Hydrodynamic characterization of Corpus Christi Bay through modeling and observation.

Christi Bay is a relatively flat, shallow, wind-driven system with an average depth of 3-4 m and a mean tidal range of 0.3 m. It is completely mixed most of the time, and as a result, depth-averaged models have, historically, been applied for hydrodynamic characterization supporting regulatory decisions on Texas coastal management. The bay is highly stratified during transitory periods of the summer with low wind conditions. This has important implications on sediment transport, nutrient cycling, and water quality-related issues, including hypoxia which is a key water quality concern for the bay. Detailed hydrodynamic characterization of the bay during the summer months included analysis of simulation results of 2-D hydrodynamic model and high-frequency (HF) in situ observations. The HF radar system resolved surface currents, whereas an acoustic Doppler current profiler (ADCP) measured current at different depths of the water column. The developed model successfully captured water surface elevation variation at the mouth of the bay (i.e., onshore boundary of the Gulf of Mexico) and at times within the bay. However, large discrepancies exist between model-computed depth-averaged water currents and observed surface currents. These discrepancies suggested the presence of a vertical gradient in the current structure which was further substantiated by the observed bi-directional current movement within the water column. In addition, observed vertical density gradients proved that the water column was stratified. Under this condition, the bottom layer became hypoxic due to inadequate mixing with the aerated surface water. Understanding the disparities between observations and model predictions provides critical insights about hydrodynamics and physical processes controlling water quality.

Integrated real-time monitoring system to investigate the hypoxia in a shallow wind-driven bay.

Corpus Christi Bay (Texas, USA) is a shallow wind-driven bay which experiences hypoxia (dissolved oxygen < 2 mg/L) during the summer. Since this bay is a very dynamic system, the processes that control the hypoxia can last on the order of hours to days. Monitoring systems installed on a single type of platform cannot fully capture these processes at the spatial and temporal scales of interest. Therefore, we have integrated monitoring systems installed on three different platform types: (1) fixed robotic, (2) mobile, and (3) remote. On the fixed robotic platform, an automated profiler system vertically moves a suite of water quality measuring sensors within the water column for continuous measurements. An integrated data acquisition, communication and control system has been configured on our mobile platform (research vessel) for synchronized measurements of hydrodynamic and water quality parameters at greater spatial resolution. In addition, a high-frequency radar system has been installed on remote platforms to generate surface current maps for the bay. With our integrated system, we were able to capture evidence of a hypoxic event in summer 2007; moreover, we detected low dissolved oxygen conditions in a part of the bay with no previously reported history of hypoxia.

A mobile monitoring system to understand the processes controlling episodic events in Corpus Christi Bay.

Corpus Christi Bay (TX, USA) is a shallow wind-driven bay and thereby, can be characterized as a highly pulsed system. It cycles through various episodic events such as hypoxia, water column stratification, sediment resuspension, flooding, etc. Understanding of the processes that control these events requires an efficient observation system that can measure various hydrodynamic and water quality parameters at the multitude of spatial and temporal scales of interest. As part of our effort to implement an efficient observation system for Corpus Christi Bay, a mobile monitoring system was developed that can acquire and visualize data measured by various submersible sensors on an undulating tow-body deployed behind a research vessel. Along with this system, we have installed a downward-looking Acoustic Doppler Current Profiler to measure the vertical profile of water currents. Real-time display of each measured parameter intensity (measured value relative to a pre-set peak value) guides in selecting the transect route to capture the event of interest. In addition, large synchronized datasets measured by this system provide an opportunity to understand the processes that control various episodic events in the bay. To illustrate the capability of this system, datasets from two research cruises are presented in this paper that help to clarify processes inducing an inverse estuary condition at the mouth of the ship channel and hypoxia at the bottom of the bay. These measured datasets can also be used to drive numerical models to understand various environmental phenomena that control the water quality of the bay.

Application of fractal flocculation and vertical transport model to aquatic sol-sediment systems.

In estuarine and coastal environments, flocculation occurs between particles of different fractal dimensions and of different densities. Questions remain concerning the level of detail required to model particle flocculation and settling in these heterogeneous systems. This paper compares the goodness of fit between two flocculation models, using measured time series particle size distribution data collected from clay, colloidal silica, emulsified crude oil, clay-crude oil, and silica-crude oil systems. The coalesced sphere (CS) flocculation model includes the effects of heterogeneous particle size and density; the modified coalesced fractal sphere (mCFS) model adds effects due to heterogeneous fractal dimension. Goodness of fit was quantified using values of a minimized objective function, the mean of the sum of the square of the relative residuals (MSSRR). For nearly all tested experimental conditions, MSSRR values varied less than 5% between the CS and mCFS flocculation models. Additionally, collision efficiency values for single-particle-type (alpha(HOMOO)) and dual-particle-type (alpha(HETT)) systems were obtained through parameter regression using the CS and mCFS models. Using the mCFS model, estimated fractal dimension (D) values obtained for clay and clay-oil systems were between 2.6 and 3.0, lower than that postulated by the CS model but higher than that estimated experimentally by the particle concentration technique. The Stokes settling velocity of a clay aggregate of a given mass is reduced with decreased fractal dimension. This results in clay-oil flocculation occurring faster than floc sedimentation in the studied hydrodynamic range. Thus, the mCFS model provides insights to the fate of spilled oil in inland and coastal waters.

Redox dynamics during recovery of an oil-impacted estuarine wetland.

Redox potentials and sediment porewater parameters were measured around the periphery of a small cove along the San Jacinto River in Texas throughout a crude oil and chemical dispersant remediation study to distinguish normal dynamics from those caused as a response to stress from oil deposition and chemical treatment and subsequent recovery. Before the application of oil and treatments, sediments displayed average redox potentials of 0-350 mV when not submerged. Within 2 days of the applications, redox potentials in these plots decreased and exhibited a range from -200 to 0 mV for a duration of 5 weeks. Applied treatments significantly reduced the sediments of the wetland. Reduced redox potentials were indicative of the corresponding sulfate reduction that was also found to be significant following the oil application. GC/MS and MPN analysis indicates this reduction is due to biological oxidation of the crude oil components by alkane- and PAH-degraders in these surficial sediments and validates the usefulness of the redox measurement as an indicator for carbon oxidation. Increases in aqueous phase total organic and inorganic carbon coincided with a decrease in pH shortly after the applications, suggesting incomplete mineralization and the generation of organic acids. While dissolved ferrous iron and sulfide have been found to be good indicators of reductive processes in petroleum-contaminated aquifer sediments; that was not the case in this wetland study. Despite the disappearance of sulfate following the oil application, dissolved ferrous iron and sulfide concentrations remained at pre-application levels suggesting the formation of mackinawite and/or pyrite. The transient exposure of surface sediments to oxygen complicates the consideration of potential solid phase pathways since aqueous iron may be removed by precipitation when oxidized or reduced, making porewater iron a poor indicator for terminal electron accepting processes in wetland sediments.

Behavior of a chemically dispersed oil in a wetland environment.

An experiment was conducted at a wetland research facility, investigating the behavior and effects of chemically dispersed oil (CDO) using an oil-spill dispersant. The research site is located on the San Jacinto River near Houston, TX. The replicated treatments included oiled control, "high-dose" CDO (1:10 dispersant-to-oil ratio (DOR)), "low-dose" CDO (1:20 DOR), as well as an unoiled control. Known amounts of oil or dispersed oil were added to the respective plots. Sediment samples were taken over a 99-day period using a 5-cm-diameter coring device. The GC/MS results for both "total target saturate hydrocarbons" and "total target aromatic hydrocarbons" were plotted over time and data were modeled using nonlinear regression. The overall (including abiotic and biotic) petroleum loss rates for the dispersed-oil treatments were not statistically different when compared to the oiled control. However, the initial concentrations for the dispersed-oil treatments were statically lower (95% confidence) than for the oiled control. From this, it can be inferred that the dispersed oil was more prone to flush off the sediments, as was visually observed. Biodegradation rates were also determined for all treatments; it was concluded that there were no differences when comparing each dispersed-oil treatment to the oiled control. The sediments from each plot were also analyzed for microbial population numbers (most-probable-number) and acute toxicity (Microtox 100% Test). Statistical analyses for both sets of data found no significant differences for the dispersed-oil treatments when compared to the oiled control.

Modeling coagulation kinetics incorporating fractal theories: comparison with observed data.

There are currently four possible approaches in modeling coagulation kinetics: the traditional Euclidean rectilinear; the Euclidean curvilinear; the fractal rectilinear; and the fractal curvilinear. The fractal model includes the Euclidean case as a subset. The primary purpose of this research is to investigate which of the rectilinear models among these best predicts the evolution of experimental observed particle size distribution (PSD). Using a fractal rectilinear model previously developed by the authors, model predictions were compared with a series of observed PSD data obtained from estuarine sediment particles in a 2m settling column, where the average velocity gradient (G) was 20 or 40s(-1). Nonlinear parameter estimation was performed to estimate two free parameters for the fractal model (the fractal dimension, DF, and the collision efficiency factor, a), and one free parameter (the collision efficiency factor, alpha) for the Euclidean model. Compared with the observed PSD, the simulation showed that the fractal rectilinear model was best, and that this model fit better for the larger size particles. The estimated DF was between 2.6 and 3.0. The research demonstrated that the alpha's have multiple values for the same observed data, depending on the coagulation model used. This finding is significant because a is currently used as a single value based on the conventional Euclidean rectilinear model.

The use of toxicity bioassays to monitor the recovery of oiled wetland sediments.

Six toxicity assays were compared to determine their efficacy in assessing toxicity dynamics during a wetland bioremediation study. The toxicity bioassays used were the Microtox 100% elutriate test, Microtox Solid Phase Test (SPT), amphipod assay, P450 reporter gene system, Toxi-ChromoPad test and a Salmonella/microsome assay. Oiled sediments were analyzed for toxicity in the petroleum biostimulation experiment conducted along the San Jacinto River, near Houston (TX, USA). The bioassays were evaluated for their ability to measure acute toxicity, chronic toxicity, and the mutagenic potential of amended oiled plots as compared to oiled and unoiled control plots. Amendments were diammonium phosphate alone or in combination with potassium nitrate, which served as an alternate electron acceptor. With exception of the Toxi-ChromoPad and Salmonella tests, the bioassays exhibited a significant increase in toxicity after oil application. Microtox bioassays detected significant sediment toxicity up to 29 d after oil and amendment application. The Microtox solid phase test results correlated strongly with gas chromatography-mass spectrometry analyses of total target saturate and aromatic hydrocarbons. The amphipod assay detected initial toxicity with a decline to day 70, followed by a significant increase in toxicity on day 140 in plots receiving nutrient amendments, which may be in response to excessive nutrient application. Low levels of enzyme induction were observed with the P450 reporter gene system assay in all oiled sediments throughout the study, suggesting low but persistent levels of polycyclic aromatic hydrocarbons. Of the six tests, the two Microtox tests and the amphipod test showed the most potential in evaluating petroleum toxicity in wetland sediments.

Chemical dispersant effectiveness testing: influence of droplet coalescence.

Thermodynamic and kinetic investigations were performed to determine the influence of coalescence of chemically dispersed crude oil droplets in saline waters. For the range of pH (4-10) and salinity (10 per thousand, 30 per thousand, 50 per thousand ) values studied, zeta-potential values ranged from -3 to -10 mV. As the interaction potential values calculated using Derjaguin-Landau-Verway-Overbeek (DLVO) theory were negative, the electrostatic barrier did not produce significant resistance to droplet coalescence. Coalescence kinetics of premixed crude oil and chemical dispersant were determined within a range of mean shear rates (Gm = 5, 10, 15, 20 s(-1)) and salinity (10 per thousand, 30 per thousand ) values. Coalescence reaction rates were modeled using Smoluchowski reaction kinetics. Measured collision efficiency values (alpha = 0.25) suggest insignificant resistance to coalescence in shear systems. Experimentally determined dispersant efficiencies (alpha = 0.35) were 10-50% lower than that predicted using a non-interacting droplet model (alpha = 0.0). Unlike other protocols in which the crude oil and dispersant are not premixed, salinity effects were not significant in this protocol. This approach allowed the effects of dispersant-oil contact efficiency eta(contact) to be separated from those of water column transport efficiency (eta(transport)) and coalescence efficiency (eta(coalescence)).

Characterizing aquatic sediment-oil aggregates using in situ instruments.

The effects of emulsified crude oil and salinity (15, 30 per thousand ) on the steady state aggregate volume distributions and fractal dimensions were determined for a range of mean velocity gradients, (G(m) =5-50 s(-1)). Aggregation was performed in a 40-L cylindrical tank with a 4-blade paddle mixer. Three-dimensional fractal dimensions (D3) and volume distributions were determined using a procedure integrating data from an electrozone and an in situ light scattering instrument. Two-dimensional fractal dimensions (D2) and derived volume distributions were determined using a recently developed submersible flow cytometer equipped with a digital camera and image analysis software. For latex beads or emulsified crude oil systems, the above listed instruments yielded consistent size distributions and fractal dimensions (D2=1.92 +/- 0.16, D3=2.94 +/- 0.12). Mean volume aggregate diameters determined using the FlowCAM were consistently larger that those determined using the LISST-100 or Coulter Multisizer due to aggregate orientations during measurements. With increasing G(m) values, all colloidal aggregates showed increasing D3 values due to reduced aggregate length. Because of the compactness of all the aggregates (D3 >2), D2 values remained constant at 2. Neither salinity nor sediment type significantly affected D2 values calculated for sediment-crude oil aggregates. However, clay-oil aggregates showed higher D3 values than clay aggregates. This suggests that colloidal oil and mixing shear are the more dominant factors influencing aggregate morphology in nearshore waters. Overall, the data suggests that the analysis methods provide consistent size distribution results. However, because of the shear and salinity of coastal waters, resulting aggregates are too compact to estimate their D3 values using image analysis alone.

Evaluation of bioremediation strategies of a controlled oil release in a wetland.

A controlled petroleum release was conducted to evaluate bioremediation in a wetland near Houston, Texas. The 140-day study was conducted using a randomized, complete block design to test three treatments with six replicates per treatment. The three treatment strategies were inorganic nutrients, inorganic nutrients with an alternative electron acceptor, and a no-action oiled control. Samples were analyzed for petroleum chemistry and inorganic nutrients. These results are discussed in the context of our related research involving toxicology and microbiology at the site during the experiment. To evaluate biodegradation, the targeted compounds were normalized to the conservative compound C3017alpha, 21beta-[H]hopane, thus reducing the effects of spatial heterogeneity and physical transport. The two biostimulation treatments demonstrated statistically-higher rates of biodegradation than the oiled no-action control. For the majority of the experiment, target nutrient levels were maintained. Further research may be warranted to optimize these bioremediation strategies as well as evaluating additional treatment strategies for wetlands and other shoreline systems.

Intrinsic bioremediation of a petroleum-impacted wetland.

Following the 1994 San Jacinto River flood and oil spill in southeast Texas, a petroleum-contaminated wetland was reserved for a long-term research program to evaluate bioremediation as a viable spill response tool. The first phase of this program, presented in this paper, evaluated the intrinsic biodegradation of petroleum in the contaminated wetland. Sediment samples from six test plots were collected 11 times over an 11-month period to assess the temporal and spatial petroleum concentrations. Petroleum concentrations were evaluated using gas chromatography-mass spectrometer analyses of specific target compounds normalized to the conservative biological marker, C(30)17alpha,21beta(H)-hopane. The analyses of specific target compounds were able to characterize that significant petroleum biodegradation had occurred at the site over the one-year period. Total resolved saturate and total resolved aromatic hydrocarbon data indicated the petroleum was degraded more than 95%. In addition, first-order biodegradation rate constants were calculated for the hopane-normalized target compounds and supported expected biodegradation patterns. The rapid degradation rates of the petroleum hydrocarbons are attributed to conditions favorable to biodegradation. Elevated nutrient levels from the flood deposition and the unconsolidated nature of the freshly deposited sediment possibly provided a nutrient rich, oxic environment. Additionally, it is suggested that an active and capable microbial community was present due to prior exposure to petroleum. These factors provided an environment conducive for the rapid bioremediation of the petroleum in the contaminated wetland.

Estimating sub-surface dispersed oil concentration using acoustic backscatter response.

The recent Deepwater Horizon disaster resulted in a dispersed oil plume at an approximate depth of 1000 m. Several methods were used to characterize this plume with respect to concentration and spatial extent including surface supported sampling and autonomous underwater vehicles with in situ instrument payloads. Additionally, echo sounders were used to track the plume location, demonstrating the potential for remote detection using acoustic backscatter (ABS). This study evaluated use of an Acoustic Doppler Current Profiler (ADCP) to quantitatively detect oil-droplet suspensions from the ABS response in a controlled laboratory setting. Results from this study showed log-linear ABS responses to oil-droplet volume concentration. However, the inability to reproduce ABS response factors suggests the difficultly in developing meaningful calibration factors for quantitative field analysis. Evaluation of theoretical ABS intensity derived from the particle size distribution provided insight regarding method sensitivity in the presence of interfering ambient particles.

Modeling crude oil droplet-sediment aggregation in nearshore waters.

This paper describes a modeling approach that simulates changes in particle size distribution and density due to aggregation by extending the Smoluchowski aggregation kinetic model to particles of different density. Batch flocculation studies were conducted for clay, colloidal silica, crude oil, clay-crude oil, and silica-crude oil systems. A parameter estimation algorithm was used to estimate homogeneous collision efficiencies (alphaHOMO) for single-particle-type systems and heterogeneous collision efficiencies (alphaHET) for two-particle-type systems. Homogeneous collision efficiency values (alphaHOMO) were greater for clay (0.7) and for crude oil (0.3) than for silica (0.01). Thus, clay and crude oil were classified as cohesive particles while silica was classified as noncohesive. Heterogeneous collision efficiencies were similar for oil-clay (0.4) and oil-silica (0.3) systems. Thus, crude oil increases the aggregation of noncohesive particles. Data from the calibrated aggregation model were used to estimate apparent first-order flocculation rates (K') for oil, clay, and silica and apparent second-order flocculation rates (K'') for oil and clay in oil-clay systems and for oil and silica in oil-silica systems. For oil or clay systems, aggregation Damköhler numbers ranged from 0.1 to 1.0, suggesting that droplet coalescence and clay aggregation can occur on the same time scales as oil resurfacing and clay settling, respectively. For mixed oil-clay systems, the relative time scales of clay settling and clay-oil aggregation were also within an order of magnitude. Thus, oil-clay aggregation should be considered when modeling crude oil transport in nearshore waters.

Partitioning of crude oil polycyclic aromatic hydrocarbons in aquatic systems.

This paper investigates the hypothesis that observed polycyclic aromatic hydrocarbon (PAH) concentrations in an aqueous system are equal to the sum of the organic phase and soluble phase molar concentrations. While the organic phase concentrations are proportional to the PAH mole fraction in the oil, the soluble phase molar concentrations are estimated using Raoult's law. A batch laboratory mixing vessel with a scalable mixing energy was loaded initially at various oil layer thicknesses (0.4-3.2 mm) which correspond to oil surface loadings (40-310 mg/cm2). The vessel was agitated at constant mean shear rates (Gm = 5, 20 s(-1)). Total petroleum hydrocarbon (TPH) samples were taken periodically to estimate the entrainment rate as a function of initial oil layer thickness. TPH concentrations were measured in-situ using a laser scattering instrument (LISST-100) and ex-situ using gravimetric analysis. At a steady-state TPH concentration (>72 h), additional samples were analyzed for PAH concentration using GC/MS analysis. TPH concentrations increased over time according to a first-order kinetic model. Generally, the first-order rate constant and steady-state concentration both increased with increased oil loading and with increased Gm. In addition, measured PAH concentrations correlated well (r2 > 0.96) with those predicted by a partitioning model. These results are useful for assessing the effects of mixing and oil loading conditions on crude oil entrainment and PAH partitioning.