Results - trel02-380032

an extended period of time. The supernatant was then poured off, the supernatant

and remaining wet solids weighed, and then water content of the wet solids

determined by oven drying. In this manner, the initial solids content of the slurry

could be calculated. The settled solids were analyzed on the Coulter particle

size analyzer for particle size distribution.

Results

Particle size analysis

Results of the particle size data for samples taken from the feed and overflow

samples while stirring versus samples taken from the settled slurries were

different. Because the settled slurries were still relatively liquid but at a higher

percent solids, it was possible to mix them thoroughly and avoid rapid settling of

coarse particles, thus producing more representative samples of all size ranges in

the solids. The feed and overflow particle size distributions reported in Table 5

are for the subsamples taken from the settled samples. The underflow contained

less than 8 percent fines by volume. Depending upon the specific gravity of the

particles, this can be converted to percent fines by weight. Because the Coulter

counter measures particle volume only (void volume is not measured), percent

sand by volume can be taken to be approximately equivalent to percent sand by

mass, assuming the same specific gravity for all particles in the material.

Percent moisture/percent solids

Results of percent moisture (Wwater/Wtotal) and percent solids (Wsolids/Wtotal)

for the process streams are summarized in Table 6. The percent solids of the feed

varied from 1.8 to 5.9 percent by weight. This is a relatively dilute feed stream.

Although the separation efficiency is enhanced by a dilute feed stream,

operational efficiency overall is lower than optimum. Percent solids of the

underflow was quite high, ranging from 75.2 to 80.3 percent, reflecting the

coarse nature of the underflow. Mean percent moisture of the underflow was

approximately 22 percent, compared with approximately 98 percent for the feed

and overflow process streams.

Chemical analysis

The results of the chemical analysis for each replicate at each sampling time

were averaged for the three process streams, and are summarized in Table 7. For

comparison, results of the bench-scale characterization for these fractions are

given in parentheses. Qualitatively, the bench-scale testing appears to have given

a relatively representative indication of the contaminant levels in the field-scale

process streams. To evaluate whether the concentrations in the process and

characterization fractions are essentially equivalent, the data were evaluated.

Contaminants included arsenic, cadmium, chromium, copper, lead, mercury,

nickel, selenium, silver, zinc, barium, PCB 1242, PCB 1260, TOC, oil and

grease, and TRPH. One-half the detection limit was used for contaminant

concentrations less than the detection limit. The Statistical Analysis System

(SAS) release 8.1 was used to perform the data analysis (SAS Institute, Inc.,

1989a, 1989b). The statistical procedures and assumptions are more fully

Chapter 2 Project Description