Thrust Bearing Temperature Rise on an Extraction-Condensing Turbine

Diagnosis: water induction into the turbine through the high-pressure main drains.

The chain is: drain pot dump valve left shut after the maintenance tag → drain pot fills → water carries over into the HP steam chest and the first stages → the indications follow mechanically.

Weight given to each indication, highest first:

1. Inlet steam temperature down 135°F at constant pressure. This is the direct measurement of what entered the machine. 760°F at 880 psig is still superheated (saturation at 880 psig is about 525°F), so what arrived was wet, cooled steam plus entrained water, not a solid slug - which explains why the unit is still running rather than tripped on a water-hammer shock.
2. Drain pot high level plus the unverified dump valve. This identifies the source rather than the effect. Together with the temperature drop it takes the diagnosis from "possible" to "almost certain."
3. Thrust bearing temperature and axial displacement together. Water entering the stages changes the stage pressure distribution and the axial thrust balance; a rotor designed for a small net thrust in one direction can see that thrust reverse or multiply. Rising metal temperature and rising axial displacement in the same direction, at the same time, is a thrust-loading signature, not a bearing lubrication problem.
4. Condenser vacuum loss. Water entering the machine flashes in the last stages and loads the condenser with vapour and non-condensables; vacuum falls. It also raises exhaust hood temperature and can lift the exhaust hood safety diaphragms (if fitted) - check them.
5. Gland seal header pressure drop. Gland seal pressure is referenced to the machine and its exhaust; a collapsing vacuum and steam-side upset moves the seal regulator's operating point. On its own it says little, but it fits the picture of a machine whose internal pressures have all shifted.

What does not support a simple bearing-failure theory: vibration is only up 0.06 in/s, and the temperature rise is smooth over 35 minutes rather than stepwise. A failing bearing typically shows vibration and temperature together.

Immediate actions, in order (target: complete within 10 minutes):

1. Open the HP main drain pot's dump line to the condenser and prove flow - manually, physically, not by reading a tag. Open the drain pot bypass and the HP header free-drain to atmosphere as permitted by the drain system design. Getting the water out of the header is the whole fix.
2. Reduce load to minimum (or to the manufacturer's minimum stable load) to cut steam flow, stage pressure ratio and thrust.
3. Back off extraction to the deaerator if that extraction is on the same line, to stop any reverse flow path into the machine.
4. Confirm gland seal pressure is above the minimum on the seal regulator and that the seal steam is dry.
5. Notify the control room and start a two-minute log: thrust metal temperature, axial displacement, vibration, vacuum, inlet steam temperature, drain pot level.

Trip now, not at 250°F, if any of these occur:

• thrust metal reaches 240°F, or continues rising at more than about 5°F per 5 minutes after the drains are opened;
• axial displacement reaches 0.016 in. or moves another 0.001 in. in either direction;
• overall vibration rises more than 0.10 in/s above baseline, or any metallic contact/rubbing noise is heard;
• inlet steam temperature drops below about 100°F of superheat (roughly 625°F at 880 psig) or any water-hammer is heard;
• exhaust hood temperature exceeds its alarm.

A thrust bearing that fails in service takes the rotor, the diaphragms and the seals with it. The reason to trip early is that the cost of a controlled trip is hours and the cost of a bearing failure is months. The temperature is a lagging indicator - the babbitt is already being damaged before the thermocouple responds.

Inspection (in this order):

1. Thrust bearing: babbitt condition, contact pattern, crush, clearance and end float versus specification. Photograph and measure, do not just "looks fine."
2. Axial position: re-establish the rotor's cold position and compare with the commissioning datum. A shifted rotor position is the strongest evidence of a thrust event.
3. Borescope the first stages and the blade path for water- droplet erosion on the leading edges, blade untwist, and any rub marks at the seals. Water entering at 900 psig leaves a distinctive erosion pattern on the first two rows.
4. Diaphragm and tip clearances - check for rub at the seals, which would show as a step.
5. Rotor bow check on turning gear (or slow-roll indicator runout) before any restart.
6. Drain pot internals and level controls: verify the level transmitter is calibrated and the dump valve actuator strokes fully. A pot that overflows because the level control is blind is a different failure from a closed manual valve.
7. Exhaust hood safety diaphragms on the condenser.

Procedural change:

The root cause is not a valve; it is that a drain was taken out of service by a maintenance tag with no check that it went back into service, and no operator-verifiable status. Fix it three ways:

• Add a positive, physical line-up verification to the tag procedure: valves in safety-related drain service carry a numbered car seal or locking device, and the tag cannot be cleared without a second signature confirming the valve is open and passing.
• Add inlet steam temperature rate-of-change and drain pot level to the turbine's alarm/trip matrix with an annunciator that is not buried - the machine gave 35 minutes of warning and nobody acted on the temperature trend.
• Change the restart checklist so that after any outage on the drain system, the unit cannot be loaded above minimum until the drains are proven flowing and the inlet steam temperature is stable for 30 minutes.

Nine months ago the same signature occurred and a bearing was inspected. Inspecting the bearing treats the symptom; this time, close out the drain line-up as a formal management-of-change item.