MARINE SAFETY INVESTIGATION

REPORT 148

Coral Sea
on 26 and 27 July 1999
 Report No 148

Navigation Act 1912

Navigation (Marine Casualty) Regulations

investigation into

the shift of cargo on board the Singapore flag bulk carrier

PADANG HAWK

in the Coral Sea

on 26 and 27 July 1999

Issued by the

Australian Transport Safety Bureau

September 2000

ISBN 0 642 20032 7

Investigations into marine casualties occurring within the Commonwealth's jurisdiction are
conducted under the provisions of the Navigation (Marine Casualty) Regulations, made
pursuant to subsections 425 (1) (ea) and 425 (1AAA) of the Navigation Act 1912. The
Regulations provide discretionary powers to the Inspector to investigate incidents as
defined by the Regulations. Where an investigation is undertaken, the Inspector must
submit a report to the Executive Director of the Australian Transport Safety Bureau
(ATSB).

It is ATSB policy to publish such reports in full as an educational tool to increase aware
ness of the causes of marine incidents so as to improve safety at sea and enhance the pro
tection of the marine enviroment.

To increase the value of the safety material presented in this report, readers are encouraged
to copy or reprint the material, in part or in whole, for further distribution, but should
acknowledge the source. Additional copies of the report can be obtained from:

Inspector of Marine Accidents

Australian Transport Safety Bureau
PO Box 967
Civic Square 2608 ACT

Phone: 02 6274 6088

Fax: 02 6274 6699
Email: [email protected]
Internet address: www.atsb.gov.au
Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

Sources of information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

Narrative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

Padang Hawk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

The nickel ore trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

The incident . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

Comment and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

Initial Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

Sample tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

Liquefaction in nickel ore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

The cargo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19

Representative samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

The adapted flow test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

The Japanese experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23

Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

Submissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27

Padang Hawk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31

Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33

Appendix C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35

Figures

1. Padang Hawk in Townsville . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4

2. Padang Hawk holds and tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8

3. Typical cross-section of laterite deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10

4. Laterite deposits in New Caledonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

5. Padang Hawks passage to Townsville . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12

6. Cargo, hold number 1 (two views) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

7. Cargo, hold number 1 (two views) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

8. Cargo, hold number 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

9. Padang Hawk events and causal chart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26

Summary Some water was pooling on the surface of

the cargo in number 1 hold. The cargo in
the forward holds appeared to be flowing
with the movement of the ship.

Between 17 and 23 July 1999, the

The master reduced speed and altered
Singapore flag bulk carrier Padang Hawk
course to put the wind and seas on the
loaded a full cargo of nickel ore from
ships port quarter. Ballast was then
barges at Kouaoua, New Caledonia. Late
pumped to correct the list. The ships
on 23 July, the ship sailed for Townsville,
course was maintained so that it entered
Australia. During the passage, Padang
the inner route of the Great Barrier Reef
Hawk was subjected to rough seas and
by Grafton Passage rather than the more
rolled heavily. At about 2200 on 26 July,
southerly Palm Passage.
the ship developed a 15 list to port.
The ship finally arrived safely in
A quick examination of the holds showed
Townsville at 2000 on the evening of 28
that the cargo in four of the five holds had
July.
settled and appeared to have liquefied.

Sources of

information

The master and crew of Padang Hawk

Queensland Nickel Pty Ltd

Australian Maritime Safety Authority

LLP Ltd, Information Service Department

International Transport Workers Federation

Nippon Kaiji Kyokai Survey Department

Meteorological Office of New Caledonia

SGS Australia Pty Ltd

Professor Tamaki Ura, University of Tokyo

Marine Casualties Investigation Board,

France

Queensland Department of Mines and

Energy

Acknowledgement
Portion of chart Aus 4060 reproduced with
permission of the Hydrographic Office,
RAN.

7
FIGURE 2.

Padang Hawk holds and tanks

Narrative Singapore. Padang Hawk had been issued

with a Safety Management Certificate on
28 April 1998 by NK.

Padang Hawk is a regular caller at

Padang Hawk Townsville with cargoes of nickel ore.
The Singapore flag Padang Hawk (figs. 1 The ship carried a cargo of nickel ore from
and 2) is a 46 635 tonne deadweight the port of Nakety, New Caledonia to
geared bulk carrier, owned by Singa Star Townsville on the previous voyage,
Pte. Ltd. of Singapore. The ship was built number 903. Padang Hawk was to return
in 1995 by Mitsui Engineering and Ship to Kouaoua, New Caledonia, on voyage
Building of Japan and is classed with number 904 to load another cargo of
Nippon Kaiji Kyokai (NK). nickel ore for shipment to Townsville.

Padang Hawk is 189.8 m in length overall, The master and mate joined the ship at
has a beam of 31 m and a summer draught Townsville at the completion of the voyage
of 11.6 m. Propulsive power is delivered number 903. They were both experienced
by a six-cylinder Mitsui MAN B&W in bulk carrier trades and the master had
6S50MC slow-speed diesel engine devel previously commanded a smaller bulk
oping 6 532 kW. The main engine drives a carrier carrying nickel ore on coastal
single fixed pitch propeller, which pro voyages around New Caledonia.
vides a service speed of

14 knots. Electrical power is provided by

The nickel ore trade
three Daihatsu 6DL-20 generators, each In 1986, Queensland Nickel Pty Ltd
producing 480 kW. (QNPL) started importing limonitic nickel
ore for its nickel and cobalt refinery at
Yabulu, 25 km north-west of Townsville.
The engine room and accommodation
The Yabulu refinery utilises nickel ore
superstructure are located at the after end
from the Philippines, Indonesia and New
of the vessel, (aft of frame 37). There are
Caledonia. At the time of the incident, in
five cargo holds; with the exception of
excess of 27.4 million tonnes of ore had
number 1 hold at 17.6 m, all are 20.8 m in
been carried in 577 vessels into the port of
length and extend to the collision bulkhead
Townsville.
at frame 216. The ship has four cranes to
service the five holds when ships equip
The Philippines, Indonesia and New
ment is required for loading or discharg
Caledonia are estimated to contain 35% of
ing.
the worlds nickel resources. The nickel
occurs in laterite deposits, which have
At the time of the incident, the shipowners been formed by the deep weathering of
held a current International Safety ultramafic1 rocks. Typically, the laterite
Management Code (ISM Code) document deposits have an upper zone in which the
of compliance for bulk carriers, issued by nickel is combined with iron oxides
NK on 24 March 1998, under the authority (limonite zone), and a lower zone where
of the Government of the Republic of the nickel occurs in complex magnesium
1 Igneous rocks containing a high percentage of iron and magnesium based minerals.
9

rich silicates (saprolite zone) (fig. 3). excluding pure saprolitic ores. The ore is
Limonite generally contains around 1.5% discharged from ships in the port of
nickel, and saprolite around 2.5% nickel. Townsville then transhipped by train to the
Yabulu refinery where an ammonia leach
Laterite deposits are mined using open process is used to refine the ore.
cut methods. The limonitic and saprolitic
ores are mined together and the ores are Saprolitic nickel ores are refined using a
graded and stockpiled separately. Mine smelting process. A large proportion of
site stockpiles generally consist of the saprolitic ores mined in the
windrows where the ore is dumped in Philippines, Indonesia and New Caledonia
rows of small piles and turned regularly to are shipped to Japan.
aid the drying process. Once the ore has
been dried sufficiently, it is transported to The incident
larger stockpiles at the ship-loading facili
Padang Hawk sailed from Townsville on
ty. The ore in many shallow water ports,
the evening of 13 July and arrived at the
including Kouaoua on the east coast of
anchorage at Kouaoua, New Caledonia at
New Caledonia (fig. 4), is loaded onto
1224 on 17 July. Loading of ore from
barges at the ship loading facility and
shore barges commenced at 1440. The
transported out to ships at anchor in
ships cranes were manned by shore labour
deeper water. The ships cranes are then
and loading from the barges was carried
used to load the ore from the barges into
out from about 0500 each morning until
the cargo holds.
about 2030 in the evenings. With the
exception of one morning, when there
The refining processes for limonitic and were a few hours of light drizzle, the
pure saprolitic ores are different owing to weather was fine throughout the loading
their differing mineral properties. QNPL operation. Soundings of the cargo holds
utilises ore largely from the limonite zone, for water were taken each day and no bilge
FIGURE 3.

Typical cross-section of laterite deposits

10

water was recorded. Caledonia. The planned route was to take

Padang Hawk north of Rcifs
During the loading operation, the ships dEntrecasteaux, west through the Coral
crew noticed that some grabs of cargo Sea, north of Bampton Reef and Marion
had water running from them as they were Reef, to Palm Passage and Townsville (fig.
lifted out of the barges and into the ships 5).
holds.
At 1700 on 24 July, the bilges of all holds
The cargo loading was completed at 2140 were pumped dry. The ships log-book
on 23 July. The crew stated that the records the course as 303 with the ship
hatches were then closed and secured by rolling heavily in a south-easterly swell.
the hatch cleats. Following a draught
survey, a pilot embarked and Padang The passage proceeded as planned. Just
Hawk sailed from Kouaoua at 2307, at a before 2400 on 24 July, north of Rcifs
draught of 11.85 m fore and aft. dEntrecasteaux, Padang Hawk altered
course to a true heading of 265, making
After the pilot disembarked in the early good a speed of 13 knots. The wind was
hours of 24 July, the master set a northerly noted as being east-south-east, force 5, and
course along the east coast of New the sea was described as rough. The

FIGURE 4.

Laterite deposits in New Caledonia

11

 FIGURE 5.

Location of
NT incident

Qld
WA

SA

NSW
Padang Hawks passage to Townsville

ACT

Grafton passage Vic

Tas

Palm

12

passage x
x Noon position 26/7
Position of incident

Kouaoua
vessel was rolling and pitching heavily in The master sent the mate, boatswain and
a south-easterly swell. At 0900 on some deck ratings to check the hatches
and the state of the main deck. The crew
25 July, the bilges in all holds were again checked and found the cargo hatches and
pumped. The rough conditions caused the the small hold access hatches secured.
vessel to roll and pitch heavily, particularly They opened an access hatch to each of
from 1600 onwards on 25 July. the holds in turn. They found that the
cargo in all but number 5 hold had settled
At 0830 on 26 July, the bilges in all holds and shifted to port. The cargo in the first
were pumped. The vessel was still rolling 3 holds appeared to be semi-liquid, like
heavily in the southerly swell and strong melted ice cream as the boatswain
winds that had veered to south to south described it.
westerly. At noon on 26 July, Padang
Hawk was in position 18 11 S 153 56 The master decided to ballast starboard
E, about 100 nautical miles east-north-east side tanks to correct the list. Numbers 3
of Marion Reef. Over the preceding day, and 5 starboard topside tanks were filled.
the average speed had been 12.7 knots.
The master sent a radio message to the
During the afternoon, the wind strength designated person, as described in the
was logged at force 5 with regular nota ships ISM Code documentation. The
tions in the logbook concerning the ships message detailed the situation and the
heavy rolling. Throughout the day, from remedial action taken. A message was
time to time, seas broke over the deck and also sent to the ships agent in Townsville,
hatch covers. By 2000, the wind was who informed the Australian Maritime
logged at force 67 and the vessel was Safety Authority of the ships message and
rolling heavily. The hold bilges were the report that the cargo had liquefied.
pumped at 2000 and again at 2100.
At 0145, the master received a reply from
At 2200, or a little before, Padang Hawk the vessels owners advising him to use
suddenly developed a 15 list to port. The double bottom tanks to correct the list.
master, who was in his cabin, immediately The message noted that countering lists by
went to the bridge and joined the second using topside tanks had caused vessels to
mate and lookout. The master altered capsize and it continued:
course from 265 to 295 to bring the
wind and sea on to the port quarter and Although your vessel is having very high
reduced the engine revolutions from 110 GM due to dense cargo, still high risk of
RPM to 100 RPM. cargo shifting to one side with the roll is
high.
The master mustered all the crew. In the
dark, with seas breaking over the port side The ships list had been reduced to about
and the ship rolling heavily about the 5 by the early morning. The strong wind
angle of the list, there was general concern and heavy swell continued and seas broke
and a very high level of apprehension regularly over the vessels quarter.
amongst all the crew. Because of the heavy swell, the master
decided to wait before turning toward

13

Palm Passage. The cargo hold bilges were Grafton Passage. Even with reduced
pumped at regular intervals throughout the engine revolutions, Padang Hawk made
day. The disposition of ballast was adjust good speed. In the early hours of 28 July,
ed in accordance with the advice from the the ship was approaching Grafton passage
owners. and, by 0400, it was safely in the calmer
waters inside the Great Barrier Reef.
At noon on 27 July, Padang Hawk was at
position 17 16 S 148 58 E. The In the afternoon of 28 July, Padang Hawk
weather had not abated and the master anchored off Townsville and final adjust
judged it unsafe to try and make Palm ments were made to the ballast to min
Passage, as this meant altering course to imise the list and reduce the ships
bring the wind on the port beam. He draught. The anchor was weighed at 1737
decided to maintain the course with the and the pilot boarded at 1820. The ship
wind astern. Fortunately the course of arrived safely alongside in Townsville at
295 took the ship directly towards 1928 on 28 July.
15

FIGURES 6 & 7

Cargo, hold number 1 (two views)

16

Comment and the hold. The holes left in the surface of

the cargo after each grab was removed,
maintained their approximate size and
analysis shape without excessive slump in the sur
rounding cargo.

The hatch covers aboard Padang Hawk are

Initial inspection hydraulically operated steel covers, with
When Padang Hawk arrived alongside at two sections folding and opening forward
Townsville in the evening of 28 July 1999, and two sections folding and opening at
investigators inspected the ships cargo the after end of the hatch. The hatch seal
holds. Cargo in holds 1, 2, 3 and 4 had consists of neoprene rubber packing.
settled and shifted. In each hold, the Locking dogs fixed at intervals around
surface of the ore showed that a portion of their periphery ensure that the hatch
the ore had apparently liquefied into a covers are watertight.
glutinous thick slurry. In number 1 hold,
water was pooled on the surface of the ore Padang Hawk presented as a well-main
and the whole cargo in this hold had tained ship with hydraulic deck equipment
settled to be almost level. The inner sides in good order. Examination of the hatch
of the hold showed the cargo still covers, coamings and other openings into
hanging to port, about 5.5 m below the the cargo holds, showed no obvious sign
deck level, and there was evidence of sig of any ingress of water. The neoprene
nificant mud splash (figs. 6, 7, and 8). packing on the hatch covers was in good
The cargo in holds 2, 3, and 4 showed condition and showed that it had sealed
varying amounts of liquefaction with a against the coaming channel bar. Inside
central mound of unaffected cargo sur the coamings there was no obvious sign of
rounded by cargo that had liquefied. any ingress of water, although some water
was lying in the trackways associated with
In number 5 hold, the cargo retained the the hatch coamings.
form in which it was loaded, with cargo
forming a flattened pyramid which Sample tests
reached to the level of the bottom of the The state of the cargo in the forward four
topside hopper tank, 3.5 m below the deck holds was caused by excessive moisture in
level. The disposition of the cargo in the nickel ore and indicated two possibili
number 5 indicated the way in which the ties: either seawater had gained access to
cargo had been trimmed and what the the holds in the rough weather, or the
cargo must have looked like in holds 1, 2, cargo had been loaded in an excessively
3, and 4 after the completion of loading. moist condition.

Test grabs were taken from number 1 If the moisture present in the cargo had
hold on the morning of 29 July, using the been due to seawater ingress through
shore-based discharge crane. The cargo poorly secured or failed hatch covers, it
lifted in the test grabs remained in a bulk would have been salty. Conversely, if
solid form, with some water or slurry the cargo had been loaded in an excessive
running from the surface when lifted from ly moist condition, the origin of the water

17

would have been rain in the mining and/or the sub-surface sample number five
stockpiling phases of the operation and hold, 38.9 %.
would consequently have been fresh.
To obtain the flow moisture point of the
Investigators took samples of the nickel nickel ore sample, SGS used the Flow
ore from just beneath the top layer and Table method described in appendix D
between 2030 cm below the cargo of the BC Code and then derived the value
surface in number 1 and number 5 holds. for TML. This resulted in a TML of 29.2
The ore was a red sand or dust with larger %. All of the ore samples represented had
pieces of solid greenish rock interspersed moisture contents that significantly
throughout. These larger pieces were gen exceeded the TML determined using the
erally less than 5 mm in size. IMO standard flow table test.

The samples were sealed in tins and deliv Liquefaction of nickel ore
ered to SGS Australia Ltd, at Port Kembla Nickel ore has not been considered to be
for analysis to determine: prone to liquefaction. This cargo is regu
larly carried with total moisture contents
whether or not the origin of the mois in excess of 36%. On this occasion,
ture was sea or rain water, however, the cargo in Padang Hawks 1, 2,
3 and 4 holds had every outward appear
the actual moisture level, ance of having suffered liquefaction
during the voyage.
the transportable moisture limit2 (TML ).
Liquefaction (refer to Appendix A) occurs
To ascertain whether the cargo had been in granular cargoes that are loaded with
contaminated by seawater, SGS used two excessive moisture and subjected to energy
different tests; the Mohrs Method from ship motions. There are any number
Chlorine test and the High Temperature of scientific definitions of liquefaction.
Chlorine test. Both tests showed conclu The 1997 annual of the American Society
sively that the water within the holds was for Testing and Materials defines sponta
not seawater. neous liquefaction as:
the sudden large decrease of the shearing
Tests of the actual moisture contained in resistance of a cohesionless soil. It is
caused by the collapse of a structure by
the samples resulted as follows: shock or other type of strain and is associat
ed with a sudden but temporary increase of
the surface sample from number 1 hold, the prefluid pressure. It involves a tempo
rary transformation of the material into a
41 %; fluid mass.

the sub-surface sample number one In excess of 90 % by weight of the ore

hold, 36.5 %; loaded by Padang Hawk had a particle size
2 The International Maritime Organizations Code of Safe Practice for Solid Bulk Cargoes (BC Code) defines the
TML of a cargo which may liquefy as the maximum moisture content of the material which is considered safe for car
riage in ships not complying with the special provisions of 7.2.2 and 7.2.3 (specially fitted or constructed cargo ships).
It is derived from the flow moisture point (flow table test appendix D.1) or from data obtained from other test methods
approved by the appropriate authority of the port State as being equally reliable.

18
of less than 1 mm, 82.5 % of the cargo number 1 hold would also have been the
had a particle size of less than 0.053 mm. most effected by any pounding and this
This cargo was typical of the limonitic probably accounts for the particularly poor
nickel ores being mined in New state of the cargo in this hold.
Caledonia. The small relative particle size
and chemical properties of the mineral By the time the test grabs were taken
mean the amount of moisture in these from number 1 hold on the morning of 29
nickel ore cargoes can be large without an July, sufficient free water had been read-
appreciable change in the appearance or sorbed by the nickel ore for the cargo to
static handling properties. Limonitic return to a solid state.
nickel ores adsorb any free intergranular
water into the mineral structure over a
The cargo
period of time. Adsorption of any free
The shipment of the limonitic nickel ore
water continues until the ore is saturated
from New Caledonia to Townsville is
and unable to take up any more water.
subject to a detailed agreement between
Handling problems may occur when satu
the cargo sellers (Nickel Mining
rated ore is subject to vibration or stress,
Corporation) and Queensland Nickel Pty
or if there is a significant amount of unad
Ltd (QNPL). The agreement specifies the
sorbed intergranular free water.
minimum characteristics of the ore and the
conditions under which it should be
At some time during voyage number 904,
shipped. This agreement includes provi
despite the fact that the crew pumped the
sion for:
hold bilges regularly and effectively on 24,
25 and 26 July, a portion of Padang an agreed maximum moisture content
Hawks nickel ore cargo liquefied and (free water) of 35 %, with a lower refer
moved to port causing the vessel to list. ence price for ore with a moisture
After listing to port on the evening of 26 content in excess 36 % and, conversely,
July, the crew inspected the cargo holds a higher reference price for ore with a
and observed the cargo flowing with the moisture content less than 34 %;
movement of the ship. On arrival in
Townsville the cargo in the forward four
a particle size not to exceed 200 mm;
holds displayed clear evidence of liquefac
tion in the fluid form of the cargo and the
flow tests prior to loading to demon
free water present on the surface of the
strate to the satisfaction of the buyer
cargo in number 1 hold. The varying
and the ships master that the shipment
degree of liquefaction of the cargo, with
will not flow; and
the forward holds being the most affected,
may be explained by the motion of the
ship in the rough sea. Padang Hawk was a statement by the seller acknowledging
pitching and rolling heavily on 25 and 26 the masters right to refuse to load the
July. All of the cargo would have been cargo based on the results of the flow
similarly effected by the rolling. However, tests.
the amplitude of the pitching, and thus its
effect on the cargo, would have been With regard to the flow tests, the agree
greater at the forward holds. The cargo in ment requires that:

19

 Ten representative samples from the stockpile discharge ports, agents, and advice relat
to be loaded aboard the ship are to be tested no ing to demurrage. It did not contain any
later than seven (7) days prior to the scheduled
arrival of the ship. specific details of the moisture limits,
density, or other cargo characteristics. Nor
All results of flow tests should be recorded,
conveyed and delivered to: did it contain any direction pertaining to
the acceptability of the cargo.
(a The Ships Master (before the
commencement of loading):
The master was not provided with the
(b Seller, and
results of the flow tests or the moisture
(c Buyer. content of the nickel ore prior to loading
the cargo on voyage number 904. There is
The master received details of voyage some doubt that the master knew that this
number 904 from the shipowners Brisbane information had to be provided under the
representative. For information on the terms of the agreement and that he had the
cargo and loading, the master was referred right to refuse the cargo based on the
to the detailed instructions provided for results of the flow tests. By failing to
voyage number 903. This telex included provide the master with this information,
standard details of the cargo, loading and however, the cargo shippers were not only

FIGURE 8.

Cargo, hold number 3

20

in breach of the agreement but also the ed the investigation with a chemical analy
provisions of the International Convention sis of the nickel ore cargo shipped by
for the Safety of Life at Sea (SOLAS). Padang Hawk on voyage number 904,
(See Appendix B). which indicated that the average moisture
content was 36.55 %.
On 27 July, the day after the incident, the
shippers did send the ship a facsimile with Representative samples
a copy of the Moisture Certificate certi The QNPL agreement with the seller stip
fying that the ore loaded had a maximum ulates that 10 representative samples
free moisture of 37 %. A second docu must be tested no later than seven (7)
ment detailed the results of a series of 10 days prior to the scheduled arrival of the
flow tests carried out between 11 July and vessel. The wording of this portion of the
16 July. These are shown in Table 1. agreement is somewhat ambiguous. It can
Table 1 be read to mean that the cargo should be
Flow test results, New Caledonia
tested at least 7 days before it is to be
Sample No. Date of test Deformation (mm) loaded when in fact the final testing
1 01/07/99 1.7 should take place just prior to the time of
2 11/07/99 1.9 loading.
3 12/07/99 2.2
4 12/07/99 2.0 In the case of Padang Hawks voyage
5 13/07/99 2.3 number 904, according to the flow test
6 13/07/99 2.1 records supplied by the shipper after the
7 14/07/99 2.0 incident (shown in table 1), 10 tests were
8 15/07/99 1.8 in fact performed in the week prior to the
9 16/07/99 2.3 ships arrival at Kouaoua on 17 July. The
10 16/07/99 2.1 last tests were conducted the day before on
16 July. Given the subsequent liquefaction
of the cargo, the results of these flow tests
These tests showed sample deformations
would indicate that either; the flow test
(the degree to which the cargo flowed in
procedure is flawed in some way, or that
the test) of 1.72.3 mm. All of the
the samples tested were not representative
samples indicated that the cargo loaded
of the cargo as a whole.
aboard Padang Hawk in Kouaoua on
voyage number 904, was fit for carriage
using the QNPL criteria which stipulates a QNPL provided information as to how the
maximum acceptable deformation of 3 samples are collected from the ore stock
mm. piles:
The method by which the samples are taken
will vary according to the type and size of
According to information obtained inde the ore stockpile. The supplier is required to
pendently on behalf of the Inspector of utilise his expertise to obtain a representa
Marine Accidents by the French Marine tive sample from the stockpile. Typically,
Casualties Investigation Board, the mois samples are taken from the stockpile at a
depth which will ensure that the ore has not
ture content of the ore was 37.4 % which been affected by the prevailing either wet or
agrees reasonably well with the informa dry climatic conditions and is representative
tion provided to the master on the of the material to be loaded. This is normal
Moisture Certificate. QNPL also provid
21

 ly done by a small backhoe and depth of flow state. If the sample does not show
sample is approximately 0.5 to 1.0 metre. plastic deformation, it is placed in a
mixing bowl and a small amount of water
Rainfall records for Kouaoua for July
is added. It is then returned to the mould,
1999 from the Meteorological Office of
tamped, and retested on the flow table.
New Caledonia showed that in the first
The whole process is repeated until the
half of the month of July, Kouaoua had
sample shows plastic deformation indicat
received 171 mm of rain, 105 mm of
ing that a flow state has been reached.
which was recorded on 14 July, three days
before Padang Hawk arrived. At the The IMO procedure requires that the mois
Kouaoua nickel mine and ship-loading ture content of a sample just above the
facility, the stockpiles of ore are stored in flow point and another just below the flow
the open. The cargo intended for Padang point are to be ascertained using a stan
Hawk would have been fully exposed to dardised drying procedure. An average of
the rain. The flow tests conducted on 14, these values forms the flow moisture point
15, and 16 July showed sample deforma of the ore. The TML is then calculated
tions of between 1.8 and 2.3 mm and thus and is 90 % of the flow moisture point
no significant increase as a result of the expressed as a percentage of water by
heavy rain on 14 July. Nevertheless, the mass.
nickel ore loaded aboard Padang Hawk on
voyage number 904 was saturated to the
The agreement between QNPL and the
point where free water was observed
seller states that the standard IMO flow
running from the cargo as it was loaded.
moisture point test procedure is not suit
able for limonitic ores. As a result, QNPL
The cargo was not fit for carriage in spite have adapted the standard IMO flow mois
of the fact that the records of the sample ture point test to form the flow test pro
tests conducted in the week prior to cedure detailed in their agreement with the
loading failed to indicate any abnormality seller.
with the nickel ore.
The equipment used, the sample prepara
Flow tests tion, and the actual methodology of the
The standard IMO flow moisture point test QNPL flow test are almost the same as
is described in appendix D of the BC those described in the BC Code for the
Code. Part of the test requires that a rep flow moisture point test. The main differ
resentative sample of test material is ence between the two tests is that the sim
placed in a standard mould and compact plified QNPL test requires no addition of
ed in a prescribed fashion with a standard water. QNPL explained the reasons why
tamper. The sample, now in the form of the adaptation of the IMO procedure was
a truncated cone, is then turned out of the necessary and these are detailed in
mould onto a standard flow table and the Appendix C.
table is raised and dropped 50 times,
from a height of 12.5 mm, at the rate of 25 The QNPL-adapted flow test only pre
times per minute. If the sample shows scribes a maximum allowable sample
plastic deformation after this process, it deformation of 3 mm and does not stipu
indicates that the sample has reached a late that the actual flow moisture point or

22

the transportable moisture limit must be Study on Prevention of Sliding Failure of

established. Nickel Ore in Bulk. The study details a
new procedure for evaluating the shear
According to the flow moisture point test strength of nickel ore and thus the suitabil
in the BC Code, plastic deformation of ity for carriage of the cargo. The test pro
the sample indicates that a flow state has cedure utilises a cone penetrometer to
been reached. The Code does not indicate measure the shear strength of a graded
an acceptable level of deformation. By sample of nickel ore suitably compacted in
the definition adopted in the BC Code, all a standard container. Advice from
of the samples tested prior to loading the Professor Tamaki Ura of the University of
Padang Hawks voyage number 904 cargo Tokyo, one of the authors of the nickel ore
had reached a flow state. study, is that when the moisture content of
nickel ore is higher than a critical value, it
loses shear strength. The Japanese study
The Japanese experience
refers only to nickel ore and does not dif
Around 4 million tonnes of saprolitic
ferentiate between limonitic and saprolitic
nickel ore is shipped to smelters in Japan
ores.
each year. There have been a number of
reported instances where this cargo has
In submission QNPL reiterated:
shifted, as a result of excessive water
content, while being carried from mines in QNPL utilises ore from the limonite zone, while
the Philippines, Indonesia and New the Japanese and New Caledonian smelters
utilise saprolite ore, which exhibits different
Caledonia to Japan. After researching handling characteristics due to its differing
marine casualty databases, the ATSB mineralogy and particle size distribution.
found there had been a recent incident Consequently the comments from Professor
Tamaki Ura may not be applicable to
where a ship was lost while carrying limonite, as they almost certainly are related
nickel ore to Japan. to saprolite.

In August 1998, the Panama flag ship Sea Stability

Prospect foundered with the loss of ten The accepted practice on bulk cargo ships
seafarers from a crew of 21. The ship had is to calculate the vessels stability prior to
a full cargo of nickel ore loaded in departing both loading and discharge
Indonesia and was en route to Hiroshima ports. Stability calculations are performed
in Japan. The ship broadcast an SOS using: data relating to the vessels draft;
signal in a 2 m sea with wind speeds of the quantity and location of known masses
less than 20 knots. There is no informa including, ballast, fuel, fresh water, etc;
tion as to whether the loss of the ship and the vessels stability curves and tables,
could be attributed to the shift of cargo, (usually in tabulated format and often
the deformation of cargo, or to high mois computer based). These calculations are
ture levels. performed, usually by the mate, to ensure
the vessel has adequate stability for the
The Japanese authorities have been con planned passage and any contingencies
cerned for some time about the propensity that might be encountered en route. On
of nickel ore cargoes to shift, and the Padang Hawk, this was not an onerous
Journal of the Society of Naval Architects task as the ship was equipped with a
of Japan (vol 187, June 2000) contains a loading/stability computer.
23

When Padang Hawk completed loading The relevant stability data for Padang
the cargo of nickel ore at Kouaoua on Hawk was obtained for the purpose of the
investigation. The vessels stability was
24 July, no stability calculations were calculated independently for the purposes
undertaken prior to the vessel departing of the report using the departure data from
the port. The master knew from previous Kouaoua, for an initial condition, and the
experience with the cargo that the vessel subsequent conditions of the starboard
had a surplus of stability after loading. topside tanks ballasted, and the double
Once the cargo had become fluid, and bottom tanks ballasted. The calculations
Padang Hawk had developed the 15 list to showed that the vessel exceeded the intact
port, the masters initial response was to stability criteria recommended in the
pump ballast into 3 and 5 starboard Code of Intact Stability for all Types of
topside tanks. No stability calculations Ships Covered by IMO Instruments
were undertaken prior to this ballasting (Resolution A.749 (18)). In using the
operation to ensure that the vessel would topside tanks on this occasion the ship was
remain stable. Once again, the master not in fact put at risk through the possible
relied on his experience and judgement in loss of intact stability. The topside tanks
correcting the vessels list without using do have a higher centre of gravity than the
the hard facts available from the ships double bottoms, but have much smaller
loading/stability computer. free surface and bigger righting effects.

The owners of the vessel were consulted While recognising the circumstances and
once the list had been corrected using the the imperative to right the ships list, the
two topside tanks. Their advice was to master took a significant risk in ballasting
transfer the ballast from the topside tanks the vessel, by adding weight centred high
to the double bottom ballast tanks. The and outboard with an accompanying free
owners expressed concern that the ballast surface, without first checking the likely
in the topside tanks compromised the effect on the vessels stability. Although
vessels stability. The ballast was subse the master was correct in his assessment of
quently transferred in accordance with the the stability, there was a risk of far worse
directions of the owners. The most critical consequences for the vessel and crew,
times during these operations were the should his intuitive judgement have been
transient conditions when the topside and faulty. It would have been prudent to use
double bottom tanks were being emptied the available resources to calculate the sta
or filled with a consequent free surface. bility of the vessel for all of the conditions
The vessels stability was not calculated prior to transferring any ballast.
for any of these operations.

24

Conclusions
tions the relevant information pertaining
to the cargo moisture content, flow
tests, and the masters right to refuse to
load the cargo under the terms of the
agreement between QNPL and the
These conclusions identify the different cargo sellers.
factors contributing to the incident and
should not be read as apportioning blame
6. The ore seller did not provide the
or liability to any particular organisation
master with the agreed data pertaining
or individual.
to the cargos moisture content and flow
tests as required by SOLAS.
The factors contributing to the shift of
Padang Hawks nickel ore cargo and the
7. The master loaded the nickel ore
consequent port list include, but are not
without insisting on the provision of the
limited to:
data concerning the moisture content
and flow tests.
1. The cargo was loaded with excessive
moisture content.
8. The mined nickel ore was stockpiled in
areas open to the ingress of rainwater.
2. The vessel was subjected to heavy seas,
which led to the cargo changing state
9. The agreement between QNPL and the
from a solid to a viscous liquid in 4 of
seller did not stipulate a reasonable,
the 5 holds.
maximum, acceptable moisture content,
based on the nickel ores ability to be
3. Insufficient knowledge of the charac carried safely by sea.
teristics of nickel ore as a cargo and its
propensity to become fluid when the
Also:
moisture content is high and it is sub
jected to sufficient physical stress.
10.The vessels stability should have been
calculated for the loaded condition
4. There is no test to specifically ascertain
leaving Kouaoua and subsequently
the transportable moisture limit of
checked prior to the pumping of ballast
nickel ore.
into the topside and double bottom
tanks to correct the list.
5. The owners/agent of the vessel did not
include in the masters voyage instruc

25

FIGURE 9.

Padang Hawk events and causal factor chart

26

Submissions

Under sub-regulation 16(3) of the

Navigation (Marine Casualty) Regulations,
if a report, or part of a report, relates to a
persons affairs to a material extent, the
Inspector must, if it is reasonable to do so,
give that person a copy of the report or the
relevant part of the report. Sub-regulation
16(4) provides that such a person may
provide written comments or information
relating to the report.

The final draft of the report, or relevant

parts thereof, was sent to the following:

The Master, Padang Hawk

Singa Star Pte Ltd

Daiichi Chuo Shipping (Singapore) Pte

Ltd

Queensland Nickel Pty Ltd

Nickel Mining Corporation, New

Caledonia

Australian Maritime Safety Authority

Professor Tamaki Ura, University of Tokyo

Submissions were received from

Queensland Nickel Pty Ltd, the Australian
Maritime Safety Authority, and Professor
Tamaki Ura. The text of the draft was
amended and portions of the submissions
included as appropriate.

27
28

Details of ship

Name Padang Hawk

IMO No. 9109354

Flag Singapore

Classification Society Nippon Kaiji Kyokai

Vessel type Bulk carrier

Owner Singa Star Pte Ltd

Year of build 1995

Builder Mitsui Engineering and Ship Building, Japan

Gross tonnage 27 011

Summer deadweight 46 635 tonnes

Length overall 189.80 m

Breadth, moulded 31.00 m

Draught (summer) 11.60 m

Engine Mitsui Man B&W 6S50MC

Engine power 6 532 kW

Service speed 14 knots

Crew 20 (Filipino)
30

 Appe dix A
Liquefaction it is essential to ascertain normally from
the shipper the current physical character
istics and chemical properties of the materi
Granular materials have void spaces al.
between the particles caused by the irregu
lar shape of the particles. These void The Code notes the characteristics of such
spaces may be filled with air, water or a cargoes as one that:
combination of both. When a cargo con generally consist of a mixture of small
taining moisture is subjected to energy particles as contrasted with natural ores
such as ship motions, the cargo particles which include a considerable percentage of
large particles or lumps.
move to compress the void spaces and
pressurise any free water present in the
spaces (pore water pressure). In addition,
moisture may be released from the mineral
structure of some types of cargo when
subjected to ships motions. This release
of adsorbed water increases the amount of
free water in the cargo and leads to a
further increase in the pore water pressure.
If the pore water pressure becomes high
enough, it overcomes the friction forces
binding the individual particles of material
and the shear strength of the cargo falls to
the point where liquefaction occurs. The
cargo becomes a viscous fluid with the
ability to flow.

The BC Code provides guidance to admin

istrations, shipowners, shippers and
masters on the standards to be applied in
the safe stowage and shipment of solid
bulk cargoes, excluding grain. There is a
specific section on cargoes liable to lique
faction and an appendix, appendix A,
which contains a list of such cargoes.

Nickel ore is not contained in the appen

dix.

However, the introduction to the Code

specifically states:
it should be carefully noted that the list
of materials appearing in appendices A,B
and C to the Code is by no means exhaus
tive and the physical properties attributed to
them are intended only for guidance.
Consequently before loading any bulk cargo

31

32

 Appendix B
SOLAS

The International Convention for the

Safety of Life at Sea, regulation 2, part A,
chapter VI, Cargo Information states:
1 The shipper shall provide the master or his
representative with appropriate information
on the cargo sufficiently in advance of
loading to enable the precautions which
may be necessary for proper stowage and
safe carriage of the cargo to be put into
effect. Such information shall be confirmed
in writing and by appropriate shipping doc
uments prior to loading the cargo on the
ship.

And further:

2 The cargo information shall

include:
.2 in the case of bulk cargo, information on the
stowage factor of the cargo, the trimming
procedures and, in the case of a concentrate
or other cargo which may liquefy, additional
information in the form of a certificate on
the moisture content of the cargo and its
transportable moisture limit.

33

34

 Appe dix C
The adapted flow test There are two main reasons for this:
procedure Firstly most of the water in an undisturbed
piece of limonite is adsorbed onto the very
high surface area possessed by the very fine
The flow test procedure used to test the grained (to almost amorphous) goethite
nickel ore carried by Padang Hawk on dominated mineralogy. This is why undis
voyage number 904 is outlined in the turbed pieces of limonite containing >40%
moisture can appear dry. Water entering a
agreement between the cargo seller and piece of limonite will probably take much
QNPL. The flow test is adapted from the more time than a flow moisture test would
BC Code procedure for ascertaining flow allow to change from being intergranular
moisture point and, like the Code, it identi free water to becoming water adsorbed onto
the goethite minerals. Unlike adsorbed
fies that: water, free water moves amongst the pieces
a flow state is considered to have been of limonite and lubricates them allowing
reached when the moisture content and them to move freely, leading to slumping or
compaction of the sample produces a level collapse.
of saturation such that plastic deformation Secondly and perhaps most importantly, the
occurs. At this stage, the moulded sides of above effect is compounded when:
the sample may deform, giving a convex or
concave profile. preparation of the sample for the flow
moisture test disturbs the material. This
The agreement goes on to set a limit on disturbance (which is greater than any
mining and loading operation does to
the acceptable amount of sample defor ore) leads to compaction or slicking of
mation: limonite pieces (1mm-10cm range), thus
reducing the permeability of the surface
Under plastic deformation, if the increase in of these pieces,
the diameter of the moulded sample meas
ured at any part of the cone exceeds 3 mm, then added water cannot be absorbed into
the cargo which it represents is considered the pieces and runs or slurries as free
unsuitable for shipment. water amongst the surfaces of the grains.
The stability coefficient of a limonite
QNPL provided the following explanation sample is probably more dependent on the
as to why the BC Code procedure is not amount of free water than on the amount of
suitable for testing limonitic ores: adsorbed water, so for example:

The IMO Test was developed for concen sample A with 40% absorbed water and
trates which have no specific structure and 0% free water is much more stable than:
involves mixing of water with the material.
sample B with 30% absorbed water and
In the case of limonitic nickel ore, this test
10% free water.
changes the structure of the macro particles
and the test material becomes a different All of this is independent of the effect of
material to the product being transported, liquefaction caused when adsorbed or
therefore it is not suitable for use in testing absorbed water is released from near satu
the suitability for transport of limonitic rated limonite by vibration.
nickel ore.
In 1986, QNPL commissioned a study of
Specifically relating to the ore structure, the behaviour of limonitic nickel ore
QNPL said that: during shipping. Gutteridge Haskins &
the sample preparation of limonite nickel Davey and Chalmers University,
ore for the flow moisture test, and the act of Gothenburg, carried out tests specifically
adding water to this sample, increases the
amount of free water present as opposed to related to the likely behaviour of nickel
adsorbed or absorbed water. ore to evaluate ship unloaders. The inves
tigations were related to the determination
35

of the shear strength of nickel ore, density

and stability of ships cargoes under the
conditions encountered at sea. A further
report by Gutteridge Haskins & Davey in
1990 titled Gag Island Nickel Laterite
Deposit Report on Ship Stability, Trial
Shipments V3/90 and V4/90 concluded in
part that no significant strength changes
have been observed within Gag Island
limonite or saprolite as a result of ship
ment. These reports form the basis of
QNPLs third-party verification of their
adapted flow test.

36