Sediments, as this paper will explore them, are defined as free particles of soil found at the bottom of a water body, whether it be clay, sand, organic material, or silt. They originate from both erosion and decomposition of natural elements, animals, and plants. When the sediments are exposed to pollutants through water or soil, they become contaminated over time.
The locations of contaminated sediments are very widespread. The most concerning areas that are affected include the Atlantic and Pacific coasts, the Gulf of Mexico, the Great Lakes Region, and many inland waterways (EPA, 1999). Research in the area of contaminated underwater sedimentation has slowly developed since the 1970’s when environmental policy was first becoming relevant.
The most prominent contaminants are called Polycyclic Aromatic Hydrocarbons (abbreviated PAH’s), which include petroleum and byproducts of petroleum. Other contaminants, like dichlorodiphenyltrichloroethane (DDT) and Polychlorinated biphenyl (PCB) are concerning because their decay half-life is extremely long, which leads them to stay at high levels of contamination for long periods of time (Ferrarese, Andreottola, Oprea, 2007).
Contamination of sediments is mostly dangerous because of how it travels up the food chain, ultimately becoming harmful to animals and humans. When contamination starts in the soils that are home to plant seeds, the toxins travel through the soil and make the plants themselves toxic. As organisms eat these plants and larger animals eat these organisms, the toxins become more potent in the process of biomagnification (EPA 1999).
By the time the contaminants reach the top of the food chain, mutations or deadly diseases can result. If small fish closer to the bottom of the food chain and digesting the free, toxic sediments cannot handle the mutations, their population will die out quickly. The biodiversity of the ecosystem will decrease, causing environmental problems that could become detrimental.
The main source of problems concerning human health effects comes with the consumption of fish. Starting in the 1990’s, fish consumption advisories were put out in the Great Lakes area due to health concerns. Potential risks for humans include neurological disorders and cancer.
The 1970’s began an influx of environmental policies that began addressing problems that the public was not aware of. The Clean Water Act, Clean Air Act, Comprehensive Environmental Response, Compensation, and Liability Act, and Water Development Act all became legislature within a couple of decades. The Water Resources Development Act of 1992 (WRDA) specifically addressed the contaminated sediments issue for the first time. It caused Congress to order the Environmental Protection Agency to write the National Sediment Quality Survey, which found there to be contaminated sediments in bodies of water in every state, while most notably outlining the severe problems in the Great Lakes region (EPA, 1992).
The sources of these types of contaminants vary, but can mainly be traced back to human waste disposal sites. Point sources, such as sewage treatment plants, sanitary sewers, stormwater drains, and waste industry discharge sites can be traced to specific areas. Contamination sources that are more widespread, like stormwater runoff, mining and manufacturing runoff, and atmospheric pollutant spread, are known as non-point sources, as they are more difficult to trace back to origins. Since these sources range from public to privately owned, there is not one sector of government that is left responsible for regulating sediment contamination.
Before the 1980s, the contamination level of sediments was determined by comparing the concentration of a chemical in sampled sediments to “background” or reference values. It was recognized that this approach does not account for the types of biological resources in an aquatic environment or the concentration at which an adverse response would be observed in these organisms. To improve on this method, sediment quality guidelines (SQGs) have since been developed for use in assessing sediment quality, meaning contaminant concentrations that cause adverse effects (SETAC, 2002).
Numerous SQGs have been established since the 1980s, each incorporating different criteria, factors and approaches to try and account for the varied conditions in which sediment contamination occurs. Generally, these approaches can be divided into two main categories (Burton, Jr., 2002):
Empirical SQGs are generally used for heavy metals and arsenic, but may incorporate other contaminats, like organics, although the available data is more limited. EqP SQGs are mainly used for organic compounds; however, we will also present an EqP example for heavy metals. Both types predict adverse ecological effects from sediment contamination by the response of benthic organisms. Benthic organisms live on or in the sediments of aquatic systems and are used as an indicator of a toxic environment because of their function as an important food chain link and food source for fish, birds, and mammals residing in the same ecosystem (Batiquidos Lagoon Foundation).
SQGs generally outline two concentration thresholds, one in which a toxic response is unlikely and one for which a toxic response is likely. The predictive capabilities of these guidelines leave great uncertainty in the “grey” region of contaminant concentrations that lie between the thresholds. For this case, it may be necessary to perform a site-specific analysis by observing the health and behavior of benthic organisms at the site. It is also important to understand that SQGs tend to only incorporate one type of contaminant which may lead to inaccurate conclusions such as adverse effects being attributed to one type of contaminant when multiple are present or toxic response limit being underestimated for a mixture of contaminants.
Table 1 outlines the most current and popular SQGs in use in practice today. Additional information on each guideline may be found in papers by the developers. For the purpose of this paper, we will only look at two examples for each of theoretical and empirical SQGs as an introduction to their usage.
Di Toro, Mahony et al. (1991)
Di Toro, Zarba et al. (1991)
Ankley et al. (1996)
Di Toro and McGrath (2000)
Persaud et al. (1993)
Von Stackelberg and Menzie (2002)
Effects Range-Low (ERL) and Effects Range-Median (ERM)
Long et al. (1995)
Threshold-Effects Level (TEL) and Probable-Effects Level (PEL)
MacDonald et al. (1996)
Smith et al. (1996)
Apparent-Effects Threshold (AET)
Barrick et al. (1988)
Ginn and Pastorok (1992)
Cubbage et al. (1997)
MacDonald, DiPinto et al. (2000)
MacDonald, Ingersoll et al. (2000)
Logistic Regression Modeling (LRM)
Field et al. (1999, 2002)
Table 1. Types of SQGs in use today and their developers.
Empirically-based SQGs are based on field and laboratory data of benthic organism responses to exposure to metal-contaminated sediments. These relationships are determined through compilation of data from many different sites with different types and concentrations of contaminants.
The following subsections provide examples on the formulation and use of two emprical SQGs approaches.
The TEL is a sediment contamination concentration at which a toxic response has started to be observed in benthic organisms. The PEL is the concentration at which a large percentage of the benthic population shows a toxic response. Using engineering judgement, the consequences of sediment contamination and need and types of applications for remediation can be inferred.
The state of Florida developed equations to determine TEL and PEL based on the concentrations at which benthic organisms from coastal waters showed toxic responses in the lab. Which are shown in Equations 1 and 2, respectively.
It is assumed that sediment contamination concentrations below the TEL are acceptable and concentration above the PEL are unacceptable. The grey region in between the TEL and PEL requires further study and judgement to determine the likelihood of environmental consequences.
There have been many developments of these types of effects levels by different agencies for different species of heavy metals. Burton references the available limits, as of 2001, for concentrations developed which are similar to the TEL and PEL, although with different names depending on the developer. These are shown in Tables 1 and 2 as a reference to the available SQGs and magnitude of the concentrations that induce toxic responses.
Table 1. Concentrations analogous to the TEL developed by different agencies, listed down the rows, for different types of heavy metals, listed as the columns, in milligrams/kilogram (Burton, 2001).
Table 2. Concentrations analogous to the PEL developed by different agencies, listed down the rows, for different types of heavy metals, listed as the columns, in milligrams/kilogram (Burton, 2001).
Tables 3 and 4 show threshold effects and probable effects concentrations for organic compounds, also arranged by Burton.
Table 3. Concentrations analogous to the TEL developed by different agencies, listed as the columns, for different types of organic compounds, listed down the rows, in micrograms/kilogram (Burton, 2001).
Table 4. Concentrations analogous to the PEL developed by different agencies, listed as the columns, for different types of organic compounds, listed down the rows, in micrograms/kilogram (Burton, 2001).
The Wisconsin Department of Natural Resource (WI DNR) uses the consensus-based SQG approach for sediment quality assessment to try and limit the uncertainty associated with concentrations that lie between the threshold-effects and probable-effects concentrations.
This method includes a Threshold Effect Concentration (TEC), the contaminant concentration below which a toxic response is not expected and formulated using a TEL-type approach, a Midpoint Effect Concentration (MEC), and a Probable Effect Concentration (PEC), the concentration above which a toxic response is expected and formulated using a PEL-type approach,. Table 2 shows the approaches recommended by the WI DNR in determing the TEC and PEC. It is also possible to formulate a site-specific TEC and PEC based on laboratory test data for the contaminants and benthos present at a site. Next, TEC and PEC are selected based on allowable risk and level of conservatism deemed appropriate for the project, as either the average result from a suite of SQGs in Table 5 or concentrations at which chosen percentiles of toxic responses are observed in the lab.
Table 5. SQGs used by the WI DNR to develop the TEC and PEC.
Once a TEC and PEC have been established, they can be used to develop the MEC by taking the average of the TEC and PEC. Now, contaminant concentrations obtained from field tests can be compared to the three thresholds to indicate the level of concern the contaminants concentrations merit, 1, 2, 3, or 4, with Level 1 being the least concerning and Level 4 the most. This methodology and a numerical example for arsenic are shown in Table 6.
Table 6. An example of the use of the TEC and PEC by the WI DNR.
Theoretically formulated SQGs employ equilibrium partitioning (EqP) to describe the bioavailabilty of contaminants from sediments to benthic organisms. Using concentrations in sediments along with material properties, the amount of contaminants from sediments that will enter the water and thus benthic population can be estimated. It is generally assumed for these methods that the entire mass of the contaminant in solution is available for benthos uptake. The certainty in this method relies on the precision of measurement of contaminant concentrations and material constants as well as an accurate description of the system to be evaluated.
To demonstrate the first application of EqP SQGs, we will consider a simple EqP procedure utilized by the Army Corps of Engineers in the investigation of a confined disposal facility in the Saginaw River for dredged sediments from a Federal navigation channel (Myers, 1991). The dredged sediments were being investigated for concentrations of PCBs.
Assumptions that have been made include equilibrium conditions, no sorption of the contaminant to sediment particles and no contamination in the pore space. All of the contaminant either exists in solution in the water or in the solid phase. Note that while this example is meant to be instructive of the theory behind EqP, more recent and complex formulations have been proposed as stated in the previous section.
A material property of organic contaminants, the partitioning constant, Kp, is a measure of the solubility of the contaminant in water and can be used to predict the concentrations of a mass of that contaminant that will be present in the aqueous phase. Kp has been measured for most chemicals and can be found in reference tables or calculated as shown in Equation 1.
In the most general equation (Equation 2), assuming equilibrium of the system, Kp is equal to the ratio of the concentration of the contaminant in the aqueous phase to the concentration in the solid phase.
Cwe = (Cse / Kp) (Eqn. 2)
While this equation is quite simple and only assumes equilibrium, both Kp and the current concentration of the contaminant in the solid phase must be assumed or measured, although average Kp values are readily available in tables for most contaminant.
Equations 3 and 4 can also be used in more stipulative scenarios, and equilibrium assumptions still apply.
Equation 3 is only applicable for closed systems which is often an oversimplification for aquatic systems. However, all values besides Kp can be easily known. In addition, this formula provides the initial value for the contaminant concentration in the sediments.
Cwe = (Cs0 / Kp) (Eqn. 4)
Equation 4 is a simplification of Equation 2 that can be made if one assumes large value for both the sediment concentration and partitioning constant.
Additional complexities could be added to this system by incorporating the uptake rate of the contaminant from the water to the benthic organisms or the rate of decay of a contaminant.
The aqueous solution formulation may be modified to capture the sorption of the contaminant to sediment particles. Conditions are still at equilibrium, but this example assumes that the contaminant exists in the aqeous phase, sorbed to particles, and in the solid phase. This describes a three-phase contaminant system which is a more accurate description for sediment contamination.
For Adsorbtion, another material property, Kads, will need to be measured or found in a reference table for the contaminant. In most cases, it is best to measure this value in a lab because it will change based on the type of contaminants and sediments.
If we assume that the contaminant concentration is small and the surface area of available sediments will not be saturated, we can use Equation 5.
As we can see, the mass of the sorbed contaminant is dependent on an accurate description of Kads and the concentration in the aqueous phase. Additionally, the mass of sediments may be hard to estimate depending on the size and mobility of the system.
This example does not account for a phenomenon known as sorbtion hysteresis in which contaminants in the pore space of the sediments replace other sorbed contaminant molecules and vice versa, a process called sorption hysteresis.
The use of SQGs in assessing the problems and risk associated with contaminated sediments at a site is an iterative process. SQGs require minimal inputs in the way of site investigation data, however uncertainty is high when the guidelines are used without calibration. Early stage predictions can be made on the spatial distributions of contamination with concentration level testing at multiple locations at a site. They can also predict the concentration trend in time through contaminant decay. This allows the SQG user to predict locations that are especially troublesome or one that do not require additional attention. A breakdown of recommended uses for SQGs by SETAC is shown in Table 7.
Table 7. Recommended roles for SQGs by SETAC.
After the preliminary investigation and data collection, SQGs can also be indicative of areas that will require additional gathering of information such as measured concentrations that fall between the TEC and PEC. As more data is analyzed, and the results of SQGs calculated, it will become evident if additional lines of evidence or sources should be explored. An example flow chart of this process is shown in Figure 1.
If SQGs are deemed an appropriate method and indicator of conditions at a site, they can be calibrated using site-specific data to make the best predictions possible. Examples of this would be laboratory tests for toxic response concentrations for particular contaminants or mixtures and benthos at the site or a more complex EqP model for the ecosystem at the site (NYSDEC).
Figure 1. The iterative process for SQGs recommended by NYSDEC.
A phenomenon called bioaccumulation is known to be possible in a sediment contamination scenario. Bioaccumulation is the increase in the concentration of a contaminant as it moves up the food chain. Starting with small benthic organisms, as larger animals consume these organisms, the concentration present in larger organisms’ systems scales with the number of contaminated benthos consumed. Although still debated, contaminants currently known to biomagnify include PCBs, mercury, and DDT. For humans, who are essentially at the top of the food chain, the description and prediction of biomagnification is especially interesting pertaining to public health and food source concerns.
Most SQGs, up to this point, do not account for bioaccumulation. However, EqP SQGs are best-positioned to describe this process. The New York State Department of Environmental Conservation’s (NYSDEC) standardized practice for the use of SQGs includes a procedure for estimating the effect of biomagnification in the ecosystem for contaminated sediments.
SQGs are often described as a useful, predictive tool in the assessment of sediment quality and contamination. While many organizations believe this to be true, it is important to consider the limitations of SQGs for use in practice.
A.) Empirical SQGs predict thresholds at which a toxic response of benthic organisms is probable. However, SQGs, in general, do not provide perspective on the risk and possible negative outcomes associated with toxic concentration levels which requires a separate risk assessment.
B.) The factors used in the development of SQGs, especially older ones, is not well-documented and limited guidance is available on which SQGs are applicable for criteria (MacDonald, 2000). For this reason the use of SQGs is somewhat controversial in practice, and are not standardized across environmental agencies.
C.) Research and development of new SQGs is expensive and their future practicality and performance are uncertain.
D.) EqP SQG predict concentrations based on equilibrium conditions. In addition, available relationships do not model release of contaminants trapped in pores of sediments where they can be sorbed on interstitial walls.
E.) As with any empirical model, it is important to be aware of the types and rates of test errors present in each empirical SQG.
Even though SQGs may leave something to be desired, they are a useful tool that can provide a first guess at the nature of a sediment contamination problem. Combined with engineering judgement, research on the SQG being used, and appropriate field and laboratory sampling and testing, SQGs are an important tool in practice for sediment contamination, remediation, and risk assessments.
In conclusion, Table 8 provides a matrix of advantages and disadvantages associated with the practical use of both types of SQGs discussed herein, empirical and theoretical (EqP). This list is not exhaustive, but provides a summary of the advantages and disadvantages explored in this paper in regards to SQGs.
Theoretical (EqP) SQGs
| || |
| || |
Andreottola, Giannai, Ferrarese, Elisa, Oprea, Irina Aura. (2007). “Redmediation of PAH-Contaminated Sediments by Chemical Oxidation”. Department of Civil and Environmental Engineering, University of Trento, Italy
Bureau of Habitat. (2014). “Screening and Assessment of Contaminated Sediment”. New York State Department of Environmental Conservation Division of Fish, Wildlife, and Marine Resources
Burton, G. Allen, Jr. (2002). “Sediment Quality in Use Around the World.” Institute for Environmental Quality, Wright State University.
Contaminated Sediment Standing Team. (2003). “Consensus-Based Sediment Quality Guidelines: Recommendations for Use & Application”. Wisconsin Department of Natural Resources.
Florida Department of Environmental Protection (2008). “Derivation of Numerical Sediment Quality Guidelines for Florida Coastal Waters Using the Weight of Evidence Approach”. Florida Sediment Quality Assesments Guidelines (SQAGs), Volume 1, Chapter 5.
Ingersoll, Christopher G., Wenning, Richard J. (2002). “Use of Sediment Quality Guidelines and Related Tools for the Assessment of Contaminated Sediments: Executive Summary of a SETAC Pellston Workshop”. Society of Environmental Toxicology and Chemistry.
Myers, Tommy E. (1991). "Polychlorinated Biphenyl Levels in the Saginaw Confined Disposal Facility During Disposal Operations, Fall 1987," Miscellaneous Paper EL-91-4, US Army Engineer Waterways Experiment Station, Vicksburg, MS.
Office of Science and Technology. (1999). “Introduction to Contaminated Sediments”. United States Environmental Protection Agency.