CPP hub oil loss from a cracked blade foot, not a blade seal (Maunalei)

Propulsion & ThrustersMain EngineMAN Energy Solutions Mark 5 CPP; MAN B&W 7L70MC-CEngineering alarm, CPP hub head tank found empty

Contents
  1. What happened
  2. Diagnosis
  3. Root cause
  4. Timeline / response
  5. Lessons
  6. Comments · 0

What happened#

Containership, MAN B&W 7L70MC-C two-stroke ME directly coupled to a 5-blade MAN Mark 5 CPP. On passage to Anchorage an alarm came up: CPP hub head tank empty. The crew reduced pitch and refilled the tank. Everyone suspected a failed blade seal.

On the way to drydock the leak got worse and the oil on board ran out. On the CPP maker's advice, the crew topped up with fresh water to keep pressure on the shared lip seal. Water got into the stern tube oil, its temperature rose, and the ME was stopped. The ship was towed in. No injuries.

Diagnosis#

  • Loss rose with load: 1 L/h at full ahead, 0.45 L/h at three-quarters ahead.
  • Hub and stern tube LO have separate head tanks but share the forward lip seal, which works because the hub side is at slightly higher pressure. If hub head is lost, stern tube oil can flow past the seal into the hub.
Hub, servo and stern tube head tanks. Hub and stern tube oil meet at the forward seal. Image: NTSB
Hub, servo and stern tube head tanks. Hub and stern tube oil meet at the forward seal. Image: NTSB
  • Diver survey: fractures at the base of blades no. 2 and no. 4.
  • In drydock, oily water seeped from the no. 4 blade foot near bolt holes 4.6 and 4.7: a through-thickness crack. The no. 2 cracks were surface only.
Blade no. 4 foot (pressure side): the through crack runs across the bolt hole counterbores. Photo: US Coast Guard
Blade no. 4 foot (pressure side): the through crack runs across the bolt hole counterbores. Photo: US Coast Guard

Root cause#

High-cycle fatigue. Ratchet marks showed the cracks started at the edges of the bolt hole counterbores. A cross-section of one counterbore measured a radius of about 0.62 mm; the design radius was 0.8 mm.

The material was also out of spec: Charpy 9.5 J against 21 J at 10 °C, low tensile, yield and elongation, silicon 0.17 % against 0.10 % max. So the no. 4 blade was more prone to fatigue. MAN's FEA showed high stress at the blade foot under the large load changes of normal service. No cracks were found on similar propellers, so NTSB calls this likely an isolated case. MAN later raised the counterbore radius to 4.0 mm, which gave about 40 % more safety factor.

Timeline / response#

TimeEvent
4 Aug, 1723Alarm; hub head tank empty. Pitch reduced, tank refilled.
7–8 AugAnchorage: can't be fixed alongside. Leaves for drydock in Portland.
11 Aug, 0856Leak worse; rpm and pitch reduced.
11 Aug, 1834Water in stern tube oil, high temp. ME stopped.
15 Aug, 0437Arrived shipyard under tow. Diver finds blade cracks.
17–29 AugDrydock: DP test, no. 4 replaced with the spare, no. 2 weld-repaired.
29 Aug–19 SepIn service limited to 43 % MCR, 90 rpm (normal 108). Then laid up.

Lessons#

  1. Don't assume a blade seal. Here the hub oil loss was a through crack in the blade foot. Log the loss rate at each load and get a diver survey or drydock early.
  2. Watch the hub and stern tube head tanks together. Trend stern tube oil temperature too. If hub head drops, oil and water can cross the shared lip seal.
  3. Keep enough hub and stern tube oil on board for a long leak. Fresh water kept seal pressure but got into the stern tube oil.
  4. Reduce rpm and pitch. Less load means less leak and less cyclic stress on the blade foot.
  5. When blades are off, DP-test the blade foot around every bolt hole and check the counterbore radius against the drawing. For new blades, ask for material test results (Charpy, tensile, chemistry) against spec.
  6. Know what spare blades you have before drydock. There was only one; no. 2 was weld-repaired, which class rules don't normally allow.

Source: NTSB report MIR-24-11 — Loss of Propulsion aboard Containership Maunalei.

Comments · 0

No comments yet.

Sign in to comment, vote and confirm fixes.Sign in