x>EPA
United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-81-102 July 1981
Project Summary
Ferrous Metals Recovery at
Recovery 1, New Orleans,
Louisiana
Louis P. Soldano
This report summarizes four final
technical reports that document a
series of five tests (referred to as Test
Nos. 4.01, 4.03, 4.05, 4.07, and
4.09) of ferrous metal recovery from
municipal solid waste (MSW) at the
New Orleans, Louisiana, Resource
Recovery Project (Recovery 1). Test
No. 4.01 documented the performance
of the trommel-unders magnetic drum
separator. Test No. 4.03 was conducted
to evaluate the efficiency of shredded
trommel-overs separator. Tests Nos.
4.05, 4.07, and 4.09 were conducted
to evaluate the efficiency and energy
consumption of a shredder and air
classifier added to the ferrous metal
recovery system. Hammer wear was
also measured for the shredder.
This Project Summary was devel-
oped by EPA's Municipal Environmen-
tal Research Laboratory, Cincinnati,
Ohio, to announce key findings of the
research projects that are fully docu-
mented in separate reports (see Project
Report ordering information at back).
Introduction
This report describes the ferrous
metal recovery testing done at a full
scale resource recovery operation at
Recovery I, New Orleans, Louisiana.
Resource recovery of MSW consists of
collecting materials (ferrous metals,
aluminum, glass, paper, etc ) from
processed or unprocessed municipal
refuse. At Recovery 1, attempts to
recover ferrous metal were done initially
with a trommel and two drum magnets.
The waste was passed through a trommel
(a large horizontal metal drum perfo-
rated with holes) The waste falling
through the holes (trommel unders) and
waste passing through the trommel
(trommel overs) was subjected to mag-
nets that recovered any ferrous metal.
Test No. 4.01 evaluated the belt
magnet performance for recovering
ferrous metal from the trommel under-
flow. Test No 4 03 evaluated the belt
magnet performance on the trommel
overflow.
The contractor for the ferrous metal
required that the product meet the
following criteria: maximum contami-
nation, 4 percent; bulk density, 21.5 to
26.0 Ib/ft3; and shred size, maximum, 5
percent less than 1 in. Since the trommel
and belt magnets could not produce
materials meeting this criteria, the
system was modified. Test No. 4.07 was
conducted to evaluate the efficiency of a
hammermill (shredder) followed by a
belt magnet in liberating and separating
contaminants from the ferrous metal
recovered from the trommel. The system
was further upgraded to include an air
knife and another belt magnet (Figure
1). Test Nos. 4.05 and 4.09 examined
the modified system's ability to protect
the shredder from dense objects and
reduce contamination of the ferrous
product. This report describes the results
of the evaluations.
-------
Test No. 4.01, Trommel-Unders
Magnetic Drum Separator
At Recovery 1, there are two primary
magnets The function of the magnets is
to remove ferrous metal from that
portion of the MSW that has either
passed through the 4-3/4-m.-diameter
circular openings in the trommel or
flowed out the end of the trommel. The
passing material is known as the trommel
underflow and the magnet, the trommel
underflow magnet. The drum magnets
used are suspended over the head
pulleys of the underflow and'overflow
conveyors
The trommel underflow magnet is an
electromagnetic type 42 in. in diameter
and 54 in long. The electromagnetic
coil is powered by direct current (DC)
supplied by a rectifier rated for 5,200
watts and 230 VDC.
The most significant size range for the
feed of the trommel is less than 4 in and
greater than 2 in In the feedstock, on a
clean metal basis, this size range ac-
counted for 72 percent of the magnetic
opportunities by weight
The magnetically removed ferrous
underflow product is cleaned by sepa-
rating light gauge ferrous material
(mainly cans) from heavier gauge (non-
can) ferrous materials. Contamination
carried over at the time the ferrous
material was extracted from the waste
stream is pneumatically removed. (The
term "contamination" not only includes
material caught on the ferrous product,
but material actually physically joined to
the product). This section of the report
discusses the efficiency of the trommel
underflow magnet as a light gauge
ferrous separator. The underflow mag-
net showed an average efficiency of 87
percent measured on an "as received"
basis. Magnetic efficiency on a clean (no
contamination) metal basis for the light
metal is 92 percent The efficiency of the
underflow magnet in terms of clean
metal was 85 percent for the heavy
ferrous metals. Magnetic efficiency was
highest for food cans (98 percent) and
lowest for composite cans (11 percent)
The magnet itself acts as a cleanup
device, since it differentially attracts
magnetics as a function of their con-
tamination As far as composite and
contaminated food cans are concerned,
an object that is less than one-third
magnetic material will probably be ex-
tracted by the magnet.
According to the test results, for every
100 Ib of trommel underflow magnet
product, 65-1/4 Ib is the desired light
Raw MSW-
Trommel >-Trommo/ overs-
III
Shredder
Discard
I
-Belt Magnet
Trommel Unders
Air Classifier -
Belt Magnet
Cyclones^
Aluminum Feed
Landfill
Ferrous
I
Ferro us
Concentrator
I
Light Ferrous
Heavy Ferrous-^
Figure 1. Final upgraded ferrous metal recovery system at Recovery 1
gauge ferrous, 22 to 14 Ib is heavy
ferrous, 2 Ib is loose contamination, 9-
1 /4 Ib is entrapped contamination (light
ferrous), and 1-1/4 Ib is entrapped
contamination (heavy ferrous) The
amount of contamination of the light
gauge, almost 12-1/2 percent, is un-
acceptable for market use
Assuming one day's waste of 650
tons, with a bulk density of 16.6 Ib/ft2, ;s
processed at approximately 62 5 tph,
the trommel underflow magnet would
produce 17-1/2 tons of magnetic metal
per day. Just under 13 tons would be the
desired light ferrous metal. Roughly 4-
1/2 tons would be heavy ferrous The
magnet would also produce 2 tons of
entrapped contamination and 1 /4 ton of
loose contaminants
On a 650-ton MSW day, if perfect
separation of heavy ferrous and loose
organics was accomplished by the
cleanup steps, the shipped material
would average 2,820 Ib/hr on an as
received basis. Of this, 350 Ib/hr would
be entrapped contamination (at 12.4
percent shipped material) This is well
above the market specification of 4
percent nonmagnetic material.
The project data resulted from the
average of test runs of 15 sec each at
nominal operating conditions The data
developed from these tests were not
extensive enough to differentiate among
the causes of measured variations in
runs. Some of these variables included
gauss strength, level of ferrous opportu-
nity, composition of feed, and relative
weight of specific materials.
Test No. 4.03, Shredded
Trommel-Overs Magnetic
Drum Separator
Ferrous metal recovery is also prac-
ticed on the trommel overflow. Material
that does not pass through the trommel
holes is conveyed to a shredder and
reduced to a nominal 90 percent by
weight less than 4 in It then passes J
under a rotary drum electromagnet. The*
ferrous removed is cleaned up by blow-
ing off loose organics and separating
the light from the heavy ferrous metals.
The magnetic drum separator used was
a Stearns rotary drum electromagnet,
72 in. long and 48 in wide. The shredder
is a Heil* shredder
The test on the trommel overflow
consisted of a number of 10-sec runs at
a nominal rate of 23.5 tph, of which the
ferrous component accounted for about
7 percent of the total weight The total
in-feed mass reporting to the trommel
overs is 37.7 percent.
Ferrous metal objects in the trommel
overflow feed are primarily in one of two
categories: tin-plated steel cans that are
too large to pass through the trommel
holes and smaller cans that fail to report
to the trommel overflow.
The design efficiency of the trommel
overs magnet is 95 percent Test runs
showed an average 27 percent effi-
ciency. If the data were projected on a
clean-metal basis, the magnetics in the
trommel overs would comprise 13.5
tons of ferrous available for recovery for
Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use \
-------
ieach 650 tons of MSW. At the measured
efficiency, 2.7 tons per day would be
recovered and 108 tons per day would
be lost. The principle reason for the low
recovery is that the air gap between the
trommel overs and the magnet belt is 48
percent greater than the design air gap
The efficiency of the trommel overs
magnet was also different for the com-
ponent categories. The magnet was
more efficient on the light gauge frac-
tion than on the heavy.
For every 100 Ib of magnetic trommel
overs product (on an "as received"
weight basis), approximately 61.5 Ib
was light gauge,4.5 Ibwascontammant
attached to light gauge, 33 5 Ib was
heavy gauge, and 0.5 Ibwascontammant
attached to heavy gauge.
Seventy-five percent of the recovered
magnetics (clean-weight basis) were
above 2 in. This size group contained 95
percent of the contaminant found in the
ferrous fraction. The particles greater
than 4 in were contaminated the most.
In all, contaminant accounted for 11
percent of the ferrous fraction. The
recovered magnetics contained 37
percent contamination, on an air dry
basis.
The samples taken from the trommel
overs feed and the product were tested
for moisture content, bulk density, size
distribution, residual magnetics, and
contamination. It was found that the
product is richer in bi-metal cans and
tin-plated steel cans than the feedstock.
Most composite can material in the
feedstock did not report to the product.
Noncan material is less highly concen-
trated in the product than in the feed-
stock.
Test No. 4.07, Hammermill
and Belt Addition
Testing was done on the trommel
overs and unders system modified by
the addition of a small hammermill so
that attached or entrapped contaminant
could be freed. A belt magnet was used
to separate metal from the liberated
contaminant.
The shredder has four rows of four
hammers each constructed of maganese
steel with hard facing The belt magnet
is a self-cleaning, permanent, magnet
type.
Several test runs were done on the
modified system: a standard run of 30
mm, a long-duration run of 60 min, and
surge runs from 10 to 60 sec with
surges in feed up to 5 tph The recovery
efficiencies varied from 95 to 98 percent.
Samples were taken of the waste at
various points in the hammermill process.
The samples were analyzed for bulk
density, particle size, magnetics, and
contamination. Hammer wear and cur-
rent and energy consumption were also
recorded.
During the standard run, the nonmag-
netic portion in the discard tailings was
67 percent whereas the percent con-
tamination in the magnetic product was
1.7 to 3.1 percent, well below the 4
percent criteria. In the long-duration
test runs on the hammermill, it appeared
that occasional large objects would not
cause difficulty. The specific energy was
8 98 hp/hr/ton and the power con-
sumption for the shredder was 32 hp In
the feed surge runs, only the 10-sec
runs were accomplished because the
motor current exceeded the nameplate
limit after 15 sec
At Recovery 1, a hydraulic ram crushes
the recovered light ferrous product to
increase bulk density before loading
onto a railcar. Data from the hammer-
mill runs indicated that the required
density was met, and in several cases,
compaction was unnecessary.
To meet the requirements for bulk
density, there had to be a trade off in
particle size Meeting the objective of
having only 5 percent (by weight) be less
than 1 in. proved impossible because
the magnetic concentration was typically
5 to 10 percent less than 1 in.
During the tuning runs, the extent to
which the hammermill product was
nuggetized was evaluated. An average
of 38 percent of the cans were nuggetized.
Results of the initial hammermill
tuning runs and the standard run showed
an observed hammer wear rate of 0.25
Ib/ton of magnetic concentrate passed.
The wear rate decreased significantly,
however, during the long duration test
when 4 additional tons of ferrous con-
centrate were shredded The overall
wear rate for the shredder using 6 tons
dropped to 0.06 Ib/ton
Several possible improvements in the
hammermill shredder design came to
light during testing Hard facing should
extend farther up the hammer to within
an inch of the pivot hole Shearing
action within the shredder might improve
if the breaker plates were modified to
contain "teeth." Additional equipment
could be included in the design: princi-
pally, an armored belt magnet (to sepa-
rate metal from dislodged contaminant)
and an air knife, which will protect the
shredder from dense objects
Results of this test of a shredder
followed by magnetic separation indicate
that the process effectively cleans and
densities ferrous metal without exces-
sive size reduction or nuggetizing
Product contamination is reduced to
less than 4 percent, density is increased
to the target figure of 21.5 to 26.0 Ib/ft3
after compaction; and particle size
remains large enough to be used by the
detmnmg industry The shredder per-
formed satisfactorily throughout testing
although the test duration was too brief
to determine a long-term hammer wear
rate.
Test No. 4.05 and 4.09, Air
Knife and Secondary Belt
Magnet Addition
For these tests, the hammermill and
the ferrous concentration system were
upgraded to alleviate the previous
experienced deficiencies and to meet
the criteria for the high ferrous product
maximum contamination, 4 percent,
bulk density, 21.5 to 26.0 Ib/ft3; and
shred particle size, maximum, 5 percent
less than 1 in The upgraded system
provides a ready adjustment of the drum
magnet by hydraulic activators, another
ferrous air knife, a light ferrous shredder,
and a secondary magnet. Since opera-
tion of the upgraded system began, the
new air knife adequately protected the
light ferrous shredder so that the ferrous
concentrator acts now only as a vibrating
conveyor The ferrous air knife was
capable of handling 15,000 ftVmin of
air. The principle performance specifi-
cations were to process 5 tph of 10 to 1 5
Ib/ft3 bulk density magnetic product
while accommodating peak loads of 6
tph of waste The air knife feed was to be
separated into three output streams
light metallic stream, light ferrous
stream, and heavy ferrous stream.
The secondary magnet used to clean
the shredded light ferrous is run on a 2
hp electric motor, and a 3K-W rectifier
energizes the electromagnet. The design
air gap is approximately 7 in.
Two processing lines, line #1 and line
#2, feed the upgraded system. Line #1
feeds the shredded trommel unders and
overs ferrous concentrate and line #2
processes scalped shredded refuse. The
estimated throughput for processing
line #2 and line #1 was 118 and 108
tph, respectively. The estimated overall
recovery for the target can ferrous metal
was 37 percent for line #1 and 48
percent for line #2. The estimated
efficiency for the adjustable air gap
-------
drum magnet was 62 percent for line #2
operation For line #1 and #2, respec-
tively, 52 percent and 64 percent of the
ferrous metal, as cans in the air knife
feed, was recovered as product.
In line #1 operation, 40 percent of the
target can ferrous metal (clean) was lost
to the air knife "heavies" stream; how-
ever, only 27 percent of the lost can
ferrous metal was "aerodynamically
light." This would bring this loss to 12
percent if the unseparable can metal
were excluded from the feed of heavies
Line #2 operation lost 31 percent of the
can metal to the heavies; 34 percent of
which was "aerodynamically light."
This brings the loss to 12 percent.
Of the can metal in the feed, 7 percent
was lost to the air knife "flies" or light
metallic stream in line #1 operation and
30 percent in line #2 operation. Only 1
percent of the can metal in the upgraded
system feed was lost from the second
magnet for both lines #1 and #2 opera-
tion. About half of the contamination
attached to the can ferrous metal in the
air shredder feed was liberated by its
processing. This reduced the attached
contamination from 7.0 to 3.7 percent
for line #1 and from 7.9 to 4.1 percent
for line #2. The contamination of the can
ferrous metal in the heavies was 9.0
percent for line #1 operation, whereas
for line #2 operation, it was 6 2 percent.
The secondary magnet of the upgraded
system carried over 9 2 and 5.4percent,
respectively, of the loose contamination
m line #1 and #2 feed. The product rate
was 1 63 tph at 26 7 Ib/ft3 and 1.34 tph
at 24 5 Ib/ft3. For both processing line
operations, the minimum required bulk
density for the product, 21 5 Ib/ft3, was
exceeded. The product composition for
the two lines was1
line #7, % line #2, %
loose
contamination
attached
contamination
can ferrous
noncan ferrous
ferrous less
than 1 in.
0.4
4.2
81.8
17.8
10.5
0.3
3.2
84.3
15.4
12.7
ferrous materials reporting to the lights
fraction. An increase in the blower
output with appropriate modifications of
the air knife interior diverters and
baffles may be the logical next step in
improving air knife performance.
In the ferrous stream samplings, four
sample sets were obtained under line
#2 operation and two under #1 opera-
tion Samples were collected from the
feed, "heavies," tailings product, and
light nonmetallics. Data for the upgraded
process was obtained on volume, bulk
density, metal weight, contamination,
and voltage and amperage used.
During this test period, 58.1 tons of
MSW were shredded. The shredder was
run for 7 7 hr under line #1 operating
conditions for ferrous recovery and 53 1
hr under line #2 conditions. For the
shredder in line #1, 24 hp were required
and the specific energy was 17 hp/hr/
ton. For line #2, 26 hp were required
and the specific energy was 15 hp/hr/
ton The total weight loss from the
hammers during the period was 1.20 Ib,
so the wear rate was then 0.021 Ib/ton.l
It was noted that hammer wear for the
four outboard hammers was substan-
tially higher than for all hammers.
During the 41.4 hr of line #2 operation,
one railcar was filled with 74,900 Ib of
product or a production rate of 0 90 tph.
For the secondary magnet, recovery
of ferrous metal was nearly 99 percent,
m spite of substantially larger-than-
design air gap
Conclusion
Ferrous metal recovery from munici-
pal solid waste is feasible at a large-
scale recovery plant However, at Re-
covery 1, the use of a trommel and
magnetic belts needed to be modified by
adding a shredder, air knives, and
additional belt magnets.
The full report was submitted in ful-
fillment of Contract No. 68-01-4423 by
the National Center for Research Re-
covery, Inc., Washington, DC, under the
sponsorship of the U S. Environmental
Protection Agency.
The performance of the air knife met
manufacturers' design specifications
only m the case of the loose, nonme-
tallic mass split reporting to the heavy
fraction. The most serious performance
deficiency was the mass split of can
Louis P. Soldano is with the Municipal Environmental Research Laboratory,
Cincinnati, OH 45268
Donald Oberacker and Carlton Wiles are the EPA Project Officers (see below).
This Project Summary covers the following reports, prepared by the National
Center for Resource Recovery, Inc, Washington, DC:
"Magnetic Drum Separator Performance Scalping Trommel Underflow at
Nominal Design Conditions Test No. 401, Recovery 1, New Orleans,"
(Order No PB 81-213 308. Cost- $8.00, subject to change!.
"Magnetic Drum Separator Performance Scalping Shredded Trommel
Overflow at Nominal Design Conditions, Test No. 4.O3, Recovery 1, New
Orleans. Louisiana," (Order No PB 81-213 316. Cost $800, subject to
change)
"Ferrous Metals Recovery at Recovery 1, New Orleans; Performance of the
Modified System. Test No 4.05 and Test No 409, Recovery 1, New
Orleans," (Order No PB 81 -213 324, Cost .$6 50. subject to change).
"Improvement of Magnetically Separated Ferrous Concentrate by Shred-
ding- A Performance Test- Test No 4.07, Recovery 1, New Orleans," (Order
No PB 81-213 332: Cosf $8 00, subject to change)
All the above reports are available only from.
National Technical Information Service
5285 Port Royal Road
Spring field, VA 22161
Telephone 703-487-4650
Donald Oberacker can be contacted at:
Industrial Environmental Research Laboratory
U S Environmental Protection Agency
Cincinnati, OH 45268
Carlton Wiles can be contacted at.
Municipal Environmental Research Laboratory
U S Environmental Protection Agency
Cincinnati, OH 45268
> US GOVERNMENT PRINTING OFFICE 19B1 -757-012/7264
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
0 0 0 0
c.- \i f
------- |