PAH depletion in weathered oil slicks estimated from modeled age-at-sea during the Deepwater Horizon oil spill

Highlights

• Evaluated relationship between age-at-sea and PAH depletion.

• PAH depletion increased with estimated age-at-sea.

• PAH vapor pressure decreased with increasing alkylation.

• A new module to predict PAH concentration distributions is recommended.

• Improved oil spill response through better estimates of health risks.

Abstract

During time-periods oil slicks are in the marine environment (age-at-sea), weathering causes significant changes in composition and mass loss (depletion) of oil spill chemicals including the more toxic polycyclic aromatic hydrocarbons (PAHs). The goal of this study was to estimate the age-at-sea of weathered oil slicks using the oil spill module of the Connectivity Modeling System and to use this age to interpret PAH concentration measurements. Percent depletion (PD) for each measurement was computed as the percentage difference between the original and measured PAH concentration in the crude oil and weathered oil slicks, normalized upon the mass losses relative to hopane. Mean PD increased with estimated age-at-sea for all PAHs. Less PD was observed for alkylated than for parent PAHs, likely due to decreasing vapor pressure with increasing degree of alkylation. We conclude that estimated age-at-sea can be used to explain PAH depletion in weathered oil slicks. We propose PAH vapor pressure can be coupled with the model to expand capacity for predicting concentration distributions of individual parent and alkylated PAHs in weathered oil along the coastline. This new module will advance the science supporting oil spill response by providing more certain estimates of health risks from oil spills.

Introduction

The Deepwater Horizon (DWH) disaster resulted in an offshore oil spill of 4.9 million barrels of oil causing one of the largest environmental disasters and response efforts in U.S. history (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling (National Commission), 2011, Beyer et al., 2016, Westerholm et al., 2021). The deep-sea oil blowout, lasting 87 days from April 20 to July 15 of 2010, had negative effects on the health of coastal communities in the Gulf of Mexico (GoM) (Sandifer et al., 2021), and multiple species of pelagic, tidal and estuarine organisms were exposed to weathered oil (Barron et al., 2012). Some of the more buoyant oil released from BP’s Macondo well’s broken riser pipe traveled up the water column from 1522 m deep and formed large surface oil slicks. At its maximum extent, Macondo oil (known as MC-252 oil) covered over 40,000 square kilometers of the ocean (Anon, 2016), and generated atmospheric pollutants that reached the Gulf coast (Montas et al., 2022). Transport of oil to the coast by winds and tides resulted in oil impact to sediments in the nearshore environment in the form of floating, submerged, and beached oil. Approximately 2113 kilometers of shoreline were oiled, of which over 900 kilometers were sandy beaches (Nixon et al., 2016), within Louisiana, Mississippi, Alabama, Florida and Texas (Berenshtein et al., 2020, Weisberg et al., 2017, Le Hénaff et al., 2012). Elevated concentrations of polycyclic aromatic hydrocarbons (PAHs), among the more toxic oil spill chemicals, were detected in various nearshore compartments, including weathered oil, oil impacted sediments, ocean water and air (Sammarco et al., 2013, Boehm et al., 2016, Xia et al., 2020, Xia et al., 2021, Montas et al., 2022).

As part of oil spill response, oil transport models were used during the disaster to accurately predict the spatiotemporal dynamics of surface oil (Liu et al., 2011a, Liu et al., 2011b, Le Hénaff et al., 2012, French-McCay et al., 2017, Ainsworth et al., 2021). This information was used during response operations to deploy different oil containment mechanisms. Hindcast simulations of oiled coastlines were successfully conducted using various models including the General NOAA (National Oceanic and Atmospheric Administration) Operating Modeling Environment (GNOME) and the oil module (Paris et al., 2012, Perlin et al., 2020) of the Connectivity Modelling System (CMS) (Paris et al., 2013). A Lagrangian photo-oxidation module was developed for the CMS to quantify the likelihood of photo-oxidative changes of surface oil and to track the transport of persistent photo-oxidized compounds (Vaz et al., 2021). For planning purposes, the GNOME model was also used to conduct post disaster simulations of oil transport scenarios based on historical winds and currents (Barker et al., 2011). Ainsworth et al. (2018), used a spatially explicit biogeochemical end-to-end ecosystem model, Atlantis, to simulate impacts from the DWH and subsequent recovery of fish guilds. Finally, a model developed by the Consortium for Simulation of Oil-Microbial Interactions in the Ocean (CSOMIO) (Dukhovskoy et al., 2021) used various computational schemes, including a two-way Lagrangian-Eulerian mapping technique also used in Perlin et al. (2020), to simulate oil microbial degradation and sedimentation.

Collectively, the model simulations mentioned prior show that during the disaster, information regarding the spatiotemporal dynamics of surface oil was available prior to oil impact and available post-disaster for future incident planning purposes (Wilson et al., 2021). However, at the time of the disaster limited information was available on the concentrations of individual toxic oil spill chemicals prior to oil impact to the nearshore and subsequent coastline oiling, whether stranded or beached oil, due to the lack of real-time direct measurements or lack of appropriate model predictions capable of simulating individual toxic chemicals. Likewise, despite the availability of models predicting spatiotemporal data for oil, relationships between depletion of individual toxic oil spill chemicals (OSCs) have rarely been examined as a function of the oil’s transport time following release from the leak point. These relationships are of particular importance because weathering gradually increases toxic components in the residual oil (Yim et al., 2011) and human and ecological health risks are dependent upon these specific chemical compounds (Chakravarty Chowdhury and Deka, 2022, Zhang et al., 2022). Estimating the individual chemical concentrations for the more toxic OSCs, such as PAHs, is a critical consideration in oil spill first response measures (Ventikos and Psaraftis, 2004, Farrington, 2020, Westerholm et al., 2021), and implementation of the necessary protective measures (Michaels and Howard, 2012). Importantly, having the capacity to estimate concentration distributions of toxic OSCs in human exposure zones would improve oil spill response by allowing for estimations of health and environmental risks to communities, and therefore improve the management of oil spills and weathered oils (Ainsworth et al., 2021, Solo-Gabriele et al., 2021).

The goal of this study was to estimate the age-at-sea of weathered oil slicks and to determine the relationship between PAH depletion and oil slick weathering at sea for individual chemicals. We used spatiotemporal data analysis to collocate measured PAH concentrations from weathered oil slick samples collected along the Gulf Coast, at the time of the DWH oil spill, and the DWH simulation output of the 3D oil spill module of the Connectivity Modeling System (oil-CMS, Paris et al., 2012, Paris et al., 2013). The oil-CMS output data from Berenshtein et al. (2020) was used to compute the age-at-sea of the weathered oil slicks. Percent depletion (PD) for each sample was computed as the percentage difference between each PAH’s original concentration in the crude oil and its measured concentration from sampling data, normalized upon the mass losses relative to hopane, a compound that serves as the conservative internal marker within the oil. Results were evaluated according to each chemical’s physicochemical characteristics.

To our knowledge this is the first study that uses data from an oil fate and transport model to estimate the age-at-sea of environmental samples. Moreover, here we focus on a group of representative parent and alkylated PAHs and provide an analysis for how depletion is influenced by time in conjunction with each chemical’s individual physiochemical properties. The aspect of model estimated time differentiates this study from others that focused on "degree of weathering" as a method of comparing low to high depletionl These two considerations, age-at-sea (time) and individual physicochemical properties constitute essential steps required to model weathered oil PAH concentrations in real-time which would greatly benefit oil spill response operations. Currently, a few oil spill fate and transport models have the capacity to predict concentrations of a limited number of PAHs in water. Here, for the first time we hypothesize that modeling can be expanded to predict the concentrations for a wide range of parent and alkylated PAHs.