IncidentNews Home  >>  Incident  >>  Entry

Countermeasures/ Mitigation

The Alyeska Pipeline Service Company was immediately notified of the incident
and sent a tug to the site to assist in stabilizing the vessel.  At the time of
the incident, the Alyeska spill response barge was out of service being
re-outfitted.  It arrived on scene by 1500 on 24 March.  Alyeska was overwhelmed
by the magnitude of the incident; by March 25, Exxon had assumed full
responsibility for the spill and cleanup effort.

Deployment of boom around the vessel was complete within 35 hours of the
grounding.  Exxon conducted successful dispersant test applications on March 25
and 26 and was granted permission on March 26 to apply dispersants to the oil
slick.  Due to the large storm that began the evening of March 26, much of the
oil turned into mousse. As dispersants aren't generally able to dissipate oil in
the form of mousse, it was no longer practical to use dispersants on floating
oil during this response.

On the evening of March 25, a test in-situ burn of oil on water was
conducted.  Approximately 15,000 to 30,000 gallons of oil were collected using
3M Fire Boom towed behind two fishing vessels in a U-shaped configuration, and
ignited.  The oil burned for a total of 75 minutes and was reduced to
approximately 300 gallons of residue that could be collected easily.  It was
estimated that the efficiency of this test burn was 98 per cent or better.
Again, continued in-situ burning was not possible because of the change in the
oil's state after the storm of March 26.  

Five dispersant trials took place
between March 25 and March 28.  Corexit 9527 was used for the trials.  Four of
the tests used C-130 aircraft with ADDS packs, and one test was applied from a
DC-6 aircraft.   By March 29 the Regional Response Team (RRT) decided that
dispersants were no longer feasible.

Because there was not enough equipment to
protect all the shorelines that could be impacted, Federal, state and local
agencies collaborated to establish shoreline protection priorities.  The
agencies decided that fish hatcheries and salmon streams had the highest
priority; accordingly, containment booms were deployed to protect these areas.
Five fish hatcheries in Prince William Sound and two in the Gulf of Alaska were
boomed, with the largest amount of  boom deployed at the Sawmill Bay hatchery in
Prince William Sound.  On April 15, the Sawmill Bay hatchery was boomed with
30,500 feet of sorbent boom and 28,600 feet of containment boom in multiple
layers.  As many as 15 to 20 boats were used daily for tending the boom and oil
recovery by towing sorbent boom.  Overall, the deflection of oil from the
hatcheries was very successful.

At the height of containment efforts, it is
estimated that a total of 100 miles of boom was deployed.  Almost all the types
of boom available on the market were used and tested during the spill response.


Due to the size of the spill, it was necessary to employ inexperienced workers
to deploy and tend booms, and this led to some boom being incorrectly used or
handled, and sometimes damaged.  Some boom sank because of improper deployment,
infrequent tending, or leakage and/or inadequacy in the buoyancy system.  Other
problems included fabric tears in boom due to debris, and tearing at anchorage
points from wave action.  In some cases, ballast chains were ripped off during
boom recovery if the boom was lifted by the chain.  One estimate suggests that
50 per cent of the damage to larger boom came during boom recovery.  For
self-inflating booms, it was important to keep the inflation valves above the
water during deployment so that the boom did not become filled with water and
have to be replaced.  

Since most of the containment boom was in  50 to 100
feet long sections, several lengths of boom usually needed to be connected for
deployment.  When several types of boom were used in one operation, there were
often problems with incompatible connectors between different types of boom. 
Bailing wire and other adaptations were used in the field for these situations.
 A universal type of connector (ASTM connector) came with some booms, but these
were difficult to handle and hook up at sea and were hard to open once they had
been submerged in cold water.  Booms to be re-used were hand cleaned early on in
the spill, and as the spill progressed were cleaned in one of the two barges
with mechanical washing facilities.

To contain oil on the open water,
containment boom was towed between two vessels (usually fishing boats) to
surround the oil and then the two ends of the boom were drawn together to close
the loop and await collection by a skimmer.

Aerial surveillance was used to
direct the deployment of booms and skimmers for open water oil recovery.  Visual
overflight observations as well as ultraviolet/infrared (UV/IR) surveys were
used by the USCG and Exxon to track the floating oil.  Satellite imagery was
also tested as a method to track oil but was not very useful because of the
infrequency of satellite passes over Prince William Sound (every 7 to 8 days),
cloud cover, and lengthy turn around time for results.

The primary means of
open water oil recovery was with skimmers.  In general, most skimmers became
less effective once the oil had spread, emulsified and mixed with debris.  To
save time, it was most practical to keep skimmer offloading equipment and oil
storage barges near the skimmers.

Weir skimmers were useful for collecting
fresh oil that was present in a thick layer on the water.  As the oil became
weathered and laden with debris, however, it was the simple weir skimmers that
were the first to clog.  Some of the larger weir skimmers had auger pumps with
cutters for chopping debris and were able to collect oil for a longer time than
the simple models.

Oleophilic disc skimmers also worked well while the oil
was fairly fresh.  Once the oil became viscous and associated with debris, these
skimmers were not very effective.

An Egmolap brand paddle belt skimmer (Egmolap
II) was used and was effective for heavy mousse and debris.  It collected very
little water under light sea conditions.  A different paddle belt skimmer that
was supplied by the Canadian Coast Guard clogged easily when working with
viscous oil.

When using rope mop skimmers, it was important to maintain the
smallest angle possible when lifting the skimmer out of the water, so that the
oil did not run down the mop and back into the water.  In situations where the
oil was viscous, it was useful to cut down the diameter of the mop from nine to
six inches and inject diesel oil into the ringers as the mop was being rung
out.

The most used skimmers during the response were the Marco sorbent
lifting-belt skimmers that were supplied by the U. S. Navy.  Once oil became
viscous, the sorbent part of the skimmer was removed and the conveyor belt alone
was sufficient to pull the oil up the ramp.  The pump that came with the skimmer
had difficulty offloading viscous oil, so that other vacuum equipment was used
to unload the collected oil.  The Marco skimmers were generally not used close
to shore because they drawbetween three and four feet.  In general, the paddle
belt and rope mop skimmers were the most useful for recovery of oil from the
shoreline.  The skimmers were placed on self-propelled barges with a shallow
draft.

Sorbents were used to recover oil in cases where mechanical means were
less practical.  The drawback to sorbents was that they were labor intensive and
generated additional solid waste.  Sorbent boom was used to collect sheen
between primary and secondary layers of offshore boom, and to collect sheen
released from the beach during tidal flooding.  Pompoms were useful for picking
up small amounts of weathered oil.  Towing of sorbent boom in a zigzag or
circular fashion behind a boat was used to collect oil and was more efficient
than towing the boom in a straight line.  Sorbent booms made of rolled pads were
more effective than booms made of individual particles because these absorbed
less water and were stronger, and did not break into many small particles if
they came apart.

During the Exxon Valdez spill response, a hopper dredge was
used to collect oil for the first time in the United States.  The oil was
gathered using a containment boom, and the draghead of the dredge was placed
under the boom below the oil surface.  The oil was then sucked up and placed in
storage containers on the dredge.  The drawbacks to using the dredge were that
it recovers large amounts of water with the oil and must be used offshore
because of its deep draft.

To transfer the recovered oil, water, and debris
mixture from the skimmers to temporary storage containers, vacuum equipment and
positive-displacement pumps were used.  Vacuum trucks on barges or air-conveyers
were most useful when used with an open-ended suction hose with a diameter of 6
to 8 inches.

Early on in the response, storage space for recovered oil was in
short supply.  To combat the storage space problem, water was decanted from
skimmers or tanks into a boomed area before offloading.  As a result, the
remaining viscous oil mixture was difficult to offload, the process sometimes
taking up to 6 to 8 hours.  High-capacity skimmer offloading pumps, in
particular grain pumps, were the most useful in transferring viscous
oil.

Because recovery equipment was in near constant use, several vessels were
set up to perform field repairs and conduct preventive maintenance.

The oil
remaining on the Exxon Valdez, was completely offloaded by the end of the first
week in April 1989.  After offloading operations were completed, the tanker was
towed to a location 25 miles from Naked Island in Prince William Sound for
temporary repairs.  Later in the summer of 1989, the vessel was brought to
California for further repairs.

Shoreline assessment was a prerequisite for the
implementation of any beach cleanup.  Assessment provided geomorphological,
biological, archaeological and oiling information that was used for the
development of site specific treatment strategies.  Cleanup operations were
scheduled around specific activities such as seal haulout activity, seal
pupping, eagle nesting, fish spawning, fishing seasons, and other significant
events as much as possible.

In 1989, hoses spraying seawater were used to flush
oil from shorelines. The released oil was then trapped with offshore boom, and
removed using skimmers, vacuum trucks (useful for thick layers of oil) and boom
(sorbent, snare, pompoms).  For hard to reach areas, or locations with weathered
oil, heated seawater was used to flush oil from the shoreline.

Converted
vessels and barges were used for beach washing operations.  It would take
several days to outfit a conventional barge with the equipment needed to heat
and pump the water.  Smaller vessels that were used for beach washing early in
the spill were re-outfitted for bioremediation later in the response.

Along
with the large scale beach washing, manual cleanup, raking and tilling the
beaches, oily debris pickup, enhanced bioremediation and spot washing were used
to cleanup the oil.  In some locations, oil was thick enough to be picked up
with shovels and buckets.  In addition, mechanical methods were used on a few
sites, including the use of bulldozers to relocate or remove the contaminated
beach surfaces.  Mechanical rock washing machines, which were manufactured for
the spill, were not used to clean contaminated rocks and return them to the
beach.

Oiled storm berm was mechanically relocated in some cases so that these
areas, which normally would not receive much wave action, would be more exposed
and cleaned by natural processes.  If the oiling in the berm was significant or
persistent it was tilled to free the oil or washed to optimize the cleaning. 
Recommendations were made to restrict the movement of berm to the upper third of
the beach to ensure its return to the original location.

Beach applications of
dispersants were tried in several locations.  Corexit 7664 was applied on Ingot
Island, followed by a warm water wash. No significant change in oil cover or the
physical state of the oil was observed as a result of the treatment. Some
ecological impacts were observed in the treated areas.   It appeared that the
effects were largely due to the intensive washing more that the use of Corexit
7664, and were evident in intertidal epibenthic macrobiota.

In addition, the
dispersant BP1100X was applied to a test area on Knight Island.  Toxicology
studies indicated that the upper and lower intertidal biota were different from
pre-application communities the day after dispersant application, and returned
to pre-treatment levels after seven days.

Exxon also tested the dispersant
Corexit 9580 in Prince William Sound.  The decision to approve a large scale
test of Corexit 9580 in August was reached after an extensive program aimed at
evaluating shoreline cleaning technologies.  The monitoring program addressed
three major issues:  migration of oil and Corexit in shoreline sediments, the
migration of sediments and oil in the nearshore environments, and the migration
of oil in the water column, each being evaluated in the monitoring program.  
The dispersant's effectiveness and impact were then compared to mechanical
shoreline cleanup methods, and this information was used to determine whether
Corexit 9580 should be used for shoreline treatment.  The Research and
Development Committee evaluating the proposal for dispersant use recommended
against broad-scale application of the product because tests had not adequately
demonstrated that removal and recovery efficiency outweighed possible adverse
effects.  The committee recommended using Corexit only on Smith Island, subject
to continued review of the effectiveness of recovery procedures by on-scene
monitors.

In May of 1989, the Environmental Protection Agency (EPA) and Exxon
conducted bioremediation trials at two test sites on Knight Island in Prince
William Sound.  On the basis of these tests and other trials later in the
summer, Exxon recommended the use of the bioremediation enhancement agents,
Inipol (Inipol EAP22--manufactured by Elf Aquitaine of France) and Customblen
(Customblen 28-8-0 --manufactured by Sierra Chemicals of California), and
subsequently treated over 70 miles of shoreline in Prince William Sound with
these agents.

Past scientific research had determined that sufficient numbers
of hydrocarbon degrading bacteria existed naturally in Alaska.  It was decided
that the limiting factor in enhancing petroleum hydrocarbon degradation was the
availability of nitrogen and phosphorus for the indigenous bacteria.  As a
result, bioremediation trials focused on agents that were basically
"fertilizers", and contained no living microorganisms.  Considerations in the
selection of bioremediation agents included ease of application, the possibility
of causing algal blooms and eutrophication in areas where nitrogen/phosphorus
concentrations would remain elevated (such as sheltered bays and estuaries), the
flushing of nutrients from the beach soon after application due to tidal action,
and the possible toxicity associated with concentrations of nitrogen based
compounds (such as ammonia).

Winter monitoring of the effects of bioremediation
consisted of surveys of more than 20 beaches in Prince William Sound and the
Gulf of Alaska.  These studies determined that oil degradation had been enhanced
on the shorelines monitored, but some debate existed over whether bioremediation
was solely, or even largely, responsible.

Cleanup operations in 1989 ceased by
the end of September.  All parties involved in the response agreed that
continuation of cleanup into the Alaskan winter would jeopardize the safety of
cleanup crews.  In addition, it was speculated that the winter storms in Alaska
could significantly remove oil from shorelines, including sub-surface oil.  By
the end of the 1989 cleanup, more than 25,000 tons of oiled waste and several
hundred thousand barrels of oil/liquid waste were collected and disposed of in
landfills.

Cleanup in 1990 began in April and ended in September.  Surveys in
the spring of 1990 showed that oiling conditions had been reduced or changed
over the winter.  Surface oil in 1990 was significantly weathered but
sub-surface oil was relatively fresh in some locat
ons.   Cleanup techniques in 1990 focused more on manual methods of treatment
such as hand wiping and spot washing as well as bioremediation.  Mechanical
equipment was used on a few sites.

Bioremediation was more extensive in 1990,
with 378 of the 587 shoreline segments treated that year receiving
bioremediation application.  In general, Inipol was applied in cases where
surface oiling existed and Customblen slow release pellets were preferred for
treating beaches with sub-surface oiling.  Generally, beaches were given one to
three treatments over several months.  Concern over the possible toxicity of
Inipol led to recommendations for application of only Customblen on some
sites.

By the spring of 1991, the scope of the cleanup effort was greatly
reduced.  Manual cleanup, bioremediation, and very limited use of mechanical
equipment were employed.  Cleanup took place from May of 1991 through July of
1991.

An important observation that resulted from the Exxon Valdez oil spill
was that natural cleaning processes, on both sheltered and exposed beaches, were
in many cases very effective at degrading oil.  It took longer for some sections
of shoreline to recover from some of the invasive cleaning methods (hot water
flushing in particular) than from the oiling itself.