Shortly after midnight on March 24, 1989, the tanker Exxon Valdez ran aground on Bligh Reef in Prince William Sound, Alaska. The response plan written for that kind of spill described equipment and timelines that the people on scene could not deliver that night. That gap between a plan on paper and what responders could actually do led Congress to pass the Oil Pollution Act of 1990, which built the Area Contingency Plan system and the unified command that spill responders still use. This piece covers what is established about the spill, the disputed volume, and how spill doctrine changed afterward.

Bligh Reef, March 24, 1989: what the record establishes about the grounding

The Exxon Valdez was a single-hulled tanker that loaded North Slope crude at the Alyeska Marine Terminal in Valdez, at the southern end of the Trans-Alaska Pipeline. It sailed on the evening of March 23, 1989. The Exxon Valdez Oil Spill Trustee Council, the state and federal body created to manage the civil settlement, puts the cargo at about 53 million gallons. To avoid ice from the Columbia Glacier, the ship left the outbound traffic lane. It did not return to the lane in time, and it struck Bligh Reef shortly after midnight on March 24, which was Good Friday.

The National Transportation Safety Board investigated the grounding and adopted its marine accident report in 1990. That report contains the board’s probable cause statement along with its findings on the navigation of the vessel, the crew and the oversight of tanker traffic in the Sound, and a reader who needs the causal record should read the statement in the report itself rather than in any summary, including this one. I am deliberately not paraphrasing the probable cause here, because this piece is about the response, and because shortening a multi-part finding into one paragraph is how the shorthand versions of this accident got started in the first place. The site’s article on reading a closed NTSB case explains how to find the docket and what a probable cause statement is designed to do. The State of Alaska’s oil spill commission published its own separate analysis in 1990, and that report is the better source on the state’s side of the response.

The popular one-line account that pins the spill on the captain’s drinking is an oversimplification and should be labeled as one wherever it appears, because the board’s finding has several parts and the criminal case did not sustain the intoxication charge. An Alaska jury acquitted the master, Joseph Hazelwood, in 1990 of operating a vessel while intoxicated and convicted him of the misdemeanor of negligent discharge of oil. I was not involved in this response in any capacity. For this piece I am working from the NTSB report, the State of Alaska’s oil spill commission report, the Trustee Council’s published material and the 1990 statute. My subject is the response and the doctrine that followed it, so the navigation of the ship gets only this section.

The volume dispute: 10.8 million gallons and the higher estimates

The figure most official bodies use is about 10.8 million gallons, or roughly 257,000 barrels, and the Trustee Council’s public material commonly rounds it to 11 million gallons. Federal and state agencies and most later reference works use the same number. It comes from cargo accounting, which compares the volume loaded at Valdez with the volume recovered when the remaining oil was pumped off the grounded ship into other tankers. That offloading, called lightering, moved most of the cargo safely and is one of the few parts of the early response that went largely as hoped.

Higher estimates also circulate, ranging up to about 38 million gallons. They come from critics of the official accounting, who generally argue that the cargo measurement method could have understated the loss. I have not been able to tie the 38 million figure to a single primary document that shows its method. For that reason I report the high end as a claim in circulation and the low end as the official figure, and I do not treat either as proven against the other. A reader who needs to rely on a number for a paper or a briefing should go to the Trustee Council’s material for the official figure and to the specific critic’s own publication for the higher one, and should read both methods before choosing how to cite it.

For planning purposes, the volume is less useful than the footprint. The Trustee Council states that oil reached about 1,300 miles of shoreline, with roughly 200 miles heavily or moderately oiled. That spread is what turned a containment problem into a months-long shoreline operation. Under the official figure, the spill was the largest in United States waters until the Deepwater Horizon blowout in 2010. Valdez has never ranked near the top of the world’s largest spills by volume under any estimate. Its influence came from where it happened and from how visibly the response fell short.

The Alyeska plan and the first three days: equipment, dispersants and the storm

First-response duty for a spill in the Sound belonged to the Alyeska Pipeline Service Company, the consortium that ran the pipeline and terminal on behalf of the oil companies that owned it. Alyeska operated under a contingency plan for Prince William Sound that the state had approved. That plan listed equipment and described a response to a large tanker spill. The state commission’s report and later accounts of the first day describe a very different situation. The barge meant to carry containment boom and skimmers was out of service and had to be reloaded, and the containment gear reached the tanker many hours after the grounding, much later than the plan contemplated. The plan existed and had been approved. What the plan did not reflect was the condition of the equipment and the staffing on the night it was needed.

The weather during the first days was calm. In hindsight that was the best window for mechanical containment at sea, and the responders did not have the gear in place to use it. Exxon pressed for dispersant use. Test applications were made, and the company, the Coast Guard and the state disagreed over whether the tests had worked and over how much application to approve. Calm water is a poor condition for dispersants that depend on wave energy to mix oil into the water column, which complicated the argument. An in-situ burn test was also carried out early in the response. Published accounts differ on how much a faster dispersant decision would have changed the outcome, and I do not think the record settles that question, so I leave it open.

A strong storm arrived around the third day and spread the slick across the Sound and then southwest along the coast. After that, the job became shoreline cleanup across hundreds of miles of remote, rocky coast. The workforce grew into the thousands, with hundreds of vessels, many of them local fishing boats hired into the effort. The lesson the reformers drew was specific. A plan had been approved without anyone verifying that the listed equipment could be loaded, moved and deployed in the hours the plan assumed. The decision process for dispersants had not been worked out before the oil was on the water.

Who was in charge: Exxon, the Coast Guard and the State of Alaska

Under the National Contingency Plan as it stood in 1989, the responsible party was expected to clean up its own spill. A federal on-scene coordinator monitored that work and could take over the response, which responders call federalizing it. In coastal waters the Coast Guard supplied that coordinator. Within the first day or so, Exxon took over the response from Alyeska and accepted responsibility for the cleanup. The State of Alaska, through its Department of Environmental Conservation, had its own legal authority, its own priorities and its own views on dispersants and shoreline methods. All three organizations had a legitimate claim to direct some part of the work, and no standing structure combined them.

President George H. W. Bush decided against federalizing the cleanup. He sent Transportation Secretary Samuel Skinner and EPA Administrator William Reilly to Alaska and directed military support to the effort. In the weeks after the grounding, the Coast Guard sent Vice Admiral Clyde Robbins to serve as federal on-scene coordinator, which raised the rank of the federal presence without changing the basic arrangement. Exxon ran and paid for the cleanup, the Coast Guard oversaw it, and the state pushed its positions from outside the chain of decision. The three parties settled disputes by negotiation, which was slow when daylight, weather and oil movement all favored fast decisions.

Local governments and the Alaska Native villages in the path of the oil, including Tatitlek and Chenega Bay, had no defined seat in that arrangement. The fishing community of Cordova organized much of its own early protective work. Marine spill response in 1989 had not yet adopted the incident command system that came out of the California wildland fire service in the 1970s, which the site’s FIRESCOPE article covers. Spill organizations worked under their own structures, and the Valdez experience was a large part of the argument for moving them onto ICS during the following decade.

The Oil Pollution Act of 1990: liability, the trust fund and mandatory response plans

President Bush signed the Oil Pollution Act of 1990 on August 18, 1990, about 17 months after the grounding. It is usually called OPA 90. The law amended the oil spill provisions of the Clean Water Act and put several separate reforms into one statute. It made responsible parties for vessels and facilities strictly liable for removal costs and damages up to statutory limits. Those limits do not apply in defined circumstances such as gross negligence or violation of applicable regulations. The law also made the Oil Spill Liability Trust Fund usable. Congress had established that fund on paper in 1986 without authorizing its use, and OPA 90 allowed it to pay removal costs and damages when a responsible party could not or would not. The Coast Guard’s National Pollution Funds Center administers it.

For response doctrine, the most important change was that tank vessels and certain facilities had to hold response plans meeting federal standards. Those plans must address a worst-case discharge, identify the contracted resources that would respond, and be consistent with the National Contingency Plan and the applicable Area Contingency Plan. That consistency requirement tied private plans to government planning for the first time in a systematic way. A company could no longer write a plan in isolation and have a single state agency approve it, as had happened with the Alyeska plan. Federal agencies now reviewed vessel and facility plans against a common structure, and the response organizations named in them had to be real and under contract.

The law also addressed prevention and the specific conditions in Prince William Sound. It set a schedule to phase single-hull tankers out of United States waters, with the last of them due out by 2015. It required escort vessels for certain laden tankers in the Sound. It provided for regional citizens’ advisory councils for Prince William Sound and for Cook Inlet, which gave communities near the terminals a formal oversight role that they had lacked in 1989. Alaska passed its own spill legislation in the same period. Readers who need the current state requirements should verify them with the Alaska Department of Environmental Conservation and not rely on summaries of the early 1990s rules.

The planning tiers: National Contingency Plan, regional teams and Area Contingency Plans

The National Contingency Plan dates from 1968, when it was written after the Torrey Canyon spill off England in 1967, and since its revision in the mid-1990s to carry out OPA 90 it has sat at the top of a structure with several tiers. The National Response Team coordinates federal agencies at the national level. Regional Response Teams, co-chaired by EPA and the Coast Guard, include federal and state members and maintain regional contingency plans. Below them, OPA 90 created Area Committees and required each committee to prepare an Area Contingency Plan for its geographic area, under the direction of the federal on-scene coordinator for that area. The Coast Guard leads in the coastal zone and EPA leads in the inland zone, and the plans define where that boundary falls.

Area Committees are made up of qualified people from federal, state and local agencies, and in many areas tribal and industry participants attend as well. The Area Contingency Plan describes the area and its sensitive resources, identifies who does what, and lists the response equipment available. The statute also requires procedures for expediting decisions on dispersants and other chemical countermeasures. That requirement answers the Valdez dispersant argument directly, and in practice regional teams have used it to set up preauthorization zones and decision procedures before a spill happens. Many Area Contingency Plans also include geographic response strategies, which are site-specific instructions for booming and protecting a particular inlet, intake or wetland so that the first crews on scene do not have to design the tactic under pressure.

A county emergency manager in Georgia or anywhere else inland can easily assume this system is a coastal matter that belongs to the Coast Guard. EPA regions maintain inland area planning that covers rivers, pipelines, rail corridors carrying crude and fuel, and bulk storage facilities. Onshore oil pipeline operators file response plans with the Pipeline and Hazardous Materials Safety Administration, and certain non-transportation facilities file plans with EPA. In some regions the regional and area plans have been combined into one integrated document. The fastest way to find the plan that covers your jurisdiction is to ask your EPA regional office or your Coast Guard sector which document applies and where it is posted.

Unified command for spills: the federal, state and responsible party seats

The command structure that came out of the post-Valdez reforms is a unified command with three standard seats. The federal on-scene coordinator, the state on-scene coordinator and the responsible party’s incident commander sit together and set joint objectives, and local government and tribal representatives join when their jurisdictions are affected. The Coast Guard moved its spill response organization onto the incident command system during the 1990s and later published an Incident Management Handbook as its field reference. I am not going to put a single adoption year on that change, because the move happened in stages and the handbook has gone through several editions. Readers who need the current Coast Guard version should get it from the Coast Guard directly.

The design solves the specific Valdez problem of three parties with legitimate authority and no shared process. Under unified command they produce one set of objectives and one incident action plan, and a single operations section carries it out. The federal on-scene coordinator keeps the authority that the National Contingency Plan gives that role. That authority includes directing the response when the coordinator judges it necessary and federalizing when the responsible party is not performing. Unified command is a way of reaching decisions together, and the federal coordinator’s legal position does not shrink by joining it.

As a communications person I pay particular attention to one practical consequence. Each party arrives with its own radios, phones and contractors. The responsible party’s oil spill removal organizations may come from out of state on their own systems. A spill command post can therefore end up with three organizations that cannot talk to one another on the water or the shoreline. The ICS 205 radio communications plan is where that gets solved or missed. Many Area Contingency Plans include communications sections, and a local radio administrator should read that section before an incident, because it shows which channels federal and industry responders expect to use and whether the county’s radios carry any of them.

What the reforms did not settle: shoreline cleanup, plan drift and exercises

The Valdez cleanup also showed that response actions have environmental costs of their own. Crews washed large stretches of oiled shoreline with high-pressure hot water. Studies supported by NOAA later found that some intertidal communities treated that way recovered more slowly than comparable oiled areas left untreated. That finding fed the practice now called net environmental benefit analysis, in which responders weigh the harm a cleanup method does against the harm of the oil it removes. NOAA and Trustee Council studies have also found oil remaining below the surface of some beaches decades after the spill, and the Trustee Council’s publications track which injured resources it considers recovered.

The deeper problem OPA 90 addressed was the plan that looked sound on paper, and a statute alone could not fix it. At a 2010 House hearing during the Deepwater Horizon response, members pointed out that the Gulf of Mexico response plans filed by several major oil companies were nearly identical and listed walruses among the wildlife to protect, an animal that does not live in the Gulf. Those were plans that had passed federal review. The national commission that examined Deepwater Horizon published its report in January 2011, and anyone who wants its findings on response planning should read that report directly. The response itself ran under a unified command with a National Incident Commander added at the top, which showed both that the post-1990 structure could expand to a very large event and that the plans feeding it could still drift away from reality.

Federal exercise guidance exists because of that risk. The National Preparedness for Response Exercise Program, known as PREP, was first issued by the Coast Guard and other federal agencies in the mid-1990s. It sets out how plan holders and area committees satisfy the OPA 90 exercise requirements on a multi-year cycle that includes equipment deployment and unannounced drills. Readers should verify the current PREP guidelines with their Coast Guard sector or EPA region and not rely on older summaries. The principle behind them follows directly from March 1989, because a plan only becomes a capability once the equipment it lists has been deployed by the people it names in roughly the time it promises.

What to do at your agency

• Have your emergency manager ask the EPA regional office or the Coast Guard sector which Area Contingency Plan covers your county, download the current version, and put a one-page summary of its geographic response strategies for your waterways on the next LEPC meeting agenda.
• Have your water utility director confirm, using that same plan, whether every drinking water intake in the county is listed as a sensitive resource, and send the Area Committee a written correction for any intake that is missing.
• Have your fire chief or hazmat officer ask each pipeline operator in the county for the emergency contact and response zone information from its federal response plan, and confirm that those numbers are entered correctly in the CAD premise records for every pipeline crossing and station.
• Have your radio system administrator read the communications section of the Area Contingency Plan and write one paragraph into the county communications annex naming which listed channels county radios can and cannot reach, and who would bring a gateway or cache radios to a spill command post.
• Have your emergency manager add a spill inject to the next scheduled hazmat tabletop that seats a federal on-scene coordinator, a state on-scene coordinator and a responsible party incident commander, and record in the after-action report which local official took the local seat and what written priorities that person brought.
• Have your EOC manager get the name and duty contact for the state on-scene coordinator from your state environmental agency and add that entry to the EOC contact directory, with the date it was verified.

Takeaways

• The Exxon Valdez grounded on Bligh Reef shortly after midnight on March 24, 1989, and the causal record is the NTSB’s 1990 marine accident report, which should be read directly because the popular drunk-captain shorthand is an oversimplification of a multi-part finding.
• The official volume of about 10.8 million gallons comes from the Exxon Valdez Oil Spill Trustee Council and government cargo accounting, while critics of that accounting have published estimates up to about 38 million gallons, and any citation should name its source.
• The Alyeska contingency plan had been approved, but on the night it was needed the equipment it listed was not ready to deploy in the time the plan assumed.
• In 1989 Exxon, the Coast Guard and the State of Alaska each held real authority without a shared command structure, and local governments and Native villages had no defined seat.
• The Oil Pollution Act of 1990 made responsible parties strictly liable within statutory limits, made the trust fund usable, required worst-case vessel and facility response plans, and phased out single-hull tankers.
• OPA 90 created Area Committees and Area Contingency Plans under the federal on-scene coordinator, and private response plans must be consistent with them, including in inland areas covered by EPA.
• Spill response now runs under a unified command with federal, state and responsible party seats, and the federal on-scene coordinator keeps the authority the National Contingency Plan assigns.
• The 2010 walrus references in Gulf of Mexico plans showed that federally reviewed plans can still drift from reality, which is the problem the PREP exercise requirements exist to catch.