AquaGate®+ For Contaminated Sediments

AquaGate+ delivers powdered amendments in a thin coating around an inner core, allowing movement of water through the matrix for treatment, while limiting the migration of contaminants. Sites where AquaBlok® and AquaGate+ materials have been applied have been monitored and studied extensively to confirm long-term performance. These materials improve remediation outcomes and are shown to reduce the cost of delivering amendment materials in aquatic settings.

AquaGate materials can be utilized in:

• In-Situ Treatment

• Active/Reactive Capping

• Permeable Reactive Barriers (PRB)

• Shoreline Sheen Mitigation

• Emergency Response

AquaGate+ Amendments

*AquaGate+ products are not limited to the use of the above examples. Other amendments may be available to address your project. Please contact us to discuss further options.

Key Benefits and Technical Advantages

See Links to Below Sections:

1. Introduction to the Use of Activated Carbon for Contaminated Sediment Remediation
   

2. Influence of Activated Carbon Particle Size on Cap Performance: Implications for Modeling
   

3. In-Situ Treatment: What is it, and what results have been demonstrated?
   

4. Importance of Uniform delivery through water column; minimal material loss
   

5. Installation support; in-house lab and technical support

Introduction To The Use Of Activated Carbon For Contaminated Sediment Remediation

Activated carbon continues to experience increased regulatory acceptance and implementation for a wide range of contaminated sediment sites throughout the U.S.

AquaGate+PAC was developed in 2010 under a DoD ESTCP project (ER-0825: In Situ Wetland Restoration Demonstration) and continued with an addition ESTCP project in 2012 (Puget Sound Naval Shipyard Pier 7 Bremerton, Washington) under ER-201131 Demonstration of In Situ Treatment with Reactive Amendments for Contaminated Sediments in Active DoD Harbors. This project was awarded the DoD’s Project of the Year in 2016.

The first full-scale application of activated carbon was in 2016 at Lake Onondaga which was designed as a mixture of Granular Activated Carbon (GAC) and sand at a very low percentage of GAC by weight. At that time, it was recognized that, in order for GAC to be placed through a water column, it must be saturated in advance and mixed hydraulically with sand. The contractor designed a large spreader system to enable the mixture to be pumped from the shoreline and placed as a slurry/mixture at the water surface.

Since this first full-scale application of a GAC/Sand mixture, other contractors have developed equipment to achieve a similar result, but many applications of GAC/Sand mixtures have also been applied with minimal or no saturation of the GAC and using placement equipment that is not specifically designed to support hydraulic delivery of the mixture from shore.

More importantly, since the Lake Onondaga application, which was performed in 2016, there has been a relative lack of any monitoring of GAC/Sand remedial designs in the industry. Since 2016, there have been over 25 sediment remediation projects using AquaGate in the U.S. and 5 additional projects in Norway. High profile sites employing activated carbon include: The Passaic River (RM10.9), Pearl Harbor (U.S. Navy) and the Scanlon Reservoir (St. Louis River, MN). More are currently in the planning stages.

“The growth and acceptance of activated carbon for sediment remediation has led to increased interest and research on the effects of particle size and the practical aspects of placement. See the next section for more information on this topic.”

AquaGate was developed to help overcome many of the challenges of placing both powdered and granular versions of adsorptive and reactive treatment materials. The following sections highlight some of the key considerations and provide important technical information that should be considered when evaluating the various approaches to adding amendments to sediments.

Contaminated Sediment Remediation: Influence of Activated Carbon (AC) Particle Size on Cap Performance; Implications for Modeling & Remedy Effectiveness

I. How did we get here? — Performance of Different Particle Sizes of AC

The application of AC for vessel-based removal of aqueous contaminants dates back to work done by the EPA to address drinking water (Dobbs and Cohen, 1980). In these applications, granular activated carbon (GAC) was used due to ease of replacement and control of residence time for adsorption. As a result of this success, early investigation for the use of AC for sediment remediation focused on GAC even though testing and applications of powder activated carbon (PAC) was being performed and generally resulted in lower contaminant concentrations than an equal amount of GAC (due to the higher surface area-to-volume ratio of PAC) (cite Upal work). This was initially driven by the difficulty of handling and placing PAC in the water, resulting in a preference for GAC. During early applications of GAC, there was recognition of the potential for improved performance of PAC, but, there was a perception that the long-term design duration of sediment remedies (i.e., 50-100 years) would ultimately allow time for GAC to provide equal performance at field scale. Unfortunately, this thinking did not consider the potential for short-term failure due to rapid upwelling. In addition, the relative lack of post placement monitoring data for GAC applications make it difficult to evaluate or confirm either long-term performance or short-term failure. At this time, there is still not good data available to compare the full-scale performance of GAC vs. PAC.

In an attempt to provide insight into the long-term performance of AC, AquaBlok has participated in a number of comparative studies of PAC vs GAC. Recently, a long duration adsorption study of PCBs by PAC and GAC was performed and the data from the adsorption test was then utilized to develop a representative CapSim model that demonstrates the implications of particle size on the testing protocols, modeling, and design of AC sediment remedies. These results are presented herein.

II. Why does this matter? — Testing Approach and Results

The long-term performance of AC is a critical component for the success of AC-based sediment remedies. For this reason, the test duration for adsorption was set to 50 weeks. The duration was determined to be long enough to demonstrate the usefulness of an AC-based sediment approach in a cap since longer durations would equate to essentially a diffusion-based system. This study’s period of 50 weeks is the first known attempt to quantitatively evaluate and compare long term adsorption of PAC and GAC. Two (2) sources of AC were tested in PAC and GAC forms, resulting in four (4) AC samples. The adsorption experiment for 11 PCB congeners (PCB-10, 25, 70, 101, 118, 138, 149, 170, 180, 194) covering low to high MW hydrophobic compounds and in the presence of TOC (10 mg/L NOM) was conducted by Danny Reible at Texas Tech. PCB concentrations were set to be similar to those at real sites.

The test results provided key data that can be used to evaluate performance of PAC and GAC under various circumstances. The key findings are summarized below:

• At the end of 50 weeks, PAC removed 210x more contaminant than GAC and was 1300x faster in adsorption, on average. The adsorption of AC is heavily influenced by particle size, even for the same base AC material – PAC removes more contaminant and removes it faster.

• When groundwater velocity is high, GAC performance is greatly reduced, and breakthrough of contaminants is far more likely. The kinetic differences (i.e., speed of adsorption) between PAC and GAC will result in different real-world capacity in environments where contaminant transport is highly groundwater driven. Because GAC is slower to adsorb, it requires a longer residence time to equal PAC adsorption.

The below figure illustrates the impact of the particle size on the amount of contaminant that can be adsorbed by PAC and GAC in a specific time frame.

The effect of changing residence time on a 15 cm thick cap as groundwater upwelling velocity changes is illustrated in terms of the available capacity (amount of contaminant adsorbed) of PAC and GAC. When the groundwater velocity is low, the residence time is high, and the available capacity of PAC and GAC will be at its maximum – although PAC will still have higher capacity as seen at every data point.

For example, at a fixed cap thickness of 15 cm, a groundwater velocity of 0.1 cm/d will provide a residence time of 150 days for this cap. At this residence time, PAC has a KD of 1010 L/kg while GAC has a KD of 107.5 L/kg. When the groundwater velocity increases by 10, the residence time will reduce to 15 days, reducing the GAC capacity to 106.5 L/kg while the PAC retains its Kd at 1010 L/kg. As groundwater velocity continues to go up, the residence time of the cap and the GAC capacity will go down in turn. Because the key is residence time, the same effect will be experienced for a fixed groundwater velocity if the cap thickness is reduced. A thinner cap will result in reduced residence time, which will also reduce the amount of contaminant GAC can remove in this time frame.

III. What Does the CAPSIM Model Say? - Implications for Remedy Performance

To evaluate potential implications, the results of the GAC adsorption tests were integrated into a CapSim model. The model was set up to be representative of actual caps that have been specified for full-scale projects. The cap model evaluated consisted of a 15cm thick chemical isolation layer consisting of GAC and sand overlain with a 10 cm habitat layer (sand). Contaminant concentration was 0.5 µg/L of PCB10. Groundwater velocity was set at 1 cm/d, representing a cap residence time of 15 days. Breakthrough was defined as 1% of the starting contaminant concentration (0.005 µg/L) at the surface of the chemical isolation layer (i.e. below the habitat layer).

Two potential placement outcomes were modeled. The first assumed a theoretically perfect/uniform distribution of the GAC amendment within the chemical isolation layer. This was performed to illustrate what the CAPSIM model typically assumes. A second approach was developed that was intended to evaluate the potential impacts of nonuniformity in GAC distribution within the GAC/Sand layer during placement. It was assumed that the GAC/sand isolation layer would be placed in three lifts, with GAC separating and layering on each lift. Results of these modeling scenarios are summarized below and model output is shown in the graphics.

• The unmodified base case scenario (theoretically uniform placement while ignoring the speed of adsorption) suggested that a cap constructed with 1% GAC would be protective for 100 years.

• When GAC kinetics and nonuniformity are added (i.e. constructing the cap in three lifts) to the model, the 1% GAC is no longer protective and will experience breakthrough in just over 3 years.

• Using the same nonuniformity scenario but doubling the GAC amount to 2% extends the cap life to 14.4 years – but still well below the unmodified base case. Further, it is necessary to increase the GAC content to 5% in order to return the cap life to 100 years.

• In comparison, a thin layer of AquaGate+PAC (delivering a lower dose of AC than the 1% GAC scenario) is protective for 100 years.

The CapSim modeling exercise demonstrated that particle size and placement implications are critical when upwelling velocity may limit residence time within the cap or when realistic field installation challenges may impact the uniformity of the amendment within the capping layer. Some of the key conclusions of the modeling exercise are as follows:

• The significant performance difference of PAC vs GAC suggests that each of these AC materials should be viewed as separate approaches during the testing, modeling, design, and specification of the AC-based sediment remedy. Also, it is critical that testing data on PAC materials is not utilized for specifications that call for a GAC/Sand approach.

• There is an increased ‘risk-of-remedy’ for GAC-Sand caps due to the nonuniform distribution that can happen during placement of GAC-Sand mixtures. The density difference between GAC and Sand will result in separation during placement through the water column. As a result, the residence time of the contaminant through GAC is effectively reduced, creating a significant potential impact on the expected performance of the remedy.

• A PAC-based approach using AquaGate+PAC will result in a reduction in material quantity and/or layer thickness. This opens the application of AC-based remedies to difficult-to-access areas (like under piers) or in challenging and sensitive environments.

• PAC-based approaches provide protectiveness at a significant reduction to material quantity, overall cost, and remedy risk.

IV – Where do we go from here? - Recommendations

The use of AC is an important part of improving remedy outcomes and remedy design for contaminated sediment sites. To improve outcomes, the treatability testing approach and modeling methodology must be carefully examined and planned to insure potentially significant issues impacting the expected performance of the remedy do not occur.

The data collected in a 50-wk adsorption test provide definitive and conclusive insight on the effective adsorption equilibrium of PAC and GAC. The faster kinetics and higher effective capacity of PAC over GAC can result in much lower material quantity and overall remedy cost for PAC based remedies. As a result of these significant differences, the following are important recommendations to consider in remedy design utilizing AC:

1. Treatability testing of AC should use the commercially available particle size of the material anticipated for application. Testing results generated for one particle size should not be attributed to another particle size.

2. It is important to understand the site-specific characteristics such as pore water velocity and ensure that data exists that can assist with appropriate inputs into modeling.

3. Modeling AC should be based upon commercially available materials as it has been demonstrated that particle size has a significant impact on real-world performance of AC-based sediment remedy.

4. Placement/installation methods should be carefully considered in combination with the type of AC utilized in the design since models assume theoretically perfect distribution of the AC to achieve the modeled results.

References

Dobbs and Cohen, 1980. Carbon Adsorption Isotherms for Toxic Organics, United States Environmental Protection Agency. EPA-600/8-80-023.

Upal Ghosh

In-Situ Treatment Approach:
Treatment of Contaminated Pore Water Through a Permeable Amendment Layer

Goal: Reduce Pore Water Concentration of Target Contaminant in the Biologically Active Zone (BAZ) of Sediment with a Low-Impact to Existing Benthic Habitat.

In-situ treatment involves the placement of a thin layer of activated carbon rich amendment directly over the existing sedimenrt surface. The activated carbon becomes mixed with the underlying sediment via bioturbation, or the natural xixing processes performed within the bioactive zone (BAZ).

• A thin layer of treatment material is applied directly to sediment surface - no modification or removal.

• Treatment material slowly mixes with sediment in the baz through natual bioturbation concentrations of contamination in the baz pore water are reduced.

Extensive data and documentation of full-scale application of In-Situ Treatment has been developed and published. Examples include our ESTCP project in Bremerton, WA and a large-scale project at Middle River, MD – Under Recent Documents, see 2021 In Situ Treatment Three Year Monitoring Report.

In summary, pore water concentrations of contamination (i.e. PCBs) have been documented to be reduced by as much as 95% following application of powdered activated carbon in the form of AquaGate+PAC. In addition, these reductions have been sustained in excess of five years following application.

Uniform Distribution of Treatment Material:
A Key Principal & Assumption

When utilizing amendments in a ‘reactive’ or treatment layer within a sediment capping design, it is of critical importance to deliver the adsorptive or reactive materials in the most uniform manner – both in vertical and horizontal directions within the capping layer. The reasoning for this is simple, the contaminants moving from the sediment up through the capping layer must come in contact with these treatment materials in order to remove them from the pore water. Therefore, if the treatment material is not distributed as uniformly as possible there is a much greater likelihood that pollutants will ‘short-circuit’ or simply not be removed as they pass through the capping material. The graphic below is an attempt to illustrate the theory behind the concept of uniform distribution.

Unfortunately, in practice, it is far more difficult to achieve this theoretically ‘perfect’ distribution of treatment materials. This is particularly true when the quantity of adsorptive material is very small, compared to the bulk of the capping layer. For example, if a specification calls for less than 5% GAC by weight within a 15cm thick sand capping layer, the small amount of GAC is going to be nearly impossible to distribute uniformly, in both vertical and horizontal directions, within the cap. In actual installations, contractors have attempted to overcome the challenge of material separation in the water by placing multiple thin lifts, to achieve a target total thickness.

As shown in the photo (L), this often results in visible horizontal layers of GAC within the capping layer, which we refer to as ‘zebra striping’. Note: The photo was taken from a pan collected during an actual GAC/Sand placement. In contrast to the above, AquaGate+PAC can be placed without mixing or, if needed, will mix effectively with sand, and provide more uniform distribution of the PAC component within the treatment layer.

This is illustrated by the below photograph, which is a pan collected during actual full-scale installation of a reactive cap on the Detroit River. In this design, the AquaGate+PAC was mixed at an approximate 50:50 ratio with sand to achieve a 12-inch thick treatment layer. Another important factor in comparison of a GAC/Sand approach to the use of AquaGate+PAC relates to the potential for losses of activated carbon in the water column during placement. In regard to GAC/Sand, there is one significant confounding factor when considering material losses during placement. This relates to the relative difficulty of accurately assessing the quantity of GAC within a GAC/Sand layer (post-placement). Thermal methods (i.e. EPA Lloyd Kahn Method) often provide wide variations in results, rendering the methods of little use. For this reason, most project specifications do not require that a minimum GAC content be verified. Therefore, to gain confidence that a minimum quantity of GAC makes it into the layer, engineers will often simply double the quantity of GAC. In cases where efforts have been made to accurately quantify the amount of GAC that resides in a GAC/Sand cap, losses of from 20-50% have been documented. It should be noted that AnchorQEA has developed an alternative method for determination of GAC content, referred to as “Density Separation Recovery of Granular Activated Carbon” which uses a heavy liquid for separation of GAC and Sand in a solution. Of course, it is important to note that, even if you can accurately quantify the total GAC content, and therefore reasonably determine losses during placement, there is still the issue of non-uniform distribution, which is not addressed in this regard.

So, if we assume the correct or design quantity of GAC is successfully placed, this begs the question: What is the Impact of GAC nonuniform distribution / layering on Cap Life and Risk-of-Remedy? For an evaluation of this question, please see the section titled: Influence of Activated Carbon Particle Size on Cap Performance – Implications for Modeling.

Installation support:
In-house lab and technical support.

As noted in the AquaGate background section, AquaBlok has been manufacturing and supporting installations of AquaGate for over 10 years (since our first ESTCP project in 2012). Since that time, we have gained significant experience and furthered the development of handling, storage, installation and quality control (QA/QC) for AquaGate products. With regard to both activated carbon and organoclay use for ‘active’ or treatment caps or in-situ treatment, it is believed that the AquaGate approach has been utilized more than any other material or placement method. Based on this experience, the following sections are presented to highlight some of the important field and technical support offered.

INSTALLATION SUPPORT

Three phases of field support offered include; Material Handling/Storage, Placement Equipment and Systems, and On-site QA/QC.

MATERIAL HANDLING/STORAGE

AquaGate is an aggregate particle which is coated with a powdered adsorptive or reactive material such as powder activated carbon (PAC) or organoclay. We tell engineers to think “peanut M&M” for a visual for size and coating thickness. As such, the coating can be damaged by too many material handling steps or use of certain methods, which are generally accepted for handling of sand or raw aggregate. In our detailed material proposals, it is generally recommended that handling steps we minimized (to 3-4), prior to placement in the water. Similarly, storage is a critical aspect of field operations since contractors generally do not want placement operations slowed due to delays in shipments to a site. For larger projects, placement is also often faster than production of AquaGate, thereby creating the need for advanced shipments and storage of material at the site, so placement can be continuous for the entire duration of the installation. Other than adequate allocation of space, the primary challenge for storage is that AquaGate must remain dry prior to placement. AquaBlok generally recommends that AquaGate in bulk bags are not stacked more than two high and all material is covered (i.e. either with tarps or in a temporary structure) to protect the product from weather/moisture. Although AquaBlok routinely provides phone support at no charge, on many occasions, AquaBlok personnel have been engaged to provide on-site technical support and recommendations for both material handling and storage of AquaGate.

PLACEMENT EQUIPMENT AND SYSTEMS

As noted above, AquaGate is a coated particle that is designed to be placed from the surface of the water – to effectively ‘deliver’ adsorptive or treatment materials to the sediment surface. As such, the material needs to be kept dry prior to placement. This means that AquaGate cannot be transported to placement equipment or placed hydraulically. This is a significant difference between placement of AquaGate vs GAC/Sand (or other granular sand mixtures) which is generally pumped out to placement equipment. While this limitation generally only impacts how the material is delivered to the placement equipment, this can be a issue where distances are significant or there are limited areas where capping material can be brought to a site. However, once AquaGate is delivered to the placement equipment, then a wide array of placement methods are possible and have been successfully employed. Examples of successful placement methods include (but are not limited to); excavator on barge, stone slinger, TeleBelt®, hopper/spreader on barge, and pneumatic spreader. See Section 6 – of the ITRC guidance on Sediment Cap Isolation Guidance” Chemical Isolation Construction Considerations for more information on equipment types and advantages of each (https://sd-1.itrcweb.org/ ).

ON-SITE QA/QC

A final phase of technical support offered by AquaBlok involves quality control and quality assurance of materials supplied and also materials placed. AquaBlok believes it is of critical importance to recognize quality control and quality assurance relate to not only the manufactured and delivered products, but also to their ability to achieve the design goals of the project. A major objective of any project QAAP should be the post-placement verification that the cap achieved the design criteria. As noted in other sections, this includes not only that the total quantity of adsorptive or reactive material was delivered to the sediment surface, but also that the placement achieved uniform distribution of the reactive material within the cap. AquaBlok has provided technical support to engineers at numerous sites to assist in evaluation of material, post-placement. At some sites, AquaBlok’s extensive manufacturing QA/QC procedures were relied upon to document the delivery, with only post-placement monitoring being performed. However, we believe it is important to verify post-placement (as-constructed cap) objectives and we stand ready to support engineers or other parties to accomplish this goal. Please feel free to contact us to learn more about specific testing and methods that can be employed to verify AquaGate materials, either post-manufacturing or post-placement.

Addressing PFAS Contamination 

Learn more about our AquaGate+RemBind Approach for in-situ remediation of soil, groundwater, and sediment.