Case Study II

This study involves the remediation of 40,000 tons of heavy metals-contaminated soil at the Thompson-Isaacson site in Tukwila, Washington. In this soil, arsenic was the principal metal of concern. The other metals that were evaluated were barium, chromium, copper, nickel, lead, and zinc. The need for the remediation arose as a result of the planned development of the site, which entailed the construction of an industrial structure necessitating substantial excavations for foundations and pedestrian tunnels [5].

Elevated levels of arsenic are believed to have occurred as a result of the dredging, filling, and straightening of the Duwamish channel beginning around the turn of the century. Presently, one of the site boundaries is the Duwamish Waterway, which once flowed through the approximate center of the site before the channel was straightened and redirected. Various site investigations [6,7] indicated that the arsenic contamination may have resulted from a variety of fill materials generated by smelting and other ore processing operations.

v Pb/i

00

*

A O

C» Nl

. * ♦ s

♦ ♦

f*

— ♦ Zn

A

*

-

ft

+

* v

i

+

/

f

A

A

i

A

A

£

A A*

*

C

n ■ "Ä

a

u

A

D "

1

c

1

rf

c

JO

3

c

)

c [

b

3

!

3

<

c

i>

(

k

0

*

i.

c

n

n

ff

c

V

17

7

o

OD

B

C7

^

1

7

0 ce

o

<7

S7

7 V

7

&

1.0 10 100 1000

10,000

1.0 10 100 1000

TTLC LEVEL (mg/kg)

Figure 3 Relationship between soluble and total concentration levels in untreated soil (Case Study I).

so ni 3

A number of factors entered into the decision to use on-site treatment for the remediation of the heavy metals. Basically, this alternative allowed the site owner to maintain control in that the treated material would remain on-site, thereby reducing the reliance on off-site hazardous waste landfill disposal and its associated long-term liabilities. Further, the use of on-site treatment at high processing rates (100 tons/hr) eliminates the excessive time associated with trucking large quantities of material considerable distances to Class I disposal sites. Finally, in terms of construction, since the characteristics of the backfilled STS-treated material are suitable for the placement of buildings, the need and costs associated with imported fill materials can be significantly reduced or eliminated.

The conditions for the use of on-site treatment were consistent with the treatment in tank by generator regulations. Consequently, an RCRA tank was designed and constructed with the appropriate membrane liner, asphalt concrete pavement, catch basins, etc. Basically, this RCRA tank was designed to not only contain the soil during treatment, but also to collect all contact and noncontact liquids, which could be subsequently transferred to appropriate storage tanks. A secondary containment system was also installed for the purpose of monitoring any leakage from this system.

Normally, full-scale or commercial processing would follow the treatability study. However, in this project, the Washington State Department of Ecology, as part of the approval of the Site Remediation Action Plan [8], requested a 4000-5000-ton pilot field test prior to full-scale remediation for the purpose of verifying the treatability data and the effectiveness of the treatment under actual field conditions. It was necessary to first construct the previously mentioned RCRA tank before the test could begin. The mobile treatment system was then erected in the completed tank. This system consisted of the feed hopper, mixing unit, chemical delivery system, and associated feed and discharge conveyors. In this project, the curing and subsequent stockpiling of the treated material occurred in the tank.

Approximately 4500. tons of soil was treated at commercial processing rates of 80-100 tons/hr during the 5-day pilot test. Material requiring treatment was excavated to the water table (about 12 ft below the surface) using a standard backhoe machine. Prior to entering the treatment unit, the excavated soil was screened to a 3A in. particle size through a two-step process using grizzly bars and a trommel screen. Stockpiles of the screened material were sampled and evaluated for heavy metals concentrations before treatment. Composite samples of the material were also evaluated before and after treatment. Control of the treatment protocol was accomplished through a mass balance, which entails continuous measurement of the rate at which material and reagents enter the treatment unit. The necessary instrumentation included a certified belt scale on the material in-feed conveyor, calibrated and computer-controlled rotary feeders for dry reagents, and in-line flow meters and calibrated delivery pumps for liquids. After the successful completion of the pilot test program, the remaining 36,000 tons of material was processed and subsequently backfilled into the excavation.

1. Treatment Levels

In this particular soil, as in Case I, a comparison of the total (TTLC) and soluble concentrations (TCLP) for arsenic, copper, and zinc (Figure 4) illustrates that the concentrations of heavy metals can vary over several orders of magnitude throughout the site. Arsenic is the principal metal of concern; copper and zinc exhibited the next highest concentrations but did not exceed TCLP limits. The data shown in Figure 4 were obtained during the pilot test program. During the production phase, the evaluation of total concentration levels was not included in the routine data collection.

An analysis of the data in Figure 4 indicates that the degree of solubility is a function of total concentration. In the case of arsenic, the soluble concentrations were about five times

-

A

A

4-

A.

it?

m

11 ■

■ 1 ■

L

--

— -

100 1000 10000 TTLC LEVEL (mg/kg)

100 1000 10000 TTLC LEVEL (mg/kg)

Relationship between soluble and total concentration levels in untreated soil (Case Study II).

Figure 4

greater at the highest TTLC levels. In other words, the TCLP levels range between 200 and 1000 times less than the TTLC levels as they decrease from 10,000 to 100. Copper exhibits similar behavior even though the ranges in TTLC and TCPL levels only span an order of magnitude. On the other hand, the levels of zinc are clustered such that the TCLP level is about 100 times less than the TTLC concentration. It is interesting to compare this soil with that of Case

1, which also exhibited concentration levels of the aggregate or mixture of metals that spanned several orders of magnitude (Figure 3). However, unlike the arsenic in this soil, the metals in the Case I soil were grouped or clustered in relatively narrow ranges within the overall distribution. Also, another notable exception was zinc, which was not clustered. In these particular arrangements (Figure 3), the average soluble concentrations (determined by the STLC procedure) were about 20 times less than the total concentration. Extraction procedures will affect the behavior of soluble concentrations in a mixture of metals. In terms of the above comparison, this would involve differences in the leaching characteristics between the previously discussed TCLP procedure and the STLC extraction, which uses a sodium citrate buffer with a tenfold dilution factor.

2. Treatment Results

The treatment protocol used two cementitious materials, cement and lime, in the ratios of 20:3 or 15:5, with polysilicate additions of approximately 0.5 gal/ton. An example of the ability of the process to reduce the soluble concentration is given for arsenic treatment (Figure 5) during the production phase. It is important to note that in this case the treatment standard for arsenic was set at 1 ppm by the site owner instead of the regulated level of 5 ppm. The analytical evaluation protocol was established at 0.2 ppm, so values below this level were not determined.

Figure 5 Comparison of untreated and treated TCLP levels of arsenic (Case Study II).

In this project, a daily production goal of 750 tons of treated material was established as being reasonable for the site conditions, which were constrained by the ability to deliver material to the process (i.e., excavation, screening, sampling, etc.), the cycle time for curing and stockpiling in the tank, and the subsequent backfilling. The ability of the system to meet the processing goals is illustrated in Figure 6 in terms of daily tonnage and daily average tonnage over the production phase of the project. At times it was not possible to meet the daily production schedules, the reasons being (1) a mechanical failure in the lime delivery system on days 3 and 4, (2) the unavailability of feed material on day 31, and (3) the reduced quantity of feed material at the end of the major portion of the project during days 37 and 38.

An additional feature of this project concerned the contaminated water collected on-site in the process. The RCRA tank allowed liquids to be collected from the processing areas that included both the screening and treatment activities. Approximately 12,000 gal of water per day was generated from a combination of spraying for dust control and rain. An on-site system was implemented that allowed this water to be incorporated into the treatment process. The water from the collection system was introduced into the first of a series of four Baker tanks. This provided suitable capacity and time for the settling of suspended solids. The liquid from the fourth tank was then returned to the polysilicate-water blending tank in the chemical mixing portion of the treatment system. The concentration of arsenic in the liquid entering this tank, as determined by EPA methods SW-846, 3010, and 6010 was on the order of 4 mg/L. Blending this liquid with the process water at a ratio of 1:20 did not impact the treatment protocol. In addition, the project benefited from a significant savings in off-site disposal costs.

0 0

Post a comment