transported from the old burn pit area were
that the liner directs leachate into the collection
moved into the landfarm. The extent of this exca-
system (Fig. 3a).
vation in the burn pit area is shown in Figure 4.
Construction
The contaminated soil was disked weekly with a
After vegetation was cleared from the landfarm
2-ft-diam. (60-cm-diam.) disk for aeration and
site, the native sandy-silt material was excavated
nutrient mixing. The disk mixed the upper 8 to
to attain the design contours at the bottom of
12 in. (20 to 30 cm) of soil. Each week, 25 lb (11.35
the landfarm. Excavated material was used to
kg) of ammonium nitrate (NH4NO3) and 2 lb
form the surrounding dike. The lowest point in
(0.908 kg) of potassium (potassium sulfate) were
the structure was approximately 2.5 ft (0.76 m)
mixed with 150 gal (568 L) of water and allowed
below the original surface of the mineral soils,
to flow into the irrigation piping. The well pump
and the soils in the lower portions of the struc-
was activated to disperse the fertilizer mixture
ture were saturated by groundwater. Construc-
over the landfarm area. Irrigation water was
tion of the facility was delayed to allow the
added to the landfarm several times during
groundwater elevation to recede.
August to keep the soil's moisture content at
After the groundwater level receded, the
2585% of field capacity.
native soils were compacted, a separation fabric
The 1992 operational season began in mid-
was placed atop the sandy-silt in the lower half
April; a wheeled loader and a large snowblower
of the excavation, a 1-ft (30-cm) lift of embank-
were used to remove approximately 80 in. (2 m)
ment material was placed over the bottom of the
of snow from the landfarm. An additional 15 in.
entire excavation, and the berm height was also
(38 cm) of snow fell after the winter accumula-
increased. A layer of 12-oz (4.07-g/m2) cushion
tion was removed. Meltwater, coupled with the
fabric was placed before HDPE liner sections
moisture from rainfall in late August and Sep-
were positioned parallel to the 230-ft (70-m)
tember 1991, saturated the material in the
axis of the landfarm, with seams overlapping
landfarm and delayed tillage until 23 June. The
56 in. (1315 cm). Heat-welds were made along
rate of fertilizer application was increased to
the seams and weld integrity was tested. Follow-
600 lb (272 kg) of ammonium nitrate, 150 lb
ing weld testing, a cushion fabric was placed
(68 kg) of triple super-phosphate, and 50 lb (23
kg) of potassium each month. Applying fertil-
over the HDPE liner and a 1-ft (30-cm) layer of
izer through the irrigation system during 1991
filter rock covered with a layer of nonwoven
resulted in uneven coverage because of leaky
separation fabric was added.
joints in the irrigation pipe, so dry fertilizer was
The leachate recovery system was a 4-in.
(10-cm) perforated PVC pipe placed in the trench
applied in 1992. A tractor-mounted broadcast
at the center of the landfarm parallel to the 230-ft
spreader was used, followed by tillage.
(70-m) axis. A riser at the lower end of the
landfarm was used for pumping water from the
Process monitoring
leachate system. The riser was connected by a
Landfarming is one of the most commonly
1-hp (10-kg cal/min) trash pump to a fertilizer
used and accepted soil biotreatment techniques
mix tank installed on a 2-ft-high (61-cm-high)
in temperate regions (Kuroda and Nusz 1994), yet
platform. A 30-ft (9-m) irrigation well for adding
information on landfarming that would expedite
supplemental water to the landfarm and a sur-
its application to cold regions was sparse. For
face irrigation system were installed. Surface
these reasons, we emphasized characterization
irrigation was through 2-in. (5-cm) aluminum
of the landfarm and the governing processes
pipes and rotating sprinkler heads. The nutri-
within it.
ent mixture was gravity fed to the irrigation
Microbial activity. We characterized the micro-
piping. Water could be delivered to the irriga-
bial activity at the landfarm by four methods. A
tion system from the well, fertilizer tank, or the
most probable number (MPN) sheen screen tech-
drain system (Fig. 3b).
nique was used to enumerate the oil-degrading
population. Radio-respirometry was used to
Soil treatment
determine the potential to mineralize specific
Approximately 500 yd3 (382 m3) of soil, previ-
hydrocarbons. Nonspecific microbial activity
ously stockpiled during cleanup of the fuel spill
in the field was estimated by measuring evolved
next to the railroad tanker body, and approxi-
carbon dioxide (CO2). This was done by alkali-
mately 3200 yd3 (2500 m3) of soil excavated and
trapping and both gravimetric and gas chromato-
7
Previous Page
