Remediation of Contaminated Stormwater Canal at Miami International Airport

M. A. Vivona, P.E. and Gregg Mooney, P.E.

undefinedOn the basis of previous studies, metals and organics contamination of sediments in a stormwater collection/conveyance canal at Miami International Airport had been confirmed. Contamination levels detected in the canal sediments exceeded Dade County and Florida state standards. A Remedial Action Plan (RAP) had been prepared and approved, requiring the removal and disposal of contaminated sediment (and associated contaminated water) as a source control measure. Canal sediments were described in the RAP and prior study documents as "very fine grey to black silts with sand and natural organic (plant) matter." Thickness was reported as varying from 1.5 to 2 feet. A preliminary concept of the proposed approach to remediation was incorporated in the approved RAP. Metcalf & Eddy, Inc. (M&E) was contracted to perform site remediation.

System Description

M&E expanded the concept presented in the approved RAP, and prepared detailed drawings and specifications for the remediation system. The remediation system design provided for dredging, dewatering of removed sediment, off-site disposal of dewatered sediment and treatment and disposal of the water generated through sediment dredging and dewatering.

The use of a truck-mounted hydraulic dredge was incorporated to pump the sediment (and associated water) from the canal. This approach was selected over barge-mounted units based on the long, narrow configuration of the canal footprint, and the ease of access along the edges of the canal. The detailed system design consisted of a slurry receiving tank (with shale shaker screen), slurry batch tanks, and a belt filter press with polymer addition for sediment dewatering. The RAP concept for filtrate treatment included oil/water separation and granular activated carbon. Due to suspended particulates and dissolved metals in filtrate from dewatered sediment, filtrate settling tanks, sand filters and an ion exchange unit to the treatment train were all added.

Since the transmissivity and yield of the underlying limestone strata are extremely high, and the canal is connected through underwater culverts to the rest of the canal system, drawdown of the water level in the canal or area groundwater was not of concern at the proposed dredge pumping rate. Thus, treated dredge water and sediment were not returned to the canal, but instead reinjected through an existing nearby infiltration gallery.

The following is a brief description of the operation. (Refer to Figure 1 for the process schematic.)

1. A truck-mounted hydraulic dredge (P-01) (Lefco's Sludge Buster) "vacuums" up a sediment/water mixture (approximately 20 percent solids) at a rate of approximately 150 GPM.
2. The slurry (sediment/water mixture) is piped on-shore to a slurry receiving tank (T-01) where, after initial screening for separation of debris, it is uniformly mixed prior to dewatering.
3. Slurry is then pumped to two 12,600 gallon slurry batch tanks (T-02 and T-03).
4. The mixed slurry is piped from the slurry batch tank to a belt filter press unit (BF-01) that dewaters the solids. Polymer is added to the slurry immediately prior to the belt filter press to improve slurry dewatering characteristics.
5. Dewatered solids are conveyed to a lined and bermed holding area, covered for temporary storage prior to analytical characterization and off-site disposal.
6. Water extracted from the slurry at the belt filter press passes through two filtrate settling tanks (T-06 and T-07, in series) to allow settling of particles. The water is then piped to two oil/water separators (T-08 and T-09, in parallel) to remove floating oils that may have been released from the sediment.
7. Water exiting the oil/water separator is piped through two sand filtration units (V-01 and V-02, in parallel) for removal of suspended particles, followed by two granular activated carbon (GAC) canisters (V-03 and V-04, in series) for final removal of trace organic compounds.
8. Treated water is piped from the GAC units to an ion exchange unit (V-05) for final polishing The final treated effluent is piped to the existing infiltration gallery for reinjection (field tests and calculations were performed to ensure that the existing infiltration gallery had sufficient capacity to receive the additional flow).

Field Process and Equipment Modifications

Once the field operation had begun, actual implementation of the remediation process was modified in two primary areas: physical configuration of the dredge unit and the approach to sediment dewatering.

Dredge Configuration

Canal sediment had been described in the RAP as loose sandy silts varying in thickness from 1.5 to 2 feet. However, during sediment removal operations, M&E actually encountered much heavier, more consolidated sediment that resisted removal by the dredge. M&E subsequently collected further sediment samples for lithologic analysis, and identified three semi-distinct layers of sediment.

The base of the canal was found to be a bed of a densely packed limestone fragments and indurated limestone bedrock. Although the loose silts in the upper layer of sediment were easily captured using the original hydraulic suction dredge, in-situ cohesiveness and cementation of the fine-grained sand and clay-sized calcite particles in the middle and bottom layers prevented the hydraulic dredge from recovering sediment at acceptable depths and production rates. To overcome this condition, M&E procured an auger attachment for the dredge suction hose to slice through and re-suspend the densely-settled middle and bottom layers of sediment, allowing subsequent capture.

Sediment Dewatering

Based on the sediment characterization presented in the RAP and supporting documents as well as bench-scale sediment dewatering tests, a belt filter press had been sized and provided to dewater the sediment. The belt filter press could not effectively retain and dewater the fine-grained sands and clay-sized calcite particles encountered during the performance of dredging activities. These particles flowed through the press and into the filtrate treatment system in sufficient quantities to force repeated shut-downs of the system due to clogging in the filtrate settling tanks.

For these reasons, the dewatering operation was modified to achieve a more effective separation of solids. The belt filter press also was removed. As a substitute, flocculent was added directly to the slurry mixing batch tank along with sediment extracted from the canal, and the resulting slurry was pumped to a roll-off dumpster to permit settling of the heavier sands and silts. Water and suspended particulates were then pumped from the roll-off dumpster into a meandering decanter maze constructed from concrete barricades and hay bales that permitted the more gradual settling of fines. Dewatered sediments were recovered from the roll-off dumpster and meandering decanter maze, and stockpiled for chemical characterization prior to disposal. Water was decanted from the maze into the original filtrate settling tanks. From the filtrate settling tanks onward, the original filtrate treatment system layout remained in operation. The modified process remained in use for the duration of the project, offering improved liquid/solids efficiency and higher production rates. Filtrate quality was not affected by the process modifications.

Environmental Safeguards and Data Reporting

Significant emphasis was placed on environmental safeguards both within the canal and at the sediment and equipment staging areas. Results of water quality monitoring of treatment system effluent were reported to regulatory agencies throughout the project. In addition to effluent monitoring, environmental safeguards for the project included the placement of impervious liners and berms in equipment and sediment staging areas; the use of turbidity curtains and oil booms in the canal during dredging; assessment of soil and groundwater in the equipment and sediment staging areas before and after implementation; and canal ecosystem evaluation.

Turbidity Control, Oil Booms and Staging Area Liners

Although no discharges were made back to the canal, controls were implemented to prevent potential turbidity increases in the canal caused by the dredging process. During the dredging operation, sediment was suspended by the hydraulic dredge. While most of this resuspended sediment was collected by the dredge, turbidity screens (silt curtains) were used to minimize the movement of suspended particles. Turbidity screens were placed on either side of the dredge and repositioned to bracket the dredge at all times as it was advanced along the length of the canal. Oil booms were used to bracket the dredge in basically the same manner to prevent the movement of potential floating petroleum hydrocarbons.

Staging area soil and groundwater were characterized prior to field mobilization to establish background conditions. HDPE liners (30-Mil) and berms were used in the equipment and sediment staging areas to prevent the release of free liquids to the underlying soil. Due to the numerous storm events that occurred during the project and the inevitable contact of stormwater with dewatered sediment stockpiles, the capture and treatment of stormwater from the staging areas was performed throughout the project. Once the meandering decanter maze had been substituted for the belt filter press, the sturdy liner and berms in the sediment staging area proved indispensable as a sump for the collection of canal water separated from sediment in the decanter maze.

In-Process and Effluent Monitoring

Effluent and in-process sampling and analysis included regular effluent sample collection and analysis required by the effluent discharge permits and regular sample collection and analysis from additional sampling points in the remediation process to verify the operational requirements and performance of each unit process.

Sampling and discharge of treated stormwater, sediment filtrate and decontamination water was performed on a "batch hold-and-release" basis. Treatment system effluent was containerized and discharged only after receipt of effluent data confirming compliance with applicable standards. This practice of testing prior to discharge was practiced through project completion.

In-process monitoring included the collection of samples at the following sample port locations for analytical parameters as noted: before and after the oil/water separators; before and after the sand filtration units for Total Suspended Solids (TSS) analysis; before and after GAC canister 1 for Total Organic Carbon (TOC) analysis; and before and after the ion exchange unit for Total Dissolved Solids (TDS) and heavy metals analysis. These analyses allowed evaluation of the performance of each unit operation.

Remediation Completion Criteria and Verification

Remediation completion criteria, verification and site closure procedures are addressed below.

Completion Criteria

The dredged canal section was approximately 1,250 feet long, 30 feet wide and 5 feet deep. Based on an estimated average sediment depth of 1.5 feet, a total sediment volume of approximately 2,100 cubic yards was projected.

While the theoretical objective of canal dredging was to extract all sediment down to the limestone comprising the original base of the canal, it was possible in practice only to capture the majority of deposited sediment. The canal water was not sufficiently clear to allow field personnel to see the canal bottom except at the extreme edges of the canal. Therefore, confirmation of dredging effectiveness was done by manually probing with poles to measure sediment depth and by collecting samples of remaining sediment using hand augers and "clamshell" sediment sampling devices. These methods indicated that the small percentage of material remaining consisted primarily of limestone fragments and medium-grained sands. These fractions were of relatively large particle size, found to contain very low levels of contamination.

Disposal of Waste Materials and Residuals

When the project was completed, all waste materials were disposed of including dewatered sediment, canal debris prior to dredging, debris separated from the sediment during the dewatering process, sand media from the sand filtration units, GAC media from the GAC filtration units, ion exchange media from the ion exchange units, liquids disposed subsequent to the disassembly of the water treatment system and HDPE liner used for the sediment and equipment staging areas.

Disposal of Dewatered Sediment

A total of 1,250 tons of dewatered sediment was chemically characterized and hauled offsite for thermal treatment in the Miami area. This weight corresponds to an estimated volume (dewatered basis) of 900 cubic yards and includes mechanically dewatered sediment from the belt filter press, settled sediment from the grit settling tank, meandering decanter maze and water treatment settling tanks, debris separated from the sediment slurry prior to the mixing tank and sand from the sand filtration units.

Disposal of Other Materials

Materials other than dewatered sediment were disposed of as follows.

• Debris removed from the canal prior to dredging was pressure-washed, placed in a roll-off dumpster, and hauled away for landfill disposal (approximately 2 cubic yards total).
• Debris separated from the sediment prior to the slurry mixing and dewatering process was disposed of with the dewatered sediment.
• Sand media from the sand filtration units were disposed of with the dewatered sediment.
• GAC media from the GAC filtration units was returned to the carbon supplier (Calgon) for regeneration.
• Ion exchange media from the ion exchange units were returned to the ion exchange resin supplier (Geopure) for regeneration.
• Liquids disposed of subsequent to the disassembly of the water treatment system were hauled away for off-site disposal (20,800 gallons total).
• Sludges removed from the filtrate settling tanks and effluent holding tanks at the conclusion of the project were hauled away for off-site disposal (413 gallons total).
• The HDPE liner used for the sediment and equipment staging areas was pressure-washed, placed in roll-off dumpsters, and hauled away for landfill disposal (approximately 30 cubic yards total).

Decontamination and Demobilization

Upon the completion of dredging, dewatering and treatment operations, each piece of equipment that came in contact with canal sediment or water from dewatered sediment was thoroughly cleaned using pressure-washing procedures prior to removal from the site. All equipment decontamination procedures were performed within the lined and bermed staging areas to allow the capture of wash water. Wash water was collected within the berms and pumped into the water treatment process (settling, oil/water separator, sand filtration, GAC filtration and ion exchange filtration) for treatment. In the final states of demobilization (after the treatment system had been disassembled and adequate treatment could no longer be provided) remaining liquids were collected by a licensed waste hauler for off-site disposal.

Post-Implementation Soil and Groundwater Characterization

Once all material and equipment had been cleaned and hauled away, site characterization for soil and groundwater was performed in a manner identical to that employed in the background characterization prior to initial mobilization to the site. The post-implementation characterization results confirmed that the environmental safeguards implemented during the performance of work at the site had been sufficient to prevent impacts to the underlying soil and groundwater.

Post-Implementation Canal Ecosystem Evaluation

A qualified biologist was employed to conduct an evaluation of the canal ecosystem subsequent to the completion of all dredging operations.

Summary and Conclusions

In summary, a total of 1,250 tons of dewatered sediment was removed from the canal south of Hangar 22 at Miami International Airport, and hauled to an offsite facility in the Miami area for disposal by thermal treatment. Sediment removal and dewatering procedures, and procedures for treating and discharging the canal water removed from the sediment during the dewatering process, were performed in general accordance with the approved RAP and subsequent approved modifications.

Major conclusions from this successfully completed project include

1. The truck-mounted hydraulic dredge, modified by the addition of a cutter head, proved effective in removing loose sediment and the majority of dense and/or cementitious material down to the underlying limestone rock.
2. Belt filter press dewatering in this application was not highly effective due to the poor quality of the filtrate, and the severely limited production rates. Alternate methods were effectively employed achieving improved solids capture and higher production rates.
3. Solids removal alone was not sufficient to achieve acceptable treated filtrate quality to meet required standards for reinjection. GAC and ion exchange were required to consistently meet or exceed water quality requirements.

Additional general conclusions were as follows.

1. Special care must be taken during the characterization and remedial planning phases to gather sufficient representative physical property data to:

• allow proper selection of remedial methods,
• effectively estimate remedial construction schedules and costs, and
• establish realistic and reasonable remedial goals.

1. Constructability and operability reviews should be performed by experienced field remediation professionals to incorporate practical experience early in the remedial process.
2. Flexibility must be provided in remedial technology selections, remedial designs, and the regulatory submittal/review/approval process to allow field adjustments to the inherent variability of site conditions.

About the Authors:

M. A. Vivona, P.E., was a Process Consultant with Metcalf & Eddy, Inc. ­p; Houston, Tx. during the time of this project. He currently is the Manager of Water/Wastewater for ICF Keiser, Inc./Georgia Wilson, Inc. in Houston, Tx.

Gregg Mooney, P.E., is Vice President of Metcalf & Eddy, Inc., Atlanta, Ga.