Federal District Court Opinions

and Consolidated Cases. No. 98 Civ. 8394 (DC). United
States District Court, S.D. New York. October 18, 2005


DENNY CHIN, District Judge

On November 9, 1998, the M/V DG HARMONY was off the coast
of Brazil, en route from Miami. At approximately 7:20 a.m.,
the vessel shuddered. Within moments, dense smoke covered
the ship. The master of the vessel, Captain Michael
Balitzki, rushed to the bridge. After checking the wind, he
turned the ship starboard. The wind cleared the smoke from
the deck, and he saw flames coming from cargo hold 3.

The chief officer, who had the watch, had already sounded
a general alarm and alerted the crew to assemble. The crew
began fighting the fire, wearing fire suits and using hoses
and pumps. The crew continued to fight the fire until late
afternoon, when the captain ordered most of the crew to
abandon ship. A lifeboat was launched at 6 p.m., carrying
away fourteen crew members and leaving only the captain and
a handful of others behind. As the fourteen crew members
looked back from the lifeboat, they saw the HARMONY ablaze,
with flames and smoke Page 2 rising high from
approximately the middle third of the ship. One of the crew
members took a photograph: (CX 110 at 1).[fn1]

The captain and the others who remained aboard the HARMONY
continued to fight the fire and operate the vessel. They
finally abandoned ship at 2 a.m., after yet another
explosion, when the captain decided that it was no longer
safe to remain on board. The vessel had been on fire for
more than eighteen hours, and portions of her deck and side
shell plating had turned red and white hot. The captain
collected the vessel’s log books and charts, and he and the
remaining crew members evacuated, Page 3 transferring from
the HARMONY via a lifeboat to the SEALAND URUGUAY, a
northbound container ship that had been standing by
approximately half a mile away.

The DG HARMONY continued to burn for three weeks. Most of
its cargo was destroyed or damaged. The vessel itself was
declared a constructive total loss and eventually was

Some eighteen lawsuits were filed in this Court, by cargo,
vessel, and other interests seeking recovery for damages
suffered as a result of the casualty. All claims have been
settled except for the claims of certain cargo and vessel
interests[fn2] against defendant PPG Industries, Inc.
(“PPG”), the manufacturer of calcium hypochlorite hydrated
(“cal-hypo”), a bactericide used to purify water. Ten of
the containers aboard the HARMONY contained 120 drums each
of cal-hypo, manufactured and shipped by PPG.

Plaintiffs contend that the cal-hypo in one (if not two) of
the containers decomposed and self-heated, resulting in
“thermal runaway,” an explosion, and the fire that
followed. Plaintiffs assert theories of strict liability,
breach of warranty, failure to warn, and negligence, and
proceed under the Carriage of Goods by Sea Act (“COGSA”),
46 U.S.C. § 1304, general maritime law, and New York
law, as incorporated by general admiralty principles. Page

PPG denies liability, arguing that plaintiffs are unable
to prove the origin and cause of the fire and that it is
impossible to determine where the fire started. PPG also
argues that its cal-hypo was not defective, it had no duty
to warn, and, even assuming such a duty existed, the duty
was not breached. PPG further argues that even if the fire
did originate in the containers of cal-hypo, the crew
failed to properly stow the cargo because it recklessly
placed the containers of cal-hypo next to a source of heat
— heated fuel oil bunkers. Finally, PPG argues that
the crew was negligent in its efforts to fight the fire.

The case was bifurcated and the issue of liability was
tried to the Court, without a jury, from April 22 through
May 18, 2004. The parties also submitted extensive
deposition excerpts. My findings of fact and conclusions of
law follow.


A. The Casualty

1. The Vessel

The DG HARMONY was a full container ship built as the
HANSA CLIPPER in West Germany in 1989. (DX 40 at 5; see Tr.
101, 197). The HARMONY was one of a series of sister
vessels. (DX 40 at 5). It had an overall length of 176.57
meters and a breadth of 27.5 meters. (Id. at 4).

The vessel had three holds with numbered bays using odd
numbers from forward to aft as follows: Page 5

Hold 1 Bays 1-7

Hold 2 Bays 9-23

Hold 3 Bays 25-39

(Id. at 12; Tr. 256; CX 31). The vessel could hold the
equivalent of 1,799 twenty-foot containers. (DX 40 at
12).[fn3] Hold 3 had twelve hatch covers (or “pontoons”),
each 40 feet long and 25 feet wide and weighing 22 tons.
(Tr. 246-47, 290).

Containers were placed into the vessel in “slots,” running
lengthwise aft to forward. Each odd-numbered bay could hold
a twenty-foot container in each slot. A forty-foot
container could be placed across two slots and the
combination of the two bays would be referred to by the
corresponding even number.[fn4] Rows of slots continued
across the breadth of the vessel. Containers could be
stacked under deck up to the top of the hold, just below
the hatch covers. Each layer in the stack was referred to
as a “tier.” Hence, each slot could be identified by a
series of three numbers indicating the bay, the row, and
the tier.[fn5] Containers could also be stacked on deck, on
top of the hatch covers, as many as six or seven containers
high. (Tr. 289-90, 315-16, 379; see CX 92). Page 6

All three holds were ventilated by electric fans and were
designed to carry dangerous goods. All three holds had CO2
extinguishing systems, and hold 1 had a water sprinkler
system as well. (DX 40 at 13; Tr. 201-02). Each hold also
had a smoke detection system that would, upon detecting
smoke, sound an alarm and turn on a red light on a control
panel on the bridge. (Tr. 203-05). The red light would
indicate whether the smoke condition was in hold 1, 2, or
3. (Tr. 205). The smoke detection system was last inspected
before the casualty on September 2, 1998, when it was
inspected by a surveyor from the Isle of Man. (Tr. 205-06;
CX 91, 92).

The HARMONY had a fire control and safety plan, which was
posted in several places on board, in German and English.
(Tr. 199-200; CX 95). The firefighting system included, in
addition to the features described above, two seawater
pumps, an emergency pump, a series of hydrants and hoses,
fire extinguishers, and firefighting suits. (Tr. 200-03).
The crew generally conducted fire and boat drills on
Saturdays when the Harmony was at sea; the drills included
running the fire pumps and running water through the hoses.
(Tr. 211-14).[fn6]

The HARMONY was originally registered in Germany. In
September 1998, just two months before the casualty, the
ownership of the vessel changed as the HARMONY was acquired
by Page 7 Navigator Shipping, Ltd. (“Navigator”), a
subsidiary formed for this purpose by Safmarine and CMBT
Lines N.V. (“SCL”). (Tr. 101, 106; see CX 14, 15). Upon the
purchase, the registry of the vessel was transferred to the
Isle of Man. (See CX 14; Tr. 133).

At the time of the purchase, the Harmony was on charter to
DiGregorio Navegacao Ltda. (“DiGregorio”), which had been
employing the vessel as a member of the Independent
Carriers Alliance (the “ICA”), a group of carriers (or
“slot charterers”) engaged in a liner service between the
east coast of the United States and Brazil. The ICA
included DiGregorio, Cho Yang Shipping Co. (“Cho Yang”),
DSR Senator Lines GmbH, Montemar S.A. Pan-American
Independent Line, Zim Israel Navigation Co., and Hanjin
Shipping Co., and it operated a tonnage center, where stow
planners prepared pre-stow and final stow plans. (Tr.
288-89, 832-33, 913-14). SCL did not have an immediate need
for the Harmony, and thus it decided to and did re-charter
the vessel to DiGregorio. (Tr. 106-08, 134; CX 16, 17).
Around the same time SCL entered into crew and ship
management agreements with Leonhardt & Blumberg (“L&B”), a
reputable ship management company. (Tr. 134-40; CX 18, 19).

2. The Crew

At the time of the casualty, the DG HARMONY had a crew of
twenty, exceeding the number (seventeen) required by the
Safe Manning Certificate issued by the Isle of Man. (CX 22,
23; Tr. 218-19, 221). The crew included the captain,
Michael Balitzki, Page 8 from Germany;[fn7] the chief
mate, Yuri Kovshel, from Russia; and the chief engineer,
Ulrich Hahnemann, from Germany. (CX 23). Many of the crew
members were from the Kiribati islands in the South
Pacific. (Id.).

Both Captain Balitzki and chief mate Kovshel were trained
in both firefighting and the handling of dangerous cargo
and were certified in the handling of hazardous materials.
(Tr. 186-88, 191, 241-43, 276-77, 279; CX 24 at 436-40; CX
98). Although Balitzki was ultimately in charge of ensuring
the proper stowage of dangerous cargo aboard the HARMONY,
he delegated the day-to-day responsibility for this task to
the chief mate. (Tr. 240-41, 282).

3. The Voyage

The voyage commenced in New York on October 26, 1998, bound
for Newport News, Virginia. The HARMONY arrived in Newport
News on October 27th, where it took on the ten containers
of cal-hypo. From there the HARMONY traveled to Savannah
and then to Miami. It departed Miami on October 30th, bound
for Rio de Janiero, Santos, and other ports in Brazil. (Tr.
222-23; CX 84 at 08210, 08211; CX 103 at 11816, 11825).
Page 9

4. The Fire

On November 9, 1998, at approximately 7:20 a.m., as the
HARMONY was off the coast of Brazil, Balitzki was in his
bedroom when he felt the vessel shudder two or three times.
(Tr. 224). Within three or four seconds, his phone rang. He
went to the phone, which was located next to a window,
looking forward. As he reached for the telephone, he looked
out the window and saw dense, white, dirty smoke. He could
see nothing else. The chief mate, who had the watch, was on
the phone, shouting into the phone. (Tr. 224-25, 312, 314;
Kovshel Dep. 247-48, 351).

Balitzki got dressed and immediately went to the bridge,
arriving at approximately 7:26 a.m. The chief mate was the
only other person present. Again, the captain saw dense,
white, dirty smoke. The general alarm had been sounded,
alerting the crew that something was wrong and that
therefore they should assemble. The smoke detection system
indicated that there was smoke in hold 3 and the engine
room. The captain sent the chief mate to the muster
station, where the crew was gathering, and also instructed
him to check the main deck. (Tr. 226-29; Kovshel Dep.
248-49, 351-52).

Balitzki checked the ship’s course, the weather conditions,
and the wind. He turned the ship to the starboard side, to
clear the smoke from the deck. He then saw the flames for
the first time. The vessel’s speed had been reduced
automatically because smoke had been detected in the engine
room. (Tr. 227-28; see Kovshel Dep. 249-50). Page 10

Communicating by walkie-talkie, the chief mate reported to
Balitzki that the vessel was on fire, in bay 26, the forward
part of hold 3, on the port side. He also reported that one
hatch cover was missing; the 22-ton hatch cover had been
blown off by the force of the explosion.[fn8] Balitzki then
checked the dangerous goods list and dangerous goods
stowage plan, and saw that the fire had broken out
approximately where the ten containers of cal-hypo had been
stowed. (Tr. 228-29; Kovshel Dep. 250-53).

The crew immediately started fighting the fire, using all
three firefighting pumps at maximum capacity, all hydrants,
and all available fire hoses. They even took some hoses
from the engine room. It quickly became clear, however,
that the fire could not be extinguished with just water.
(Tr. 229-31; Kovshel Dep. 248-50; see CX 110 at 2).

After approximately an hour, the port side fuel oil tank
in hold 3 breached. The smoke, which had changed in color
gradually from white to gray, turned to black. The fire
started to spread, including onto the deck. At some point,
the crew used Page 11 the C02 system to try to extinguish
the fire. The captain did not believe the C02 would work
because a hatch cover had been blown off, but he tried it
anyway. It did not work. (Tr. 231-33, 321; Kovshel Dep.
253). Approximately three hours after the fire started,
Balitzki sent a distress signal, to a rescue center on land
to coordinate rescue measures and send firefighting
equipment. (Tr. 232, 324). Another ship had offered help
within the first hour or so, but the captain had declined
assistance because he thought he could bring the fire under
control. (Tr. 322). Other ships also offered assistance,
but Balitzki declined; he needed foam to fight what had
become a petroleum fire, and he did not want to risk having
a tanker come alongside the burning HARMONY. (Tr. 323; see
Kovshel Dep. 395).

At approximately 3 p.m., Balitzki decided that most of the
crew should abandon ship. He and five others, including the
chief engineer and the chief mate, were to remain on board.
The other fourteen crew members abandoned ship in a
lifeboat at 6 p.m., while it was still daylight They were
picked up by a German containership, the SEALAND URUGUAY,
that had been following the HARMONY about a mile behind
since noon. (Tr. 233-34, 322; Kovshel Dep. 257).

Balitzki and those who stayed behind took all the
remaining fire hoses and nozzles and fixed them to railings
and other extensions with the pumps running, intending to
create a water curtain to protect the super-structure and
engine area. At approximately 1:30 a.m., there was another
explosion, on the Page 12 starboard side of the
vessel.[fn9] Balitzki decided it was no longer safe to
remain on board, and at 2 a.m. he and the five others
abandoned ship, transferring via a lifeboat to the SEALAND
URUGUAY. (Tr. 235-36). The HARMONY continued to make speed,
as its engine continued to run. (Tr. 236).[fn10]

Three or four days later, Balitzki re-boarded the HARMONY,
to release both anchors. (Tr. 236-38). The ship was still
on fire, although the fire had “calmed down a little bit.”
(Tr. 238). The fire continued to burn, and did so for a
total of approximately three weeks. (Tr. 239-40).

Eventually, the vessel was declared a total loss. The
HARMONY had been purchased for $16.43 million but the
estimated cost of repairs was more than $18 million. Hence,
the decision was made to scrap the vessel. (Tr. 116; CX

4. The Investigation

The various interests retained fire investigators shortly
after the incident. For approximately two days in Page 13
January 1999 and three days in February 1999 six fire
investigators inspected the HARMONY in Curacao. These
included Dr. Geoffrey Philip Bound, retained by the cargo
interests and vessel owners, and Dr. Roger McCarthy, for
PPG. In general, the investigators viewed the fire scene as
a group, taking photographs and videos as the debris was
removed, bit by bit. The investigators were permitted to
walk around on the vessel, getting close to objects,
including the remains and burnt-out ruins of containers.
(Tr. 498-504). Eventually, the containers (or their
remains) were removed and examined, with the investigators
observing and inspecting each container during the process.
(Tr. 547). Fluorescent spray paint was used to mark the
container numbers on the containers (or ruins thereof) as
they were removed; the various investigators conferred and
agreed on the container number before the number was
painted onto the container. (Tr. 553). As Dr. Bound

Investigating a containership [fire] is like taking apart
a 3-dimensional jigsaw. You only get one chance to look at
the evidence as it becomes uncovered. . . .

(Tr. 501). The investigators focused their attention on hold
3 and, in particular, the area in hold 3 where the
containers of cal-hypo had been stowed.

B. The Cargo

1. Cal-Hypo

Cal-hypo is a bactericide used to purify water. It is a
molecule containing calcium chlorine and oxygen in a form
that Page 14 is a “fairly high-energy material.” (Tr.
599). When it decomposes, it liberates oxygen and chlorine.
Cal-hypo will decompose at room temperature. The higher the
temperature it is exposed to, the greater the rate of
decomposition. (Tr. 599; Simmons Dep. 21). When cal-hypo
decomposes, it liberates heat and heats itself up. Cal-hypo
reacts faster when it gets hot. (Tr. 600, 1312-13). When
cal-hypo is stored in a drum, heat is retained inside the
drum. The larger the mass, the harder it is for the heat to
escape, and the decomposition is likely to increase.
Similarly, the higher the ambient temperature (the
temperature of the atmosphere surrounding the drum), heat
loss becomes more difficult and the temperature inside the
drum will increase. (Tr. 601-03, 1313-14).

When cal-hypo reaches “decomposition temperature,” there is
a rapid breakdown of the material, liberating oxygen. The
rapid decomposition throws white particulate material
around, which has the appearance of fine smoke. A great
deal of heat is also liberated, which can ignite materials
in the immediate vicinity, which may result in a white or
cream-colored smoke. The term “critical temperature” or
“critical ambient temperature” (or “CAT”) refers to the
ambient temperature at which the heat from the material
inside the drum cannot escape fast enough, heat is retained
inside, and the material in the drum becomes hotter,
increasing to the point of self-decomposition. The reaction
becomes circular and “runs away” — the material
explodes, Page 15 decomposes, and a fire ensues. (Tr.
603-04).[fn11] The term “self-accelerating decomposition
temperature” (or “SADT”) is used in the United Nations
testing procedures, and is the ambient temperature at which
the bulk of the material in a specific package (once it is
within 2°C of the ambient temperature) will rise by a
temperature of 6°C within the period of seven days. (Tr.
605-06, 676). CAT and SADT are similar, but the SADT does
not necessarily result in thermal runaway while the CAT
does. (Tr. 606-07).

The placement of drums of cal-hypo adjacent to each other
will make it more difficult for the heat to dissipate.
Hence, the critical temperature for a stack of drums will
be lower than for an individual drum, because it is harder
for the heat to dissipate. (Tr. 604-05, 643-44, 1316). A
container stuffed with drums creates a double risk —
the stacking of a larger number of drums creates more heat
and the container walls inhibit the ventilation of the
drums. A containerized stack of drums is likely to have a
lower critical temperature than the same stack of drums not
stuffed in a container. (Tr. 605; see also Simmons Dep. 239
(“the mass of the material in the package is a factor” in
determining the CAT), 244 (CAT is lower in “larger
packages”); Banks Dep. 69 (“SADT varies with respect to
Page 16 package size.”), 70 (“[L]arger packages have lower
SADTs.”); Ferguson Dep. 201-02).

2. The Transportation of Hazardous Materials

The International Maritime Organization (the “IMO”), which
is affiliated with the United Nations (the “UN”), has
developed a code, the International Maritime Dangerous
Goods Code (the “IMDG Code” or the “Code”), to govern the
carriage of dangerous goods aboard ships. (Tr. 111, 802-03;
CX 2). There is an International Convention for the Safety
of Life at Sea (“SOLAS”), to which the United States and
other member nations of the UN are signatories. The members
of the IMO and the signatories to the SOLAS convention have
adopted the IMDG Code. (Tr. 114-15; CX 2, Vol. I, at

The Code is relied on by shippers, manufacturers, tonnage
centers, stowage planners, and ships’ crews to make
decisions on the carriage and stowage of dangerous goods.
The Code is “absolutely critical” in the system for the
transportation of dangerous goods by sea: “It is the bible
for people who go to sea.” (Tr. 112-13; see also Tr. 681).

Cal-hypo is a Class 5.1 oxidizer designated as “UN 2880” in
the IMDG Code and the Hazardous Materials Table of the
Department of Transportation (“DOT”) regulations, 49 C.F.R.
§ 172.101. (CX 3 at 127; CX2, Part 3 at 5138).[fn12]
The IMDG Code Page 17 provides stowage instructions
specifically for UN 2880, as follows:

Stowage Category A. “Away from” sources of heat where
temperatures in excess of 55°C for a period of 24 hours or
more will be encountered. “Separated from” powdered
metals, ammonium compounds, cyanides and hydrogen

Packing, Stowage & Segregation See also General
Introduction and introduction to this class.

(CX 2, Part 3 at 5138). Category A includes cargo ships,
and stowage is permitted on deck or under deck. (CX 2, Part
1 at 0122). The reference to 55°C (which is equivalent to
131°F) is to the ambient temperature, or the air
temperature surrounding the cargo, and not to the
temperature of the heat source itself. (Tr. 266,

In addition to the individual schedules, the IMDG Code
includes a general introduction that sets forth general
principles, including the following:

14.2 Substances, materials and articles should be stowed
as indicated in the individual schedules in the various
classes in accordance with one of the categories specified
below. . . . Page 18

. . .

14.6 In view of the high protective advantages, stowage
under deck has been recommended whenever possible, except
that, for certain articles of class 1 whose principal
hazard is the production of smoke or toxic fumes, stowage
on deck has been recommended. . . .

. . .

14.12 Where it is necessary to prevent pressure build-up,
decomposition or polymerization of a substance, the
packages should be stowed shaded from radiant heat, which
includes protection from strong sunlight.

14.13 When it is recommended in individual schedules that
the substance should be shaded from radiant heat, stowage
under deck should be “away from” sources of heat,
including sparks, flame, steam pipes, heating coils, etc.

(CX 2, Part 1 at 0122-24; see generally Tr. 681-84).[fn14]
Where there is a conflict between instructions in an
individual schedule and the general introduction, the
individual schedule takes precedence. (Tr. 684-86; CX 2,
Part 1 at 0122 (14.2), 0127 (15.1.6)).

The transportation of hazardous materials is also governed
by DOT regulations, including 49 C.F.R. § 173.21,
which provides in part:

Unless otherwise provided in this Page 19 subchapter,
the offering for transportation or transportation of the
following is forbidden:

. . .

(f) A package containing a material which is likely to
decompose with a self-accelerated decomposition
temperature (SADT) of 50°C (122°F) or less . . . with an
evolution of a dangerous quantity of heat or gas when
decomposing . . ., unless the material is stabilized or
inhibited in a manner to preclude such evolution. The SADT
may be determined by any of the test methods described in
Part II of the UN Manual of Tests and Criteria.

(CX 3 at 357-58). One means of stabilizing or inhibiting
decomposition is the use of refrigerated containers. (Tr.

The IMDG Code provides that cargo having an SADT of “over
35°C” (95°F) must be transported at a “control temperature”
of 10°C less than the SADT. (CX 2, Part 1 at 0155).[fn15]
The control temperature is the maximum temperature at which
substances such as self-reactive materials can be safely
transported. (Id.). Hence, if a cargo has an SADT of 45°C,
the control temperature would be 35°C. (Tr. 673-74).

The DOT regulations also require a shipper “who offers a
hazardous material for transportation” to certify that the
regulations have been complied with, including as to
classification, packaging, and labeling. (CX 3 at 291 (49
C.F.R. § 172.204)). The shipper must tell the
carrier what the UN number is so that the carrier can then
refer to the IMDG Code to Page 20 determine the stowage
requirements. (Tr. 885).

3. The Cal-Hypo on the HARMONY

The cal-hypo at issue in this case was manufactured by PPG
at its plant in Natrium, West Virginia in mid-October 1998.
(Tr. 1248-51, 2078). PPG booked the shipment of ten
containers of UN 2880 with Cho Yang, which processed the
booking request through the ICA tonnage center. (CX 4; Tr.

The cal-hypo was placed in 136-kilogram (or 300-pound)
fiber drums made of a thick cardboard with metal rings at
the top and bottom. The temperature of the cal-hypo was
approximately 34°C (93.2°F) when it was placed in the
drums. Four drums were placed on each wooden pallet and
each pallet was then immediately shrink-wrapped in plastic.
The shrink-wrapped pallets were then loaded into the
containers, 30 pallets in 3 layers of 10 pallets each, 5
pallets on each side of the container, for a total of 120
drums in each container.[fn16] As a consequence, the
tightly-wrapped containers were packed into the
unventilated containers, stacked three-high, with very
little air space and a large quantity of Page 21 cal-hypo
(36,000 pounds) in each container. PPG did the packing and
loading — selecting the pallets, loading the drums
onto the pallets, shrink — wrapping the drums onto
the pallets, and then placing the pallets into the
containers. The bottom layer sits on the floor of the
container, and there is no separation between the pallets
on each side of the container or between the layers. The
containers were provided by Cho Yang and inspected by PPG.
(Tr. 563-64, 608-10, 2078-84; see PX 111; OX 30, 31, 43;
see also Patterson Dep. 113, 134-35; Ferguson Dep. 29).
The containers were not refrigerated.[fn17]

The containers were shipped out by truck approximately 36
hours after the cal-hypo was manufactured, under
circumstances that did not permit it to cool down. (Tr.
2078). In transit, the containers were exposed to some sun,
although there would have been some cooling at night
(although not markedly). (Tr. 2078-81; see PX 169 at 0876.
The temperature inside a container can be significantly
higher than the ambient temperature. (Tr. 2079-80; PX 168
at 3).

The ten containers were loaded onto the HARMONY at Newport
News, Virginia in late October 1998. (CX 4; Tr. 222). They
were accompanied by a three-page dangerous goods summary.
Page 22 (Tr. 255; CX 105). The summary showed that the ten
containers contained cal-hypo and identified the cargo as
class 5.1 UN 2880. (CX 105). PPG also provided dock
receipts for the ten containers describing the material as


(See CX 6; Tr. 697, 700-01). The reference to “PG 5138” was
to page 5138 of the IMDG Code. (CX 2, Part 3 at 5138). The
reference to “ERG #45” was to an outdated Emergency
Response Guide publication. (Tr. 702). The dock receipts
included the following certifications, signed by a
representative of PPG:

It is declared that the packing of the container . . .
has been carried out in accordance with the general
introduction IMDG code, paragraph 12.3.7. . . .

This is to certify to that the above-named materials are
properly classified, described, marked, labeled, and are
in proper condition for transportation, according to
applicable regulations of the [DOT].

(See CX 6; Tr. 699-01).

The containers were also accompanied by a Material Safety
Data Sheet prepared by PPG. (See Tr. 704-05, 887-88). The
sheet warned that the product was a “Strong Oxidizing
Agent” that could cause fire or explosion with
contamination. (CX 184 at 5406). The sheet also stated that
“[p]roduct decomposes at 180 C releasing oxygen gas.” (Id.
at 5407). It also stated that the product was to be stored
“in a cool, dry, well-ventilated place . . ., away from
heat, sparks, flames, direct sunlight, Page 23 and other
sources of heat.” (Id. at 5408). With respect to stability,
it stated that the product was “[u]nstable above 117 C.”
(Id. at 5409). It also stated, with respect to
transportation information, that the product was a 5.1
oxidizer, UN 2880. (Id. at 5411).

In general, when the HARMONY arrived in a port, the vessel
would receive a general loading plan covering all cargo that
was going onto or off the vessel in that particular port.
(Tr. 249). When dangerous goods were loaded on board,
Balitzki, the captain, would receive dangerous goods
manifests. He would sign for them and then pass them on to
the chief mate, who was the cargo officer. (Tr. 243). The
chief mate would check the ship’s documents of compliance
to ensure the proper stowage of dangerous goods. (Tr.
249-50). Once the chief mate determined that the cargo was
permitted by the document of compliance, he would then
check the IMDG Code to determine whether there were any
specific stowage restrictions. (Tr. 254-55; see Kovshel
Dep. 398-400). The captain and the chief mate would also
consult the Storck Guide, which set forth guidelines on
segregation. (Tr. 283-86; CX 83).

On this voyage, the ten containers of cal-hypo were stowed
in hold 3, in bays 25 and 27, in slots 270802, 270602,
270402, 250802, 250602, 250402, 250804, 250604, 250404, and
250406. The port sides of containers 270802, 250802, and
250804 were adjacent to the No. 5 heated port side bunker
tank in hold Page 24 3. (PX 92).[fn18] There was a gap of
some twelve inches between the tank wall and the
containers, because cell guides — bars — were
affixed to the wall to keep the containers in position
while the ship was at sea. (Tr. 361-63). Containers 270802
and 250802 also sat directly above the port side bunker
tank, although there was a gap, less than twelve inches,
between the tank top and the bottom of the container
because there was a securing device to hold the containers
in place. (Tr. 374). Hence, the latter two containers were
exposed to the heated bunker tanks both on the bottom and
on the port side. (PX 92). The adjacent containers —
270602 and 250602 — sat on top of the water ballast
tank, which went down to the outer skin of the ship. (Tr.
369; see PX 92).

The HARMONY had three main bunker tanks: one on the port
side, one on the starboard side, and one in the center.
(Tr. 302-03, 309). Fuel was drawn on alternating days from
the port side tank and the starboard tank. (Tr. 309). The
fuel was transferred from one of these storage tanks to the
service tank. (Tr. 304, 309). When fuel was to be
transferred from one of the storage tanks, that tank would
be heated for half an hour to an hour to warm up the fuel
— to make it “pumpable” to facilitate transfer.
Otherwise, the tanks were not heated. (Tr. 304, 308-11;
Kovshel Dep. 244, 2885). On the date of the fire, the fuel
had not been drawn from the port side bunker tank, no. 5,
the Page 25 tank in question.[fn19]

The temperature in hold 3 ranged from approximately 35°C
to approximately 40°C (or approximately 95° to 104°F). (Tr.
260, 262, 305, 1749, 2115). Balitzki had been in hold 3
about two days before the casualty, and the hold
temperature was slightly above body temperature —
“probably” 40°C. (Tr. 262).[fn20] Kovshel was not concerned
because he knew that the temperature in the hold would not
reach 55°C and therefore there was no risk of violating the
Code. (Kovshel Dep. 400-01)

Both the captain and the chief mate were aware of, and
approved, the stowage positions of the cal-hypo near bunker
tank 5. (Tr. 259-60). The captain explained that he was
comfortable with this stowage plan because it was not
prohibited. (Tr. 260). Likewise, the chief mate approved of
the location. (Kovshel Dep. 242-45, 398-99). The HARMONY’S
Document of Compliance showed that cargo designated class
5.1 could be stored in hold 3. (CX Page 26 21; see also CX
20, 93; Tr. 250-54, 687-88).

The HARMONY had a list of all the dangerous goods
containers on board when the vessel departed Miami on
October 30, 1998. (Tr. 256; CX 30). The vessel also had a
bay plan showing the stowage positions of all the dangerous
cargo containers. (Tr. 256; CX 31). There was no other
hazardous material stowed in hold 3.[fn21] There were some
additional hazardous materials stored under deck in holds 1
and 2. (Tr. 389-90; Kovshel Dep. 360).

4. The P&I Club Circular

On September 1, 1998, in New York, Balitzki received a box
of correspondence, videos, and other materials from
Leonhardt & Blumberg. (Tr. 298-99; CX 39). Included was a
circular from the Swedish Club dated July 1998 setting
forth a “Member Alert.” (CX 40). The circular stated in

In the course of investigating a serious fire onboard a
container vessel, the fire experts appointed by us have
undertaken a research of how Calcium Hypochlorite
(hydrated) reacts to heat exposure. The preliminary
results have such serious implications regarding the
transportation of this material that we have decided to
immediately advise the shipping community of the findings.
Our fire experts have issued the statement below, which
will be brought to the attention of amongst others the
IMO. Page 27

. . .

“The current IMO recommendation is that material is not
to be exposed to a heat source in excess of 55°C for
longer than a 24-hour period. Preliminary results of the
research indicate, however, that for the type and size of
packaging used routinely to ship this material around the
world, the temperatures at which the runaway reaction is
likely to begin may be much lower than that recommended by
the IMO (IMDG Code — Page 5138).

“It is our understanding that the ambient temperature of
holds of certain container ships can reach the critical
ambient temperature of this material, which preliminary
results of our research suggests is considerably lower
than that previously published.

“Accordingly, it is suggested that this material is not
shipped in the holds of such ships and if stowage on deck
is arranged, heating of this material by direct sunlight
must be avoided.”

(CX 40).

5. PPG’s Prior Knowledge of the Risks Presented By the
Shipping of Cal-Hypo

The Swedish Club circular was issued based on a fire and
explosion that occurred in containers of cal-hypo on the
M/V CONTSHIP FRANCE in October 1997, about one year prior
to the HARMONY fire. PPG concluded that the cause of the
explosion was the stowage of the UN 2880 adjacent to heated
bunker tanks. (Ferguson Dep. 573; Lee Dep. 319-22).

PPG knew that cal-hypo “is decomposing all the time. It is
relatively unstable.” (Allen Dep. 86). Its product safety
manager believed that cal-hypo should be stored in “a dry,
cool Page 28 place,” one that was “certainly not warm,
certainly not hot,” meaning 90°F (or 32.2°C) or less. (Lee
Dep. 176-78; accord Ferguson Dep. 102 (“cool” means less
than 90°F); Allen Dep. 575). In fact, the product safety
manager testified: “our literature says 90 degrees F.” (Lee
Dep. 177). He also testified that these guidelines should
apply not just for storage but also for transporting
cal-hypo. (Id. 177-78 (“[c]ertainly” below 90°F); see also
Ferguson Dep. 134).

PPG had done some temperature monitoring in the 1980s in
containers in container ships and found temperatures in the
110° to 115°F range (approximately 43° to 46°C). (Lee Dep.
222, 227). The product safety manager was of the view that
UN 2880 should be restricted to deck stowage, but he was
persuaded that this was a decision that should be left to
the master of the ship. (Id. 73-75).

PPG was also aware that cal-hypo, and in particular UN
2880, had been involved in numerous land-based explosions
or fires over the years, including some resulting in the
destruction of warehouses or large stores. For example, the
former product safety manager of PPG, who retired in April
2000, recalled 10 to 12 PPG cal-hypo fires or explosions,
on-site or off-site, over some 20 years. (Id. 10, 16-18,
130-31). He was aware of another 20 fires or explosions not
involving PPG over the same period. (Id. 131-32; see id.
132-42). There was a cal-hypo fire in a PPG warehouse in
the late 1980s, as a result of which PPG changed its
procedures for storing one-ton bags of cal-hypo to require
Page 29 monitoring of temperatures. (Patterson Dep. 196,
199, 302, 304-05). One of these changes was to ensure that
the “packout temperature,” i.e., the temperature of the
product as it went into the bags, was 90°F (32.2°C) or
less. (Lee Dep. 178-79). In addition, the one-ton bags were
used only when they could be transported by truck within a
radius of eight to ten hours, and only in refrigerated
containers set to a temperature of “probably 70 degrees F,
80 degrees F, something like that,” considerably below
90°F. (Id. 180-81; cf. Ferguson Dep. 390-91 (temperature of
refrigerated trucks carrying “super sacks” of cal-hypo is
set to 60° to 70°F)).

6. PPG’s Use of the 300-Pound Drum

Prior to 1994, PPG did not ship cal-hypo in package sizes
larger than 45 kilos (100 pounds). Customers began
requesting larger packages, however, because larger
packages enabled them to repackage the product more
efficiently. In 1994, PPG began shipping in 400-pound
drums. (Kuhn Dep. 43-44). It chose the 400-pound fiber
drums because that was what was “available” — that
was the drum PPG was using for domestic customers. (Id.

In 1998, as it continued to use 400-pound drums, PPG also
began to ship cal-hypo overseas in 300-pound (or
136-kilogram) drums. This was also done at the request of a
customer. (Id. 61). PPG did not have any special procedures
for utilizing the 300-pound drums. (Patterson Dep. 200). No
new instructions were provided to the loaders when the
300-pound drum Page 30 was introduced. (Id. 274-75). PPG
did not do any testing on the use of the 300-pound drums
for shipping cal-hypo. (Banks Dep. 126-27; see Ferguson
Dep. 809 (“There were no studies done when we [PPG]
instituted the 300-pound drums.”)). Nor did PPG obtain any
material, results, documentation, or reports from the
manufacturer of the drums with respect to any testing done
on the drums. (Banks Dep. 130-31). PPG conducted no testing
of the use of any fiber drums to hold cal-hypo. (Id.
139-40). Nor did PPG consider the impact of stowing
136-kilogram fiber drums of cal-hypo packed three-high in a
20-foot container. (Id. 149-50). Nor did PPG conduct any
tests or research regarding ventilation when cal-hypo is
stored in a shipping container. (Id. 300; Lee Dep. 263-64).
Prior to 1999, PPG had not done any SADT testing of various
sizes of drums holding cal-hypo. (Banks Dep. 170-71; see
id. 380; see also Lee Dep. 161-62, 171-72, 748-49).[fn22]

In August 1999, after the fires on the CONTSHIP FRANCE and
the HARMONY, PPG stopped shipping cal-hypo overseas in
drums larger than 45 kilos. (See Kuhn Dep. 63-64, 70; PX
31; Patterson Dep. 111-12).[fn23] Page 31

7. Other Lawsuits Involving Cal-Hypo

On October 15, 1997, the M/V CONTSHIP, while docked in
Papeete, Tahiti, caught fire following an explosion in a
container holding UN 2880 packed in 193-kilogram (or
425-pound) drums. Eventually, suit was brought against the
manufacturer, PPG. See Conti Zweite Cristallo Schiffarhrts
GMBH & Co. KG v. PPG Indus., Inc., No. 99 Civ. 10545, Tr.
at 1357-67 (S.D.N.Y.). Ruling from the bench following
trial, Judge Stanton found that the fire was caused because
the container was placed directly above the bottom center
fuel tank of the vessel and the vessel’s fuel tanks were
heated to “exceptionally high temperatures.” Id. at
1359-60.[fn24] Judge Stanton dismissed the claims against

On September 7, 1991, the M/V RECIFE caught fire at sea
while bound for North Carolina after an explosion in a
container holding chlorine replacement cartridges that
contained a different form of cal-hypo, UN 1748. Standard
Commercial Tobacco Co. v. M/V “RECIFE,” 827 F. Supp. 990,
992-93 (S.D.N.Y. 1993). The cal-hypo, which was not
manufactured by PPG, was in pellets, packed ten to a
plastic tube, four tubes per box, and four boxes per
carton, with 1,005 cartons packed into the container. Id.
at 992. The container was stowed on deck, where it was
exposed to sunlight, as the result of which the temperature
inside the Page 32 container reached 156°F (68°C). Id.
Suit was brought against the vessel. The court concluded
that the fire was caused by the stowage of the container in
direct sunlight, but concluded that the vessel interests
were not negligent because the stowage was in compliance
with the IMDG Code. Id. at 997-98.

On March 14, 1970, the M/V NICOLAOS D.L. caught fire as it
was preparing to leave the Port of Savannah, Georgia.
Ionmar Compania Naviera, S.A. v. Olin Corp., 666 F.2d 897,
898 (5th Cir. 1982). The fire was caused when a cal-hypo
compound (70% cal-hypo and 30% inert substances)
accidentally spilled and mixed with organic material
(sawdust), igniting a fire. Id. at 898, 902-04. Cargo
interests sued the vessel, the stevedore, and Olin, the
manufacturer of the cal-hypo compound. On appeal, the
circuit court concluded that Olin had a duty to warn the
vessel interests and stevedore of the hazards of which they
could not reasonably have been expected to be aware, but
remanded for further findings on the failure-to-warm claims
and on the issue of whether the stowage was negligent. Id.
at 904-05. The court noted that the cal-hypo compound
“present[ed] a serious fire hazard,” and that when it
“decomposes, enough heat is produced to ignite any
combustible material.” Id. at 899.

8. Testing of Cal-Hypo

Tests were performed on the stability of cal-hypo after
the fire on the HARMONY.[fn25] Page 33

In 1999, PPG asked Safety Consulting Engineers, Inc.
(“SCE”) to perform tests on cal-hypo in two types of
containers: 100-1b plastic buckets with screw-on lids and
300-pound (or 136-kilogram) fiber drums with clamp-on lids.
(PX 50 at 0029, 0032). An oven was built and a container of
cal-hypo was placed into the oven, sitting on a wooden
pallet. The heat was turned on and thermocouples
(temperature recording devices) measured the temperature.
(Id. at 0030). These were SADT tests performed per UN
guidelines. (Id. at 0029). Once the temperature of the
material reached 2°C below the test chamber temperature
(the ambient temperature), the time was noted and the test
was continued for up to seven days to determine whether the
sample temperature increased by at least 6°C. (Id. at 0030,

In the one run performed on 300-1b drums, the oven
temperature was set at 50°C and the temperature increased
by more than 6°C — indeed, it rose to a maximum of
60.9°C in 177.2 hours. (Id. at 0032, 0037; OX 9).[fn26] One
run, using 100-1b drums and a test temperature of 54°C,
resulted in the material going Page 34 into “full
reaction,” with a maximum temperature of 225°C reached in
200 hours. (Id. at 0032, 0035). A third test, using a
100-1b drum and a test temperature of 57°C, also resulted
in an increase of more than 6°C, as the temperature
increased to 67.1°C in just 79.5 hours. (Id. at 0032,

Hence, three of the five runs failed the SADT test. One
run (using a 100-1b drum and a test temperature of 52°C)
resulted in an increase of 5°C and was deemed “borderline,”
while one run (using a 100-1b drum and a test temperature
of 50°C) resulted in an increase of only 4°C. (Id. at 0032,
0038, 0039; see Ferguson Dep. 658-63, 674-83).

In 2001, PPG engaged Touchstone Research Laboratory, Ltd.
(“Touchstone”) to perform some experiments on cal-hypo.
Touchstone built an “oven” large enough to hold a 20-foot
container, which was filled with drums of cal-hypo.
Thermocouples were placed among the drums and elsewhere in
the container. The oven was heated and the thermocouples
recorded temperature changes. The test was run several
times, with adjustments for, among other things, size of
drum and oven temperature. (Tr. 621-32; CX 51).

One test was run in May 2001 using a container filled with
eighty 196-kilogram drums of cal-hypo (stacked two-high),
with the oven temperature set at 41°C (106°F). (Tr. 624-26;
CX 51). The starting temperature of the various materials
was approximately 25°C and it increased to 41°C in roughly
220 hours. (Tr. 628; PX 57). The cal-hypo inside the drums
then increased Page 35 slowly but steadily until roughly
580 hours, when there was a very sharp increase in the
temperatures inside the drums of calhypo. (Tr. 628-29). In
fact, at 609 hours, the oven automatically turned off. As
the Touchstone report explained:

After a total of 609 hours the Cal-hypo in the shipping
container became self-accelerating causing the control
system to trigger the alarm and shut down the heating

(CX 51 at T0833) (emphasis added).[fn27]

C. The Cause and Origin of the Fire

1. The Experts

PPG’s principal expert witness on the cause and origin of
the fire was Dr. Roger McCarthy. I did not find his
testimony credible. He had no recollection of having done a
cause and origin investigation regarding a container vessel
or cargo ship prior to the HARMONY. (Tr. 1659). Although he
had authored some 60 articles, not one involved a vessel
fire or a cause and origin investigation. (Tr. 1646-49). He
had no recollection of using the Code before this case.
(Tr. 1659). His company, Exponent, Inc. (“Exponent”),
provides a wide range of engineering and scientific
consulting services, but he was not listed on Exponent’s
website under “marine services” or “fire and explosions,”
although he was listed under “data/risk analysis,” Page
36 “mechanical engineering,” and “vehicle engineering,”
among others. (Tr. 1660-63). He was mistaken as to the
meaning of the term “trim”; the trim of the HARMONY was
down by the stern, meaning unburned fuel that hit the floor
would flow toward the stern. (Tr. 1709-11).