PHYSICAL SEPARATION PROCESSES

ERDC TN-DOER-C13

July 2000

should be conducted only if the physical data indicate feasibility. The physical data may also lead

to strategies for compositing samples for the contaminant distribution (fractionation) testing.

Selection of analytes for chemical analysis is a key component of the feasibility evaluation, and will

significantly impact the overall cost of the site characterization. Ultimately, this is a regulatory

issue. For preliminary site characterization, some cost savings can be achieved by limiting the

number of analytes where possible. This may be done by doing a full suite of metals and organic

compounds on a few bulk samples likely to represent the worst-case contamination. Compounds

not detected in the bulk analysis can be eliminated from most of the fractionation testing, although

a full suite should be done on all fractions of some samples. A 250-ml (8-oz) sample of dry or

high-solids sediment is sufficient to conduct the following analyses, including standard quality

assurance/quality control (QA/QC): Michigan Metals (arsenic, barium, cadmium, chromium,

copper, iron, lead, mercury, nickel, selenium, silver, and zinc), and organic compound groups

including base neutral/acid extractable (semivolatiles, BNA), polyaromatic hydrocarbons (PAH),

polychlorinated biphenyls (PCB), and total recoverable petroleum hydrocarbons (TRPH). Addi-

tional material is required initially to produce this volume in each resulting fraction. Necessary

starting sample volumes can be estimated from the bulk grain size distribution. If the volume of

any fraction obtained following separation is insufficient, fractions may have to be composited, or

limited analysis conducted, to obtain as much information as possible from those samples. Multiple

samples may be composited only if this is representative of the manner in which the material will

be blended for processing.

Existing characterization methods include the Fingerprint method and the TDG method. The

Fingerprint method was developed by Heidemij Realisatie, now ARCADIS Realisatie (Olin et al.

1999), and is a pilot scale operation using hydrocyclones to produce six size fractions, followed by

gravity separation of the density fractions. The TDG method is a bench scale characterization

procedure developed by ARCADIS Realisatie and TNO Institute of Environmental and Energy

Technology (Olin et al. 1999), producing three size fractions and three density subfractions. Test

development for contaminant distribution determinations is an ongoing research effort under the

Dredging Operations and Environmental Research (DOER) Program. Efforts are being directed to

streamline and standardize the process and minimize costs, employing mineralogically significant

size cut points (silt/sand and clay/silt) and an innovative density separation process. The procedures

under development use equipment that is widely available (Olin et al. 1999). Contaminant

distribution studies examining relative contaminant distribution behavior and the contribution of

the clay fraction to overall contaminant levels are being conducted in connection with this effort.

Guidance on this work will be provided in a subsequent DOER technical note.

PHYSICAL SEPARATION PROCESSES: The information quoted or summarized in this

section was taken from Olin et al. (1999), to which the reader is referred for a more in-depth

discussion of physical separation processes and equipment. "Much of the philosophy of volume

reduction comes from hundreds of years of mining experience worldwide...The remediation

engineer has the same challenge; to remove small amounts of contaminants from complicated and

diverse feeds. Volume reduction uses a fundamental understanding of the physical and chemical

characteristics of the feed soil or sediment and a simple, inexpensive treatment train to remove clean

material. This results in a smaller mass of contaminated material to be either further treated or

disposed." The focus here is on physical separation, but physical separation processes can be, and