Improving Response Planning by Using Numerical Models to Test Geographic Response Plans

After the tragic Exxon Valdez oil spill occurred in 1989, the U.S. government enacted legislation to require greater preparation for these incidents in an effort to better protect the environment. The Oil Pollution Act of 1990 (OPA 90) created Area Committees to cover all areas of the country. These Area Committees are required to prepare an Area Contingency Plan (ACP) in accordance with 40 CFR 300, the National Oil and Hazardous Substances Contingency Plan (NCP). The ACP is required by Title IV, Section 4202 of the OPA 90 which amended Subsection (j) of Section 311 of the Federal Water Pollution Control Act (FWPCA) (33 U.S.C. 1321 (j)) as amended by the Clean Water Act (CWA) of 1977 (33 U.S.C. 1251) to address the development of a National Planning and Response System.

As defined by the Clean Water Act Sections 311(a)(19) and (j)(4), the ACP is “developed to be implemented in conjunction with the NCP and RCP (Regional Contingency Plan), in part to address removal of a worst case discharge and to mitigate or prevent a substantial threat of such a discharge from a vessel, offshore facility, or onshore facility operating in or near an area designated by the President.” Over time, ACPs have become more detailed to assist with the preparations and prevention for a response.

One of these improvements was the addition of Geographic Response Plans (GRP) or Geographic Response Strategies (GRS). These plans are developed to assist responders in an actual event with guidance on where and how to deploy containment and recovery equipment to best protect sensitive sites in the ACP Planning Area. The U.S. Coast Guard's Area Contingency Planning Process Job Aid published in December of 2012 defines a GRP as “a highly focused plan that uses a large scale to identify Environmentally Sensitive Areas/Economically Significant Areas (ESA) and detail the protection and response strategies that may be employed. A GRP is effective for organizing and displaying information in a functional format for immediate use during both planning and response and for assistance in developing a Common Operating Picture.”

The GRP/GRS is not required in the event of an incident in the applicable area. They are intended to help the Unified Command determine how much equipment should be ordered for certain regions forecasted to be impacted by the spill. Information typically included in the GRP/GRS are:

• (1) A satellite image of the area covered

• (2) A graphic display of the strategy showing where the equipment is deployed as well as locations of staging areas, sensitive environmental sites, boom anchor locations, etc.

• (3) A list of equipment required to perform the strategy, including length and type of boom, number and type of skimmer, number of personnel to deploy, etc.

• (4) A list of emergency contacts in the local region, including government agencies, stakeholders, natural resource trustees, and Oil Spill Removal Organizations (OSRO)

Typically, a team of members from the Area Committee, including federal and state government agency, industry, and OSRO personnel, produce the GRP/GRSs during workshops where they sit around a table and discuss the best strategies for different locations. While these experts bring a wealth of experience, they cannot predict how a multitude of oil products will behave in different seasons under varying environmental conditions.

One method typically used to test the GRP/GRSs is to schedule deployments of oil spill response equipment in the field to see how well the plan works. Occasionally, some type of product is released into the water to simulate the oil, such as rice hulls or oranges, to observe where the product goes in the waterway. While field deployment exercises are immensely valuable, they cannot replicate all possible situations, such as how oil would move differently during different seasons or how varying oil types will behave. These exercises are at the mercy of poor weather conditions and can be very expensive.

Area Committees around the United States vary in their views of the detail required for the GRP/GRS. In some regions, every aspect of the response is listed. In others, only general information is provided, such as which sites should be considered as the highest priority and whether deflection, exclusion, or collection booming should be used. Planners in these areas believe that, due to the many variables involved in any oil spill response effort, it is futile to attempt to plot every detail of equipment deployment. The Coast Guard's Area Contingency Planning Process Job Aid does provide specific recommendations on information to include in the GRP/GRS to ensure that ESAs are sufficiently protected.

Change 1 of the U.S. Coast Guard Marine Environmental Response and Preparedness Manual published in September 2018 provides specific guidance to Area Committees on how the GRS should be validated. The Manual recognizes that it is impossible to use field deployment exercises to test the large quantity of GRSs listed in each ACP. Area Committees are encouraged “to employ a risk based decision-making methodology” in their efforts to validate these important strategies. Table 1 below is provided in the USCG MER Manual to assist Area Committees with this process.

Validation Level I of these requirements describe the validation of the strategies in a Desktop setting. The Coast Guard indicates that all GRS data in the ACP should be validated to a minimum of this level. Computer simulations are included as an option to supplement a simple “gut check” evaluation of the strategies. These computer simulations are the best method for testing all possible scenarios of a response and ensuring that ESAs can be protected.

Several oil spill trajectory and fate models are available through government or commercial resources. The RPS model, OILMAP, provides the typical capability to predict where the oil will move and how the oil will behave and weather, as well as the ability to employ response actions, such as booming, skimming, dispersant application, or in-situ burning, to evaluate how effective they will be in different scenarios. Extensive environmental datasets can be accessed to ensure the quality of the inputs to this modeling effort. Quality input data results in quality trajectory and oil fare information. Using these types of computer simulations can significantly benefit the ACP planning process by allowing Area Committee members to quickly analyze numerous response options and oil spill scenarios over a short time period.

GRS data from three different ACPs are now presented with computer simulations run in the RPS OILMAP model. These examples illustrate how modeling can show when a strategy works and when a strategy is ineffective. Modeling can also assist with changing the strategy by adding or simply editing the equipment deployments planned to create a new strategy capable of protecting the ESAs. GRS from the Plymouth to Salisbury, Massachusetts ACP (USCG Sector Boston), the Philadelphia ACP (USCG Sector Delaware Bay), and the Southeast Louisiana ACP (USCG Sector New Orleans) are used for this process. The following figures are derived from screenshots from the OILMAP model.

OILMAP was used to model a release of 1 million gallons of heavy fuel oil into Boston Harbor from a vessel at the Black Falcon Cruise Terminal. The oil was released at the surface over a period of four hours. The booming strategy simulated was according to the GRS located in the ACP for the Boston area. Boom laid at the mouth of the Boston Main Channel and the Reserved Channel are not immediately effective because the oil slick spreads beyond the identified boom deployment locations before the boom could realistically be put in place (Figure 1). Once the boom is deployed, it effectively traps the oil for further recovery efforts (Figure 2).

According to the model, the booming strategy for protecting Pleasure Bay (Figures 1 and 2) is effective at keeping oil out of this ESA. While the double chevron boom configuration stops oil from entering Pleasure Bay, the adjacent areas are impacted.

Along the Neponset River, four sequential lines of boom are deployed in accordance with the GRS. The model shows that three of the booms fail at different times due to high velocity currents, but they do sufficiently block the oil from reaching further up the river. The fourth boom deployed does not interact with oil due to the presence of the other boom (Figure 3).

The Southeast Louisiana ACP focuses on prioritizing sensitive sites rather than developing specific strategies for these sites. The Area Committee in that Planning Area believes that it is impossible to plan for every possible spill scenario. They feel that the most important preplanning that can be done is to ensure that the ESAs are accurately prioritized. The GRS in this ACP specify the prioritized locations with three general types of response strategies: Collection Booming with On-Water Recovery, Exclusion Booming, and Deflection Booming. OILMAP was used to run several cases without boom, with booming according to the priorities and general response strategies set in the ACP, and with user-defined booming specifically chosen to minimize shoreline oiling based on the modeling results.

Figure 4 shows the area of Louisiana used for this test with the USCG Sector New Orleans GRP locations indicated by red dots. The smaller scale map in the inset shows the area of the test with a blue star.

The scenario used to conduct the Louisiana validation involved a spill of 65,000 barrels of South Louisiana Crude oil at the surface over a period of eight hours. For the first evaluation, Southeast Louisiana ACP GRS Test #1, no boom was deployed in order to investigate areas of highest impact based on the predominant environmental conditions. Without deploying any boom, 2,441 barrels of oil impacted the shoreline as shown in Figure 5.

For the next simulation, Southeast Louisiana ACP GRS Test #2 booming was added based on the high priority locations and using the chosen general response strategies for boom as defined in the ACP. In this scenario, 2,045 barrels of oil reached the shoreline. Implementing the general response strategies listed in the ACP resulted in a 16% reduction in shoreline oil from the scenario with no boom deployed.

For the final test, Southeast Louisiana ACP GRS Test #3, the results from Tests #1 and #2 were considered, and the user determined the booming strategies that would be most effective using a trial-and-error method. Multiple strategies were tested using different seasons of the year, during different times of the day. Although many iterations were run, the best result was chosen, minimizing the shoreline oiling. Using the OILMAP model, this process took less than an hour to execute. The result of Test #3 was 1,772 barrels of oil impacting the shoreline, a 28% reduction in impact from Test #1 with no boom deployed. A user could continue this process to further refine the strategy.

In the scenarios simulated in Louisiana, the boom did not fail due to excessive currents, waves, or wind. The boom placement and configuration were the leading cause of whether oil did or did not reach the shore. Updating the boom configurations in the model to maximize effectiveness allowed for an iterative approach to find the most effective booming strategy for this area.

The GRS in the Philadelphia ACP include similar detail to the GRS in the Plymouth to Salisbury, Massachusetts ACP. Specific equipment deployment strategies are included for the identified sensitive sites. For this scenario, 1,500 barrels of Arabian Light Crude oil was simulated at the entrance of the Delaware River. The oil released over a period of four hours at the water surface. All GRS included in the ACP for this area were implemented in the model.

Figure 8 shows the GRS test conducted in this region. Strong currents in the river cause the boom to fail at the Delaware River Nature Preserve and along the East side of Pea Patch Island. Using the computer simulation quickly shows that these strategies will not be successful due to the environmental conditions in the river. Conversely, the booming strategy selected for the mouth of the Salem River is successful as shown in Figure 9. No oil escapes from the boom during this strategy. Area Committee planners can be confident that this strategy can be deployed to protect the sensitive sites.

Using an oil spill trajectory and fate model can successfully assist Area Committee members with testing the GRS included in their ACP. By implementing an iterative approach, similar to that used in the Southeast Louisiana ACP GRS Test, users can deduce the best possible strategies for each ESA quickly while considering all times of the year and day.

Government agencies can save considerable funding typically used for field deployment exercises by using computer simulations as a first round to testing the GRS in the plan. For a typical test of one GRS, a USCG-approved OSRO will charge approximately $8,500. Considering each ACP includes hundreds of GRS, actual field deployment tests for each GRS would amass to millions of dollars. A model like OILMAP can be licensed for a much lower cost than conducting field deployments in an ACP Planning Area. Computer simulations can be used as a first level of validation of the GRS in order to improve oil spill response planning while at a significantly lower cost.