RESULTS AND DISCUSSION

Table 2. Itemized costs ($) of treatment options for 500 yd3 (382 m3) of contaminated soil.

Bioventing

Landfarming

(percent time blower on)

(percent lined and covered)

Item

100%

50%

20%

100%

0

Earthwork subcontract

3950

1950

Mobilization/decontamination

200

Electrical subcontract

2760

N/A

System piping (venting, water)

750

0

Bottom liner

480

2400

0

Regenerative blower

1000

N/A

Polyethylene cover

250

0

Electrical power (

||content||

.11 per kW)

1445

723

301

0

Electrical start-up draw

217

108

45

0

(15% of total electrical)

Nutrients

450

Total

11,502

10,671

10,186

5250

2600

Total cost per yd3

23

21

20

11

5

Total cost per m3

17.58

16.06

15.29

8.41

3.82

Random samples taken at the beginning of the

RESULTS AND DISCUSSION

field study were analyzed for TPH components

to determine the variability at to. Results (Table 3)

Initial samples and contaminant homogeneity

Measuring field bioremediation rates is diffi-

show that distribution of TPH and other petroleum

cult because of the nonuniformity of contaminant

fractions had a Coefficient of Variation (CV) less

distribution; it is also costly. Yet, bioremediation

than 5%, were homogeneous, and demonstrated

rate comparisons, based on direct field measure-

that using mixed soil in permeable bags provides

ments, are the true measure of treatment effects. To

uniform to samples. Thus, this technique provided

effectively compare landfarming to bioventing, we

equal starting concentrations on which to base

must either develop a better system for field moni-

remediation rates.

toring, or wait a longer time before measuring so

On the basis of previous studies of relative rates

that differences would be more pronounced, or

of degradation, the diesel soil was expected to

know that the effects of the treatments would be

degrade faster than the heavy oil because of the

vastly different. Because the effect of the treatments

smaller percentage of more recalcitrant, heavy-oil

is our unknown and a comparison of rates over time

compounds.

is important, the only acceptable option was to de-

velop a better system for field monitoring.

Initial incubation period: elapsed time 0 to 54

To determine a bioremediation rate at any given

days, 29 July to 21 September 1995

time, we must be able to compare contaminant

concentrations to the initial concentration of the

Diesel

contaminant in the soil. To make valid rate com-

For diesel-contaminated soil, all treatments de-

parisons, it is essential either to provide known

creased relative to the to values during the first 54

contaminant concentrations at the start (to) or take

days of treatment (Fig. 2). Landfarming with nu-

an extremely large number of samples to quantify

trients was the most effective and had significantly

the random error caused by spatial variability.

(P < 0.05) lower TPH concentrations at t54 than

Because of the "destructive nature" of taking

did the other treatments. TPH concentrations in

samples (i.e., you cannot actually resample the

the landfarming without nutrient treatment also

same volume of soil) and the difficulty in deter-

decreased relative to the to values and had lower

mining exactly how many samples would be re-

(P < 0.10) TPH concentrations than the bioventing

quired to sufficiently reduce the variability, we

treatments, but not to the same degree as TPH

chose to use a method that provided uniform

concentrations when nutrients were added.

sample concentrations at the start, as described in

Bioventing treatments, both with and without

the Sampling and Analysis section.

nutrients, were less effective than the landfarm-

4