Formatted Title:

Rapid Field Verification of Organoclay Sand and Granular Activated Carbon Sand Capping Blends

Background/Objectives:

At the Gowanus Canal Superfund site (Brooklyn, New York), a multilayer reactive capping system is being installed as part of the approved remedy. The base reactive layer is an 8-inch thick organoclay (OC)-sand layer containing a minimum of 25 dry weight percent (wt%) OC to block upwelling of mobile non-aqueous phase liquid (NAPL). Situated immediately above the OC-sand layer are various granular activated carbon (GAC)-sand blends ranging from 6 to 50 wt% GAC and up to 36 inches thick depending on capping zone type to treat dissolved phase hydrocarbons. The current construction quality assurance (CQA) method for verifying the OC/GAC contents of the respective blends in upland stockpiles prior to placement in the canal consists of loss on ignition (LOI) testing performed at an off-site laboratory with a turnaround time of three days. To mitigate construction delays, rapid on-site testing methods to verify the OC and GAC contends of the respective blends were sought.

Approach/Activities:

After a technology screening process, portable x-ray fluorescence (pXRF) was selected based on ease of use, minimal sample preparation, reagent/chemical-free methods and rapidity. At the outset, pXRF did not seem viable since both OC and GAC contain appreciable amounts of carbon and pXRF is not capable of detecting carbon, therefore a key challenge was determining whether OC and GAC contained other “pathfinder” elements that could be consistently detected and accurately quantified. For different reasons, a key challenge for both media became the materials processing (comminution) of the pure and blended materials as well as their homogenization to ensure reproducible results. For the GAC-sand blends, a correlation was developed on the NYDOT sand specified for the Gowanus Canal project. Given concerns that a variety of quarried “sands” of different mineralogical origin may pose changes to developing a correlation, the subsequent analysis of the OC-sand blends was expanded to include a: 1) playground sand (PGS) purchased from a large retailer, a beach sand from the surf zone at Cape May Point (CMP) in New Jersey; and limestone (LIM) quarry sand from the Chicagoland area. Flourosorb-400 (FS) was also blended with the PGS given the recent interest in PFAS remediation.

Results/Lessons Learned:

Standard operating procedures were developed for the GAC, OC/FS, sand and their blends that included water rinsing, drying and milling to fines. A hand-held hammermill was successful for the GAC-sand blends but led to compaction and clumping of OC/FS. An angle grinder “Cobracrusher” attachment easily pulverized OC and blends in minutes. The pXRF analysis technique was largely the same for all media and the probe was operated in the mining mode to capture the broadest range of elements and elevated concentrations. In most cases, the pure materials and reagent blends at 5 wt% increments up to 50 wt% were measured with five to six replicates each for each blend target to account for variability. The strongest correlations for GAC were linked to sulfur (S), silica (Si) and calcium (Ca), with sulfur being linked to the source material (bituminous coal) for the GAC production. Goodness of fit (R²) values were greater than 0.95 for S and Ca for the GAC-GC sand blends. For the OC/FS-sand blends, the main pathfinder element was chlorine (Cl), owing to the reagent manufacturing process, but strong correlations were also found for Ca, iron (Fe), potassium (K), Si and titanium (Ti). Chlorine correlations had R²>≥0.98 for all sand types analyzed, illustrating the wide applicability of the technique for sediments projects nationwide. For all reagent blends, blend verification was possible in less than 4 hours per blend including oven drying.

Alise Lampi

Affiliation

Refine Results

Rapid Field Verification of Organoclay Sand and Granular Activated Carbon Sand Capping Blends