Uoited States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-81-189 Oct. 1981
Project Summary
Performance Testing of
Four Skimming Systems
H. W. Lichte, M. K. Breslin, and G. F. Smith
A test program (conducted during
the 1979 season) evaluated skimmer
performance in collecting oil floating
on water using several wave condi-
tions, tow speeds, and skimmer
operating parameters. Performance
tests were conducted at the U.S.
Environmental Protection Agency's
(EPA) Oil and Hazardous Materials
Simulated Environmental Test Tank
(OHMSETT) on four commercial oil
spill cleanup devices: the Sapiens
Sirene* skimming system, the Oil Mop
remote skimmer, the Troil/Destroil
skimming system, and the Versatile
Environment Products Arctic Skimmer.
Tests described in this report were
sponsored by the OHMSETT Inter-
agency Technical Committee (OITC).
The 1979 OITC members were the
EPA, U.S. Navy-SUPSALV, U.S.
Navy-NAVFAC, U.S. Coast Guard,
U.S. Geological Survey, and Envi-
ronment Canada. A 16-mm film "600
Foot Ocean," produced to summarize
the results presented in this report, is
available through the EPA, Office of
Research and Development, Oil and
Hazardous Materials Spills Branch,
Edison, New Jersey 08837.
This Project Summary was devel-
oped by EPA's Municipal Environ-
mental Research Laboratory, Cincin-
nati, OH, to announce key findings of
the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
'Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use.
Introduction
Results and methods used for tests
sponsored by the 1979 OITC are
presented in this report for the following
commercially available spill cleanup
equipment: (1 ) Sapiens Sirene Skimming
System, c/o Sapiens, 23-27 Avenue de
Neuilly, F75116 Paris, France; (2) Oil
Mop Remote Skimmer, Oil Mop Pollution
Control, Ltd., Toronto, Canada; (3)
Troil/Destroil Skimming System, Hyde
Products, Inc., 810 Sharon Drive,
Cleveland, Ohio 441 45; and (4) Versatile
Environment Products Arctic Skimmer,
60 Riverside Drive, North Vancouver,
British Columbia, Canada V7H1T4.
Each system was shipped from a
foreign country on loan to OHMSETT.
Tests were conducted to evaluate (1)
best oil collection performance, (2)
environmental conditions limiting op-
eration, (3) mechanical problems, and
(4) device modifications for improving
performance, or operating limits, or
both.
In the full report, quantitative per-
formance data to support conclusions in
the above areas are presented based on
the following parameters calculated
from steady state test results.
(1) Throughput Efficiency (TE) — Per-
centage of oil entering the skim-
mer that is recovered. This pa-
rameter is important for advancing
skimmers.
TE= Oil recovered
Oil distributed
(encounter rate)
(2) Recovery Efficiency (RE)— Percent-
age of oil in the fluid recovered by
-------
the skimmer; applies to all devices
in this report and is useful for
evaluating storage required to
contain fluid recovered at a spill.
RE = Oil recovered
Total Fluid
Recovered
(3) Oil Recovery Rate (ORR)— Volume
of oil recovered per unit time; also
applies to all devices in this report
and is useful in determining time
needed to clean up a spill of
known volume.
DRR= Volume oil recovered
Unit of time
Direct comparison of test results
should be avoided because all skimmers
were operated differently. The Oil Mop
remote skimmer and the Versatile
Environment Products Arctic skimmer
were operated as both stationary and
advancing skimmers, whereasthe other
two were operated as advancing skim-
mers only.
Sapiens Sirene
Skimmer System
Skimmer Description
The Sapiens Sirene, as tested, is a
two-stage oil skimming system com-
posed of five components (Figure 1). The
first stage is the oil herding section (side
floats); the second is the oil collection
section (rear float, hoses, and pump).
The five components are:
(1) a 14.5-m-long float of inflated
flexible fabric with an increasing
boom draft from the forward to aft
(right side or wing section);
(2) a 7.5-m-long oil inlet section that
includes the narrowing funnel
leading into the suction box with a
torpedo-like float supporting the
oil/water inlet in the center;
(3) another 14.5-m long float of
inflated flexible fabric with the
boom draft increasing from for-
ward to aft end (left side or wing
section);
(4) an aluminum suction box, with
floats, that is clamped onto the
upper part of the apex of the rear
funnel to accept oil collected; and
(5) 20 m of 110-mm hose and two air
driven, double-acting diaphragm
pumps (162 mVhr capacity) to
remove the collected fluid from
the Sirene to the collection
barrels.
Conclusions
From July 9-20,1979,43 oil recovery
performance tests were conducted with
the Sapiens Sirene skimmer (Figure 1);
31 tests with a high viscosity oil and 12
with a medium viscosity oil.
Best Performance—
Consistently, the highest values of
RE, TE.and ORR were obtained during
tow tests with waves. This result was
surprising since waves generally cause
poorer performance in oil skimmers.
The tests in high viscosity oil produced
better results than the tests in medium
viscosity oil. Medium viscosity oil was
entrained and lost from the system
more easily than the high viscosity oil
because of interfacial shear forces.
The best skimmer performance data
(highest numerical results) are presented
along with accompanying test conditions
in Table 1. Because of the skimmer's
operating principle, the highest values
of TE, ORR, and RE did not occur under
the same test conditions. Test oil
logistics prevented using enough oil to
saturate the system over the entire tow
test, thus absolute maximums for ORR
and RE were not determined.
No oil was lost because of wave
splashover. The cylindrical design of the
continuous flotation elements caused
the oil and water to be splashed
forward, in front of the boom; this was
true even at the highest tow speed run
in the roughest wave condition. Another
reason for lack of splashover was the
virtual absence of device heave relative
to the water's surface. The great
amount of flotation coupled with the
concave skirt design, which tends to
hold the device to the water's surface,
acted to maintain a relatively large,
constant freeboard.
Operating Limits—
Based on quantitative and qualitative
test results, the operating limits of the
Sirene skimmer appear to depend on
the following three items:
(1) Oil entrainment phenomena at
tow speeds above 0.75 knot cause
oil to escape the skimmer before it
can be pumped out. Such losses
occurred (1) beneath the points of
attachment between the side sec-
tions and the rear collection sec-
tion, (2) beneath the large floats
on either side of the oil/water
inlet, and (3) out the water outlet
located beneath the oil suction
box in the aft end of the device.
(2) Limited pump capacity allows oil
to build up in front of the oil inlet
and, therefore, is subject to
entrainment and shedding be-
cause of water flow beneath the
oil.
(3) Oil cannot flow easily to the oil
suction box after it enters the oil
inlet. This allows the oil slick to be
subjected to water passing below
it for a longer period of time, thus
increasing shedding and entrain-
ment of oil droplets.
The pumping system did not severely
emulsify the collected oil and water.
This is evidenced by the similarity
between the recovery efficiency samples
obtained by allowing gravity to separate
the oil and water and those obtained by
centrifuging oil/water samples.
TE was not affected by slick thickness
whereas RE and ORR were directly
dependent.
Device Modifications—
Device modifications recommended
for improving the performance of the
Sapiens Sirene system are:
Table 1. Best Performance - Sapiens Sirene (High and Low Viscosity Oil).
Performance
Parameter
High Viscosity—
TE
RE
ORR
Low Viscosity —
TE
RE
ORR
Highest
Value
100%
71.0%
39.7 m3/hr
99%
66.5%
39.8 m3/hr
Tow
Speed
(kt>
0.50
1.25
1.0
0.75
1.25
1.25
Wave
HxL
(m x m)
0
0.6 HC
0.6 HC
All waves
0.7 HC
0.5x11.6
Test
No.
1
27
26
36, 40.
45
45
46
-------
Oil Suction Hose
and floats
Aluminum
Floats and
Transition
Piece
Narrowing
Fabric Funnel
(second stage)
Rear float (second
stage)
Port Side Float (first stage)
Variable Depth Skirt
Adjustable Skirt
Ballast Chain
Direction of Tow
Starboard Side Float
(first stage)
Figure 1. Sapiens Sirene skimming system.
(1) Extend the oil inlet across the
entire rear section of the device.
Eliminate the floats on either side
of the oil inlet or place outside the
side flotation elements. The oil
would travel directly from the side
sections into the narrowing funnel
behind the oil inlet, thus elimi-
nating the severe angle change
that developed at the point where
the side sections join the rear
section.
(2) Improve the narrowing funnel
behind the oil inlet to allow oil to
move more easily through it to the
oil suction box. If oil could be
transported through the funnel
area at the same flow rate as the
encounter rate, there would be no
pool of oil to be subjected to the
interfacial shearing forces of the
water passing beneath the pool
and out the water exit. Longitudi-
nally arranged flotation elements
spaced across the funnel would
allow oil to pass easily by keeping
the fabric above the slick's sur-
face.
(3) Increase the system pumping
capacity by improving the ar-
rangement of the two double-
acting diaphragm pumps used to
transfer the oil/water fluid from
the suction box to the collection
barrels. The pumps would be 12%
more efficient if used indepen-
dently of each other rather than in
a common inlet, common outlet
arrangement. A doubling of
present pump capacity is recom-
mended.
(4) Replace the center torpedo float
on the oil/water inlet with two
floats, spaced at one-third and
two-thirds the distance acros»the
mouth, to eliminate turbulence
generated by the float directly
upstream of the oil suction box.
If the recommended modifications or
ones that serve the same purpose are
incorporated into the system, another
test program should be performed at
OHMSETT. The system shows promise
through its innovations in design and
material use.
Oil Mop Remote Skimmer
Skimmer Description
The Oil Mop remote skimmer model
(Figure 2) was fabricated by Oil Mop
Pollution Control, Ltd., Toronto, Canada,
as a preliminary model of a full-size unit
to be used for Arctic oil spill recovery
service. The skimmer is designed as an
unmanned unit controlled by an umbili-
cal electric cable. The operating principle
is that of oil slick sorption onto a bank of
polypropylene rope mops, rotating in the
vertical plane to produce zero velocity
relative to the surface of the water
during forward motion of the skimmer.
Conclusions
From August 6-10, 1979, 19 oil
pickup performance tests were con-
ducted with the Oil Mop remote skimmer:
-------
Sump Pump
Sump Pump
Rotatable
Prop
Dirve Rollers
Foam-Filled Hulls
Fixed Prop
Plywood Bulkhead Added for Test,
Drive Rollers
Figure 2. Oil Mop remote skimmer.
6 with high viscosity o\\ and 13 with
medium viscosity oil.
The primary test objective was to
generate design information for future
construction of a larger version to be
built -for Arctic service in Canadian
waters. The following test conclusions
relate to the design criteria for the larger
skimmer:
(1) At least three powered rollers
must be provided to prevent
slippage of the oil mop, especially
when saturated with high viscosity
oils.
(2) The mop-to-oil slick contact length
and rotational mop speed of the
full-scale skimmer should be
selected after conducting a series
of oil-mop saturation tests with
the various viscosity oils expected
to be encountered. The oil-mop
saturation times can be compared
with various values of skimmer
length divided by mop speed.
Sufficient time was not available
during the single test week to
determine mop-oil saturation
time for the two test oils. In tests
with oils of 185 cSt viscosity,
however, ORR was unaffected by
reducing the mop-to-oil slick
contact length from 1.9 m to 1.2
m. ORR performance did fall off
rapidly, however, when the con-
tact length was reduced to 0.6 m.
(3) To maximize oil recovery rate, the
full-scale skimmer hulls should
be open on the sides, if possible,
to allow oil to enter from the sides
as well as the front. This will
increase the ORR by allowing
more oil to come in contact with
the tops of the oil mops floating
above the water surface. The
skimmer beam should be maxi-
mized to increase the oil-mop
surface area being laid down on
top of the slick as it enters the
front of the skimmer.
(4) A positive displacement type off-
loading pump is necessary to
ensure rapid offloading of col-
lected oil over a wide oil viscosity
range.
Best Performance—
The objective of these tests was to
obtain design information for a larger
unmanned Oil Mop remote oil skimmed
-------
The highest numerical results obtained
in these tests may not be the maximum
obtainable with the full-scale Oil Mop
skimmer. The highest numerical values
of the three performance parameters of
TE, RE, and ORR for the present
OHMSETT tests are summarized in
Table 2.
Operating Limits—
Based on numerical and qualitative
test results, the operating limits for this
skimmer appear to depend on two
factors: (1) oil mop-to-oil slick contact
area and (2) slippage of oil-soaked mops
through the squeezing roller assembly.
First factor. During stationary tests
with medium viscosity oil, areas of clear
water appeared under the point where
the oil soaked mops were lifted out of
the water at the stern of the skimmer.
This indicated that the mops, at least on
the side facing the oil slick, were fully
saturated with test oil. The ORR
performance increased when the entire
skimmer hulls were lifted clear of the
water by an overhead crane, thereby
exposing the tops of the floating oil mop
to splashing contact with oil from the
sides. In the full-scale device, the ORR
can be increased by maximizing the
skimmer beam and opening the skimmer
hulls along the length of the skimmer to
allow oil to splash onto the mops from
the sides.
Second factor. It is essential that
future skimmers include three powered
rollers instead of two as in the present
Oil Mop remote unit. Two rollers are
needed to squeeze oil from the saturated
rope mop. A third roller, operating
against one of the other two rollers, is
needed to provide the mop tension to
maintain the oil mop rotational speed.
By visual observation, in almost all tests
with high viscosity oil and in some with
medium viscosity oil, slippage of the
mops occurred at the two squeezing
rollers. This resulted in the mops
remaining on the oil slick after they had
become fully saturated, reducing the
net oil pickup per unit time.
Troil/Destroil Skimmer System
Skimmer Description
The Troil/Destroil skimmer system
assembled for OHMSETT testing com-
bined the Troilboom Giant 1.5-m boom
manufactured byTrelleborg AB, Sweden,
and the Destroil Model DS210 skimmer
pump manufactured by DESMI A/S,
Denmark.
Table 2. Best Results - OMI Remote (High and Medium Viscosity Oil).
Performance
Parmeter
High Viscosity —
TE
RE
ORR
Medium Viscosity —
TE
RE
ORR
Highest
Value
30%
96%
2.6 m3/hr
43%
93%
2.7 rtf/hr
Tow
Speed
(ktl
0.5
0.5
1
0.5
0.5
1
Waves
HxL
(m x ml
0.3 x 4.2
0
0
0.3 x 4.2
0.2 x 7.0
0
Slick
Thk.
(mm)
9
6
6
9
9
9
The Troilboom consisted of four 6.4-m
sections of a 1.5-m-high collection
boom (Figure 3). At the center of the
boom is a 3.5-m-wide opening with an
additional section of boom attached that
provides a pocket to collect the swept oil
and contain the floating skimmer pump.
The boom panels are supported by
curved fiber glass battens that provide a
concave boom profile while under tow.
The boom is towed by an independent
external load line that connects to the
battens by individual bridles. This
arrangement allows each boom section
to conform to waves and maintain a
nearly constant waterline.
The Destroil skimmer pump (Figure 4)
is a hydraulically-driven screw pump.
Oil is recovered as it flows over the
central hopper weir into the exposed
pump screw. Skimmer flotation is
provided by two fixed-position floats and
one that is adjusted by remote ballasting
with compressed air. The pump is driven
by a remote diesel-hydraulic powerpack
that provides pump power and air
ballast control. The pump discharges
through a 127-mm, flexible discharge
hose. The screw and hopper have a
macerator cutting edge for chopping
debris that may enter the pump with the
oil.
Conclusions
The Troil/Destroil skimmer system
was tested at OHMSETT August 15-24,
1979. The tests were conducted to
measure the recovery performance of
the combined boom and skimmer
system and observe the interaction of
the boom and floating skimmer.
Best Performance—
Table 3 shows the best skimmer
performance for high and low viscosity
oils. The skimmer performance param-
eters RE and ORR were at their highest
when the boom preload oil volume was
at the test maximum. Because of the
high skimmer pump capacity, it was
necessary to change the preload charge
volume. Tests were conducted with
various boomed preload volumes to
determine performance changes and
guidance for operator control.
Operating Limits—
The Troil/Destroil skimmer system,
as tested, has the following operating
limits:
(1) The maximum towing speed at
which the Troilboom can retain
collected oil, without significant
loss, is 1 knot.
(2) The maximum pumping rate of
the Destroil skimmer pump is
approximately 37.4 mVhr.
At a boom towing speed of 1 knot, the
Troilboom lost oil at a rate of approxi-
mately 2.3 mVhr. when the towing
speed was increased to 1.25 knots, the
oil loss rate increased to approximately
23 mVhr. Oil losses were consistently
observed to be the result of vortex
shedding occurring near the side walls
of the skimmer collection pocket.
The Destroil skimmer has an advanc-
ing screw pump that increases pumping
rate as the viscosity of the pumped
mixture increases. The maximum vis-
cosity of the test oil was approximately
925 cSt.
The test series did not determine the
pumping capabilities of the skimmer
with higher viscosity mixtures. The
skimmer pump was capable of ingesting
a quantity of floating debris deposited
during one of the test runs. The boom
and skimmer system showed good
wave-following in test waves up to 0.47
m harbor chop. The independent towing
bridle allowed the boom to maintain a
relatively constant waterline while in
the wave.
-------
Bridle Lines
Pocket
Boom Panels
Fiberglass Battens
Figure 3. Troilboom illustration.
Remotely Adjustable Air Ballast Float
Hopper
Mechanically Adjustable Floats (3 positions)
r:
Screws
Hopper
Pump Discharge
Figure 4. Destroil skimmer pump.
-------
Recommendations
The Troil/Destroil skimmer system,
as tested, should be used at speeds not
exceeding 1 knot. The device can be
used in moderate waves without signifi-
cant performance reduction. Once
rigged, a single operator can control the
recovery operation by adjusting pumping
rate and skimmer height. Skimmer
performance for RE and ORR can be
increased if the operator allows a
precharge oil volume to remain in the
skimmer pocket while operating the
pump. At least 3.8 m3 is necessary to
provide an appropriate precharge volume
in the boom pocket for good performance.
The ORR of the skimmer is limited by
the pump capacity. A larger pump with
about three times the pumping capacity
should be considered for the Troil/
Destroil skimmer system. The boom and
skimmer must be able to survive a
variety of weather and towing condi-
tions. Towing bridles and boom stiffener
battens should be made stronger so that
the boom can survive greater towing
loads. The bridle attachment points
should be redesigned to provide for
quick rigging adjustments to allow a
proper boom towing attitude.
Versatile Environment
Products Arctic Skimmer
Skimmer Description
The Versatile Environment Products
Arctic Skimmer (Figure 5) is a non-self-
propelled advancing weir skimmer with
an adsorbent rotating belt. The skimmer
is equipped with a hydraulic power pack
that powers the collection belt mecha-
nism, the oil offloading pump, and a
water pump for ballasting and powering
the water jet booms. The skimmer can
be operated with the power pack and
control station onboard the skimmer, or
with the power pack removed, the
skimmer can be remotely controlled by
means of a 27.4-m umbilical hose
bundle.
Conclusions
The Versatile Environment Products
Arctic Skimmer was tested at OHMSETT
October 15-23, 1979. This skimmer is
an air-transportable, remotely controlled
version incorporating the original
Bennett oil collection principle. The
Arctic Skimmer is a prototype version
developed for Environment Canada and
intended for cold-weather use.
The objectives of the Arctic Skimmer
ssts were to observe the operator
TableS. Peak Performance - Troil/Destroil Skimmer System (High and Low
Viscosity OH).
Performance
Parameter
Highest
Value
Tow
Speed
(kt)
Boom
Preload
Waves
HxL
(m x m)
High Viscosity—
RE
ORR
Low Viscosity—
RE
ORR
93%
20.9 m3/hr
91%
23.7 m3/hr
0.75
0.75
0.75
0.75
3.8
3.8
3.8
3.8
0.26x4.2
0.26x4.2
calm
calm
Figure 5. Versatile Environment Products Arctic skimmer as tested at OHMSETT.
control and mechanical performance.
Skimmer performance was measured
over several tank test conditions to gain
operator control experience that would
maximize skimmer performance, and
the following conclusions were deter-
mined from the test series:
(1) The skimmer can be controlled
from either onboard with the
power pack mounted on the
skimmer or remotely with the
power pack removed and con-
nected to the skimmer through
the hydraulic control lines. The
several adjustable skimmer
settings can be preset for remote
skimmer operation.
(2) The water jet nozzles can effec-
tively concentrate and sweep oil
into the skimmer at speeds from 1
knot to 4 knots in both calm and
wave conditions.
(3) The settings of the three adjustable
skimmer doors are critical for
maximum performance at each
speed. A graph of skimmer door
settings was developed from
performance tests.
Best Performance—
Table 4 lists the best skimmer
performance for high and low viscosity
oils. In addition to regular towing tests
with a 3-mm slick, several tests were
performed to establish the maximum
ORR of the skimmer. In these tests, at
least 25.0-mm thickness of oil was
presented to the skimmer and the
skimmer then adjusted for maximum
recovery rate.
S. GOVERNMENT PRINTING OFFICE:1981--559-092/3312
-------
Operating Limits—
Each of the performance parameters
can be maximized by the operator as
follows:
1. To maintain high throughput effi-
ciency, the skimmer should be
operated at speeds not greater than
2 kt. The maximum safe skimming
speed in calm water is 5 kt.
2. To maintain high recovery efficiency,
a thick oil layer should be maintained
at the collection belt. The belt should
be raised to be even with the oil/
water interface.
3. The oil recovery rate of the skimmer
is limited by the maximum pumping
rate of the skimmer pump. The actual
rate will vary with the viscosity and
discharge hose requirements. Maxi-
mum pumping rates observed for the
8-mm discharge hose with 4.5-m
head was 23.6 m3/hr.
The Arctic skimmer was tested in
waves 0.18-m high by 9.4-m long. The
maximum height of wave in which the
skimmer can perform is about 0.3-m
height, which is about the maximum
depth of the bow door. As the skimmer is
able to respond to longer period waves,
the actual height of the wave can
increase if the relative wave height at
the skimmer mouth remains nearO.3 m.
Recommendations
(1) The Arctic skimmer is intended to
be transported partially disassem-
bled. The skimmer could arrive
with the belt mechanism removed
and lowered, flotation collars
removed, and additional equip-
ment stored within the skimmer.
Clear rigging, reassembly, and
system check-out procedures
should be fixed on the skimmer to
facilitate quick deployment.
(2) The skimmer operating manual
should be updated to include
Table 4. Peak Performance - Versatile Environment Products Arctic Skimmer
(High and Low Viscosity Oil).
Tow Waves Slick
Performance Highest Speed H x L Thk.
Parameter Value (kt) (m x m) (mm)
High Viscosity—
TE
RE
ORR
Low Viscosity—
TE
RE
ORR
96.3%
85.0%
20 m3/hr
99.4%
97.4%
19.4 m3/hr
2
0
0
1
0
0
calm
calm
calm
calm
calm
calm
3.4
25.0
25.0
2.8
25.0
25.0
optimal skimmer door settings,
belt speed, and pump control
calibration curves.
(3) An auxiliary cooler should be
provided for the hydraulic reservoir
for extended operation in warm
weather.
(4) The rigging for the water jet
sweep booms should be simplified
to allow quick adjustment from
the skimmer.
(5) The squeeze belt collection sump
could be enlarged to contain belt
splashover at high speed.
(6) The adjustment screws for the aft
and middle gill door should be
modified to make readjustment
quicker.
The full report was submitted in ful-
fillment of Contract No. 68-03-2642 by
Mason & Hanger-Silas Mason Co., Inc.,
Leonardo, New Jersey 07737, under
the sponsorship of the U.S. Environ-
mental Protection Agency. Technical
direction and evaluation of the Oil Mop
Remote skimmer and the Versatile
Environment Products Arctic Skimmer
were subcontracted to PA Engineering,
Corte Madera, California.
H. W. Lichte. M. K. Breslin, and G. F. Smith are with Mason & Hanger-Silas
Mason Co., Inc., Leonardo, NJ 07737; D. J. Graham andR. W. Urban are with
PA Engineering, Corte Madera, CA 94925.
Richard A. Griffiths is the EPA Project Officer (see below).
The complete report, entitled "Performance Testing of Four Skimming Systems,"
(Order No. PB 82-101 353; Cost $9.50, subject to change) will be available
only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Oil and Hazardous Materials Spills Branch
Municipal Environmental Research Laboratory—Cincinnati
U.S. Environmental Protection Agency
Edison, NJ 08837
United States
Environmental Protection
Agency
Center for Erwironmenta) Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
Third-Class
Bulk Rate
DTV
?jo s EARBOKN ST
&r,n TI
------- |