NEIC
EPA-330/1-97-001
IMPERMA^ENCE OF IRON TREATMENT OF
LEAD-CONTAMINATED FOUNDRY SAND
August 19, 1996
Douglas Kendall, Ph.D.
Senior Chemist
National Enforcement Investigations Center. Denver
U.S. Environmental Protection Agency
Office of Enforcemer
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Enforcement and Compliance Assurance
Office of Criminal Enforcement, Forensics and Training
EPA-330/1-97-001
IMPERMANENCE OF IRON TREATMENT OF
LEAD-CONTAMINATED FOUNDRY SAND
August 19,1996
Douglas Kendall, Ph.D.
Senior Chemist
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
Diana A. Love, Director
Denver, Colorado
-------
IMPERMANENCE OF IRON TREATMENT OF LEAD-
CONTAMINATED FOUNDRY SAND
The brass cast at foundries often contains lead, some of which comes
from recycled automobile radiators. Lead is present even in brass valves used
for drinking water. The foundry sand used to make molds for casting the brass
becomes contaminated with lead. The sand is used many times, but since new
sand must be continually added to maintain the properties of the sand, some
old sand must be removed from the process. The lead-contaminated waste
sand often fails the TCLP test and is a characteristic hazardous waste when
the lead concentration in the TCLP extract is over the regulatory limit of 5
mg/L.
Attempts have been made to render the lead-contaminated waste sand
non-hazardous by mixing it with metallic iron. Hazard is judged by the lead
concentration in a TCLP extract, which is the EPA test for toxicity under
RCRA. Whether materials containing well over 1000 mg/kg lead should be
considered nontoxic just because they pass the TCLP test is not a subject for
this report. This report explores the chemistry behind the iron treatment,
with emphasis on the permanency of the treatment. Materials which are
placed in a landfill for final disposal should continue to meet the standards
imposed on them before disposal.
The data and conclusions in this report are based on samples from the
NIBCO, Inc. foundry in Nacogdoches, Texas, and from NIBCO waste treated
with iron metal and placed in the Nacogdoches landfill. While the conclusions
are based on results from this one site, they are applicable to the many other
foundries using the same processes and generating similar wastes.
-------
The lead-contaminated waste sand fails the TCLP test. Mixing iron
metal with the sand (8% to 12% iron by weight is the NIBCO recipe, according
to their response to a Region 6 request for information under Section 3007 of
RCRA) causes the mixture to pass the TCLP test. NIBCO hoped that this
mixture could be disposed of in a municipal landfill. Indeed, waste sand mixed
with iron was placed in the Nacogdoches municipal landfill from 1987 to 1992.
On September 12 and 13,1995, samples of the material in the landfill were
collected by Region 6 contractors. On Sept. 13,1995, samples were collected
from the foundry itself. This report describes and interprets analyses and
experiments conducted on these samples at the NEIC laboratory. These
experiments help explain the chemistry behind the iron treatment and show
that the iron treatment will not permanently prevent the release of lead into
the environment.
ABSTRACT
1. Waste foundry sand contains high concentrations of lead, copper and zinc,
constituents of brass. Lead ranged from 1400 mg/kg to 5800 mg/kg in the
waste sand from the landfill. This material fails the TCLP test; the lead in the
TCLP extract is over 5 mg/L.
2. When iron metal is mixed with the waste sand, there is no reaction with
the lead; the lead is in no way altered or immobilized. Lead-contaminated
waste sand mixed with about ten percent iron will pass the TCLP test; the
lead in the TCLP extract will be less than the regulatory limit. Zero valent
iron will reduce lead (II) ions, removing them from solution. Copper will also
be reduced by iron metal, but not zinc.
-------
3. The key fact supporting the conclusions of this study is that waste sand
removed from the landfill after several years burial often fails the TCLP test.
This is due to the oxidation of iron so that zero valent iron is no longer
available to reduce the lead. Both lead and copper concentrations in the TCLP
extracts of this hazardous material indicate that metallic iron is no longer
active. If it is still present, it must be inacitvated by a coating of oxides.
4. Iron oxides, especially hydrous ferric oxide, can adsorb lead. Under some
circumstances enough lead can be adsorbed so that lead in the TCLP extract is
below 5 mg/L. However, adsorption alone is not a reliable method of
permanently immobilizing lead. Adsorption is too dependent on sample
history, pH, and other factors to be trusted for permanent immobilization.
There is no certainty that the material will continue to pass the TCLP test
with time. This contention is confirmed by TCLP tests of waste sand collected
from the Nacogdoches landfill.
5. The distinction between whether only adsorption is occurring or whether
oxidation/reduction is also occurring can be made by direct measurement of
redox potential or by examining copper concentrations in the TCLP extract.
Adsorption does not decrease copper concentrations in the TCLP extract nearly
as much as reduction.
SAMPLE COLLECTION
On September 12 and 13,1995, a team from A. T. Kearney, led by Bret
Kendrick, conducted sampling activities at NIBCO, Inc. and the Nacogdoches
Municipal Landfill in Nacogdoches, Texas. The EPA was represented by
Bobby Williams of Region 6 and Dr. Douglas Kendall of NEIC in Denver,
-------
Colorado. Dr. John Drexler of the University of Colorado was also present and
contributed to the sampling activities. The "Sampling Event Trip Report"
from A. T. Kearney contains complete details of the sampling process and of
the samples collected.
Samples were collected at the Nacogdoches municipal landfill from two
cells which contained waste sand and other wastes from the NIBCO foundry.
Samples were obtained by boring into the landfill with drilling equipment.
From each boring several samples were collected so as to preserve depth
information. Sample splits were sent to NEIC in Colorado, and separate splits
were sent to Environmental Science and Engineering, Inc., for TCLP and EP
toxicity tests. A set was given to NIBCO. After the samples were received in
Denver, a portion of most of them was placed in new containers for analysis at
NEIC. The remainder of each sample, in the original containers, was
transferred to Dr. Drexler in Boulder. Several samples were sent intact to Dr.
Drexler for specialized testing, so these samples were not analyzed at NEIC.
Two soil borings were done in cell A. The following samples were from
the first boring: NLF-A-01, -02, and -Core 1. The following samples were from
the second boring in Cell A: NLF-A-03, -04, and -05. In addition, two samples
were collected from ground level in cell A.
Three soil borings were done in cell 01. The following samples were
from the first boring in cell 01: NLF-01-01, -02, -03, -04, -05, -06, -Core 2. The
following samples were from the second boring: NLF-01-07, -08, -09, -10, and -
11. The following samples were from in third boring in cell 01: NLF-01-12, -
13, -14, -15, -16, -17, and -Core 3.
-------
Samples were also collected at the NIBCO brass foundry in
Nacogdoches on September 13, 1995. The materials sampled included waste
foundry sands from Unit 1 and Unit 2, green sand, hydrofilter sludge,
baghouse dust, resin sand and silica sand. These samples are designated with
"NIB" in the tables of analytical results. Full descriptions are in the A.T.
Kearney trip report.
On October 12, NEIC received a 5 gallon container filled with iron
shavings or turnings from Texas Foundries in Lufkin, Texas. This material is
similar to the iron used to treat the NIBCO waste sand placed in the
Nacogdoches landfill. This material is labeled as iron from Lufkin in NEIC
experiments shown in Table 3 and discussed later.
EXPERIMENTAL
Elemental Constituents
Table 1 presents the results of the analysis of aqua regia (hydrochloric
and nitric acids) digestions of most of the samples collected in Nacogdoches.
This digestion method is good for lead, iron, copper and zinc. The samples
were analyzed as received, with no extra drying. The concentrations of 26
elements are shown in units of mg/kg. The quality control results were
adequate for the purposes of this report. Results from the field duplicates
indicate that some of the material is not homogeneous, which is expected from
material containing pieces of metal. Substantial amount of lead, copper and
zinc are present in many of the samples, which is consistent with the samples
containing residues from the brass foundry work done at NIBCO. In the
landfill samples lead ranged from 1400 mg/kg to 5800 mg/kg, whereas iron
-------
ranged from 4.2% to 16.5%. The total lead and iron results are repeated in
Table 2 for comparison with the lead TCLP results.
TCLP Tests
TCLP tests were performed on most of the samples by Environmental
Science and Engineering. The results were provided to NEIC by Region 6.
The results for lead are included in Table 2 along with total lead and iron
results from Table 1. Some of the TCLP results for lead, including the results
for samples from the Nacogdoches landfill, are above the regulatory limit of 5
mg/L. Figure 1 is a graph of the lead concentration in the TCLP extract
plotted against the total iron for samples from the Nacogdoches landfill.
Treatment of a TCLP Extract with Different Forms of Iron
A series of experiments were conducted to better define the chemical
reactions which occur when lead-contaminated sand is mixed with different
forms of iron and then subjected to the TCLP test. The experiment separated
the TCLP test into two parts. Part one was the extraction of the lead-
containing sand. Part two was the treatment of the extract with iron or other
material. This separation into parts does not alter the conclusions as to the
role of iron in a standard TCLP test.
The first part of the experiment was to prepare a large volume of a
TCLP extract of the waste sand. A portion of a sample of waste sand collected
in a five gallon bucket from the Unit 1 foundry at NIBCO was used for the
-------
TCLP extraction. TCLP fluid number 1 was used (pH = 4.93 ± 0.05) for this
and all other TCLP extractions done at NEIC for this report. The standard
TCLP extraction is done with two liters of liquid. Four extractions were done
so that eight liters of extract were collected and available for further
experimentation. The first three entries in Table 3 are analyses of this extract
and were measured at various times as controls for the experiments described
below. The TCLP extract of the Unit 1 waste sand had high concentrations of
lead, copper and zinc, as shown in Table 3.
The TCLP extract was then treated with various materials. The
treatment was done by mixing the materials with the extract and tumbling for
18 hours, the time period of the TCLP test. Proportions were chosen so that
the ratio of solid used for treatment was consistent with the NIBCO recipe of
10% iron in the iron/waste mixture. After the tumbling, the solids were
removed by filtering and the resulting extract was analyzed by ICP.
Table 3 lists the various materials used for treatment. Untreated is the
TCLP extract described above. J.T. Baker ferric oxide is a reagent grade
chemical. The rust samples of 3/12 and 3/20 were prepared by mixing Lufkin
iron with TCLP extraction fluid and bubbling oxygen so that extensive
oxidation occurred. The rust was periodically collected by centrifugation and
dried. After drying it was used in the treatment experiments. The "baked
rust" is rust prepared as just described, and then heated in a muffle furnace
for 3 hours at 450 degrees Centigrade. The "iron from Lufkin" is iron shavings
or turnings from the Texas Foundries in Lufkin, Texas. Zinc metal, iron
filings and iron chips were laboratory chemicals. Rusted iron is iron from the
Lufkin foundry which was mixed with water and allowed to oxidize for several
-------
8
days. The experiments showed that not all the iron had been converted to
oxide, and that some metallic iron remained.
The results of the treatment experiments are in Table 3. The TCLP
extracts were analyzed by ICP-OES (inductively coupled plasma - optical
emission spectroscopy) for lead, copper, iron and zinc. For each of the
treatments, and for the untreated extract, the concentration of lead, copper,
iron and zinc in the extract is listed. The results for lead and copper are
shown in a bar graph, Figure 2, except that the effects of zinc treatment are
not shown. Zinc metal reduces the lead and copper levels to below the
detection limits and would not show in the graph.
TCLP Extractions of Material from Landfill
TCLP extractions were done on five samples (some of them composites)
of material from the Nacogdoches landfill, and they are listed in Table 4. It
was necessary to composite several samples in order to achieve the 100 g
minimum required for the TCLP test. Sample 07 from cell 01 was
supplemented from 01-11, which was a duplicate. Three samples from cell 1
were composited: 01 (55.33g), 04 (18.96 g), and 05 (27.01 g). For sample 10
from cell 1 the MSD split (unspiked) was mixed with the original sample.
Samples 04 and 05 from cell A were mixed 1:1. Sample 12 (53.67 g) from cell 1
was mixed with sample 16 (46.36 g).
The results are in Table 4. The results for lead and copper are plotted
in Figure 3. Lead concentrations in the extracts ranged from 2.81 mg/L to 11.6
mg/L. Three of the samples were over the regulatory limit of 5 mg/L. Copper
concentrations ranged from 0.830 mg/L to 164 mg/L.
-------
46 Hour TCLP Extractions with Measurement of Redox Potential
In order to obtain additional information on the relative importance of
lead reduction by iron metal and lead adsorption by hydrous ferric oxide,
additional TCLP extractions were done. The samples which were extracted
and the analytical results are listed in Table 5. The extractions were standard
TCLP extractions with two exceptions. The time period was extended to 46
hours, and the tumbling was periodically stopped so that redox potentials
could be measured.
The TCLP fluid was used as a blank. The second sample in Table 5 was
2 L of TCLP fluid with 10 grams of "rust" prepared from iron metal at NEIC.
The third sample was 2 L of TCLP fluid with 10 grams of iron metal. The
fourth sample was 100 grams of lead-contaminated waste sand from the Unit
1 foundry at NIBCO, the same material used in the treatment experiments
described above. The fifth sample was 90g of waste sand with lOg of the iron
metal from the Texas foundry. For the final extraction in Table 5 hydrous
ferric oxide (HFO) was prepared from ferric nitrate nonahydrate and IN
NaOH. Several hours after preparation, the HFO was collected by
centrifugation and then used in a TCLP extraction. The extract contained 90g
of waste sand and HFO prepared with enough ferric nitrate nonahydrate to
contain lOg of iron.
The redox potential was measured with a platinum electrode in
combination with a reference electrode. About every twelve hours the
tumbling was stopped, the redox potential was measured in all six samples
listed in Table 5, and tumbling resumed. Potentials were recorded after one
minute equilibration. While the values were not absolute measurements, the
-------
10
results were clear. The two samples with iron metal had potentials about 400
mV less than the other samples, which were all at about the same potential.
The samples with iron were more reducing than the samples without, and this
situation prevailed throughout the 46 hour period of extraction.
After 46 hours, the samples were filtered according to the TCLP
method, the pH of the extracts was measured, and the extracts were preserved
by reducing the pH to below 2 using nitric acid. The extracts were analyzed by
ICP - OES with the results shown in Table 5. There was some carryover from
the high level samples, so the detection limits have been adjusted accordingly.
DISCUSSION
There is no doubt that waste sand from brass foundries contains
substantial amounts of lead, copper and zinc. This is shown in Table 1, which
contains the results from the analysis of aqua regia digests of samples from
the NIBCO foundry and of foundry waste from the Nacogdoches, Texas,
landfill.
There is also no doubt that significant amounts of lead, copper and zinc
are leachable so that high concentrations of all three metals occur in TCLP
extracts of the waste foundry sand. This is shown by the data in Tables 2, 3, 4
and 5 for TCLP extracts of waste sand not mixed with iron or other materials.
The TCLP (toxicity characteristic leaching procedure) is the EPA test for
determining if a solid waste is a characteristic hazardous waste or if a solid
waste or a waste treatment residue can be land disposed. The regulatory limit
for lead is 5 mg/L. If the concentration of lead in the TCLP extract is over this
amount, then the waste is hazardous and the material can not be disposed of
-------
11
in a landfill. Copper and zinc are not regulated in hazardous wastes, but their
concentrations were measured and are discussed here in order to gain
additional insight into the chemistry of the foundry wastes, and because they
could have environmental impacts if leached from a waste.
While the TCLP extraction probably does not mimic exactly what
happened with the NIBCO wastes in the Nacogdoches landfill, it does provide
significant information on what did happened. The TCLP scenario assumes
mixing with municipal waste, while the NIBCO waste was buried separately
from other waste. Some of the waste from NIBCO, which presumably all
passed the TCLP test when it was buried, no longer passes the tests after
burial for a relatively short time. In particular, the amount of oxygen
reaching the wastes and the pH and volume of water contacting the wastes is
not know. Nevertheless, it is possible to reach some conclusions form the
results of TCLP tests about the wastes which have been buried.
The addition of about ten percent iron to the waste sand before TCLP
extraction changes the situation considerably. Consider the standard
reduction potentials of the metals of concern here: zinc, -0.76 volts; iron, -0.44
volts; lead, -0.13 volts; and copper, 0.34 volts. These are all for the
reduction of the +2 oxidation state to the metal. This means that iron can
reduce lead and copper from the +2 state to the metal, and zinc can reduce the
three other metals. These are the potentials under standard conditions and
must be adjusted for the ionic strength, pH, and acetate buffer of the TCLP
extraction or the landfill using the Nernst equation, if exact calculations are to
be done. None of these corrections will change the relative order of the metals,
including the conclusion that copper (II) will be reduced to a lower
concentration than lead (II) by iron when all are present in a solution.
-------
12
In Table 3 experiments numbered 8 through 12 all involved iron
treatment of a TCLP extract high in lead, copper and zinc. Lead and copper
are reduced to well below 5 mg/L. The fifth sample listed in Table 5 is also a
mixture of iron with waste sand and its TCLP extract (the time was extended
to 46 hours) is low in lead and copper.
The 46 hour extractions in Table 5 were stopped about every 12 hours,
and the redox potential was measured with a platinum electrode, as described
in the experimental section. These measurements showed that the added iron
metal was still active at the end of the TCLP extraction, even when extended
to 46 hours. Iron metal will reduce lead to below the regulatory limit if
present during a TCLP extraction. However, this does not mean it is a
permanent treatment as iron may not remain unoxidized.
There is another mechanism by which lead might be remover from
solution either in a TCLP extraction or in a landfill, and that is adsorption. As
evidenced by the TCLP results for untreated sand, the silica sand or other
constituents in the waste foundry sand don't adsorb enough lead from the
extract to reduce the amounts below the regulatory limit. On the other hand,
it is well known that iron oxides and hydroxides do adsorb metals from
solution.
Hydrous ferric oxide is the amorphous, hydrated ferric oxide which
forms first under most conditions when ferric ions precipitate as an oxide.
Information on hydrous ferric oxide (HFO) for this report was obtained from
the book by Dzombak and Morel. Adsorption by HFO is the most relevant type
of adsorption for waste foundry sand treated with iron. HFO will adsorb lead
(II), copper (II), and zinc (II) ions. They are adsorbed in that order, with lead
-------
13
most strongly adsorbed and zinc the least strongly adsorbed. This differs from
the situation for reduction, where copper is more easily reduced than lead.
The percent adsorption of each of these ions is a function of pH, and over a
fairly narrow pH range the percent adsorption can go from low to high.
However, this pH edge is not a constant for a particular ion, but strongly
depends on the ratio of metal ion to HFO binding sites. For the TCLP
situation it is prudent to look at experimental evidence rather than to try to
calculate the degree of adsorption, since the ionic strength is high and the
metal ion concentrations are high. Dzombak and Morel do not give all the
necessary adsorption parameters to do calculations for TCLP extracts of waste
foundry sand.
In the presence of oxygen, iron (III) is the most stable oxidation state of
iron. Thus when iron oxidizes in the presence of oxygen, iron (III) and HFO
can form. In the absence of oxygen, iron (II) is the most stable form of iron. In
the pH range of 5 to 6, ferrous ions are much more soluble than ferric ions, and
this is seen in the data. When TCLP extracts were treated with fully oxidized
iron, as in Table 3, or when a TCLP extraction was done with HFO, the iron
concentration in the extract was less than 1 mg/L. When the TCLP test is
done with iron treatment, iron concentrations in the extract reach the level of
several hundred mg/L, which is means significant amounts of ferrous iron
were present in solution.
The various experiments will now be considered in terms of adsorption.
Table 3 and Figure 2 contain the results of treating TCLP extracts with rust or
HFO prepared at NEIC from iron metal, and reagent ferric oxide (probably
crystalline). Ferric oxide adsorbs a small percentage of the lead and even less
copper and zinc. The rust samples adsorb from 80% to 90% of the lead, about
-------
14
30% of the copper and a few percent of the zinc. These experiments illustrate
that adsorption can under some circumstances decrease lead concentrations in
solution almost as much as reduction, but that for copper there is a significant
difference between the concentrations when reduction is possible and when
only adsorption is occurring.
The results shown in Table 5 involve TCLP extractions with iron metal
and with HFO. These show that when all of the iron in the 10% recipe is
converted to HFO, lead can be adsorbed to such an extent that the
concentration is the extract is significantly below the regulatory limit. Copper
concentrations are not decreased nearly as much by adsorption as they are by
reduction with metallic iron.
Table 4 shows results of TCLP extractions of waste sand collected from
the Nacogdoches landfill. Four of the five samples have lead concentrations in
the extracts over the regulatory limit. As shown in Table 1, these samples
have percentage levels of iron (total). It can be determined if zero valent iron
is still present to reduce the lead, or if adsorption is the predominant lead
control mechanism, by observing the copper concentrations. The four samples
with the highest lead levels have more copper than lead in the extract.
Sample 5 with the lowest lead concentration has an even lower copper level,
indicative of reduction occurring.
The first page of Table 2 contains results from samples removed from
the landfill in 1995 after being buried several years. The lead concentrations
in the extracts span a wide range, with many over 5 mg/L. The samples
contain major amounts of iron. Figure 1 shows the lead in the extracts plotted
against the total iron concentration; there is no obvious correlation. If there is
-------
15
a correlation, it must depend on the form of the iron. Since copper was not
measured in these samples, it is not possible to use the copper levels to see if
reduction is occurring. Since these samples presumably passed the TCLP test
when first buried, it is clear that whatever mechanism was decreasing the
extractable lead is no longer completely effective. There is no zero valent iron
present, since if it were lead would be reduced to well below 5 mg/L. There is
not enough HFO or any other material present to adsorb enough lead so that
the solution lead levels are below the regulatory limits. While laboratory
experiments with HFO show that it can adsorb lead, it is clear that in a
landfill there is no guarantee that HFO will form in such amounts and in such
a way to guarantee permanent prevention of lead leaching. Indeed, after only
a few years in the Nacogdoches landfill the iron treated waste foundry sand
had started to fail to TCLP test.
SUMMARY AND CONCLUSIONS
When metallic iron is mixed with lead-contaminated foundry waste
there is no reaction. The form of the lead does not change, and the lead is not
entrapped or immobilized. When the mixture comes in contact with an
aqueous solution, as it does during the TCLP test or in a landfill, lead will
start to leach into the solution. If metallic iron is present, then the lead
concentration in solution will be decreased by an oxidation/reduction reaction
to levels below the regulatory limit. If fresh metallic iron is regularly
introduced into the mixture, then soluble lead can be kept at low levels.
However, if the mixture is placed in a landfill and left alone, the iron will
oxidize, thereby losing its ability to reduce lead ions. The iron oxides, which
will sooner or later be the remnant of the metallic iron, can adsorb lead from
solution, but not reliably enough to conclude that leachable lead will never
-------
16
exceed the regulatory limit. This is amply demonstrated by the data from the
samples of actual waste removed from the Nacogdoches landfill, many of which
had lead concentrations in their TCLP extracts well over the limit. Iron
addition is NOT a way to permanently treat lead-contaminated waste.
-------
17
REFERENCES
1. D.A. Dzombak and F.M.M. Morel, Surface Complexation Modeling,
Hydrous Ferric Oxide, John Wiley and Sons, 1990.
2. Henry Freiser, Concepts and Calculations in Analytical Chemistry, CRC
Press, 1992.
3. H.A. Laitinen, Chemical Analysis, McGraw-Hill, 1960.
4. James F. Pankow, Aquatic Chemistry Concepts, Lewis Publishers, 1991.
5. R.V. Parish, The Metallic Elements, Longman, 1977.
-------
18
Concentrations in mg/Kg (wet wt.)
TABLE 1
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
Nacogdoches, Texas, Landfill (NLF)
Aqua-Regia Digestions
SAMPLE Cell & No.
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
NLF A- 01
bore 1
10" to 24"
3370.
53.
14.
24.9
38.
23.
23.2
2280.
167.
31800.
155000.
5800.
950.
812.
48.7
437.
280.
ND
7.2
120.
58.2
374.
2030.
106.
29.1
62200.
NLF A-02
bore 1
24" to 36"
7490.
28.
12.
33.5
29.
42.
12.2
5410.
92.8
25600.
99900.
5030.
1860.
399.
11.1
222.
750.
ND
6.5
400.
209.
1060.
1090.
146.
13.8
14800.
NLF A- 03
bore 2
4" to 24"
2150.
18.
14.
9.9
9.
11.
7.2
1280.
58.0
9690.
102000.
1420.
531.
378.
5.0
82.1
130.
ND
2.3
50.
47.8
190.
386.
61.8
9.4
6270.
NLF A- 04
bore 2
72" to 96"
2950.
36.
20.
16.7
28.
9.
9.2
1640.
94.8
26600.
112000.
2540.
691.
533.
14.6
255.
290.
ND
6.5
170.
62.1
349.
1100.
81.1
21.9
5730.
NLF A-05
bore 2
24" to 48"
5790.
20.
17.
19.7
12.
25.
10.1
3170.
75.2
9130.
120000.
2030.
1540.
573.
8.0
104.
570.
ND
1.7
420.
97.2
527.
395.
125.
25.3
10400.
NLF 01-01
bore 1 c)
30" to 42"
3420.
ND
ND
21.
20.
20.
6.
1100.
205.
7900.
151000.
1550.
390.
937.
37.
200.
ND
50.
ND
ND
9.0
280.
585.
82.
23.
12500.
NLF 01-02
bore 1 c)
60" to 120"
3880.
ND
ND
35.
ND
20.
5.
1300.
96.
13200.
75900.
2400.
600.
480.
18.
260.
400.
ND
3.
ND
12.8
200.
999.
98.
19.
16400.
LOD
2.
1.
3.
0.1
1.
1.
0.1
10.
0.1
0.9
2.
2.
2.
0.5
0.3
0.9
30.
4.
0.2
30.
0.05
9.
0.3
0.2
0.2
1.
a) LOD is the calculated limit of detection; ND means not detected.
c) LOD ten times higher due to analysis of diluted sample.
-------
19
Concentrations in mg/Kg (wet wt.)
TABLE 1 (cont.)
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
Nacogdoches, Texas, Landfill (NLF)
Aqua-Regia Digestions
SAMPLE Cell & No.
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
NLF 01-03
bore 1
120 to 180"
3850.
ND
ND
27.
ND
20.
5.
900.
90.
11300.
51900.
2120.
520.
284.
89.
200.
400.
ND
4.
ND
9.9
160.
926.
115.
28.
19200.
NLF 01-04
bore 1
180 to 240"
5370.
30.
ND
50.
ND
ND
4.
1500.
185.
7940.
109000.
1630.
750.
705.
46.
181.
500.
60.
3.
ND
14.2
260.
621.
127.
36.
10900.
NLF 01-05
bore 1
duplicate
of NLF 01-01
2780.
ND
ND
21.
ND
ND
6.
1200.
248.
8360.
165000.
1470.
430.
994.
47.
209.
ND
50.
ND
ND
9.2
280.
595.
78.
27.
11700.
NLF 01-07
bore 2
24" to 60"
3350.
ND
ND
25.
ND
ND
6.
1000.
122.
11100.
88100.
2200.
500.
516.
20.
234.
ND
ND
ND
ND
11.3
240.
849.
80.
23.
17300.
NLF 01-08
bore 2
60" to 120"
3810.
20.
ND
30.
ND
ND
6.
1300.
91.
13400.
65300.
2630.
630.
406.
9.
249.
400.
ND
4.
ND
13.3
230.
1050.
89.
20.
19200.
NLF 01-09
bore 2
120 to 180"
2460.
ND
ND
14.
ND
ND
3.
1000.
137.
12600.
84100.
1460.
350.
478.
20.
248.
ND
ND
3.
ND
8.2
140.
757.
74.
15.
6360.
NLF 01-10
bore 2
180 to 240"
6410.
ND
ND
50.
ND
ND
5.
1700.
104.
11700.
77900.
2430.
1500.
469.
22.
210.
700.
ND
4.
ND
21.1
310.
954.
107.
41.
14200.
LOD
20.
10.
30.
1.
10.
10.
1.
100.
1.
9.
20.
20.
20.
5.
3.
9.
300.
40.
2.
300.
0.5
90.
3.
2.
2.
10.
a) LOD is the calculated limit of detection; ND means not detected.
-------
20
Concentrations in rag/Kg (wet wt.)
TABLE 1 (cont.)
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
Nacogdoches, Texas, Landfill (NLF)
Aqua-Regia Digestions
SAMPLE Cell & No.
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
NLF 01-11
bore 2
duplicate
of NIB 01-07
3390.
20.
ND
27.
20.
ND
6.
1100.
173.
10900.
106000.
2100.
530.
727
28
292.
400.
ND
3.
ND
12.3
240.
862.
83.
21.
17000.
NLF 01-12
bore 3
42" to 60"
2870.
ND
ND
24.
ND
ND
6.
1000.
105.
12200.
90600.
2170.
460.
496
17
220
ND
ND
ND
ND
11.8
200.
847.
71.
18.
18400.
NLF 01-13
bore 3
108 to 120"
4600.
ND
ND
30.
ND
ND
5.
900.
57.
16800.
40300.
2330.
560.
246.
9.
221.
600.
ND
4.
ND
10.6
140.
1060.
99.
21.
18300.
NLF 01-14
bore 3
120 to 180"
3290.
ND
ND
24.
ND
ND
4.
900.
80.
14100.
55400.
2080.
450.
323.
18.
201.
ND
ND
4.
ND
9.7
150.
935.
82.
18.
11900.
NLF 01-15
bore 3
180 to 240"
4240.
20.
ND
35.
ND
ND
4.
1300.
145.
12900.
77300.
2290.
660.
482.
111.
210.
500.
ND
3.
ND
13.6
240.
1030.
101.
37.
14300.
NLF 01-16
bore 3
duplicate of
NLF 01-14
2640.
ND
ND
22.
ND
ND
4.
900.
59.
11600.
42200.
1910.
400.
259.
12.
191.
ND
ND
4.
ND
9.4
130.
811.
54.
14.
11900.
LOD
20.
10.
30.
1.
10.
10.
1.
100.
1.
9.
20.
20.
20.
5.
3.
9.
300.
40.
2.
300.
0.5
90.
3.
2.
2.
10.
a) LOD is the calculated limit of detection; ND means not detected.
-------
21
Concentrations in mg/Kg (wet wt.)
TABLE 1 (cont.)
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
NIBCO - Nacogdoches, TX
Aqua-Regia Digestions
SAMPLE NUMBER
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
NIB 01-S
Unit 1
foundry
sand
3600.
ND
ND
10.
ND
ND
5.
2500.
5.
18800.
3890.
1670.
980.
45.
ND
371.
400.
ND
4.
ND
74.6
450.
1030.
88.
ND
8900.
NIB 02-S
Unit 2
foundry
sand
1740.
ND
ND
7.
ND
ND
5.
1700.
ND
24800.
1620.
2650.
510.
21.
ND
130.
ND
ND
9.
500.
55.5
400.
897.
39.
ND
10200.
NIB 03-S
Unit 1
foundry
sand
2680.
40.
ND
9.
20.
ND
6.
2400.
4.
54900.
2840.
3140.
780.
39.
ND
492.
ND
ND
14.
ND
67.8
450.
3050.
65.
ND
9670.
NIB 05-S
Unit 2
ball mill
sand
780.
ND
ND
3.
ND
ND
3.
600.
3.
17200.
2970.
1120.
200.
21.
ND
110.
ND
ND
6.
ND
16.8
100.
921.
21.
ND
3200.
NIB 06-S
Unit 1
waste sand
2920.
ND
ND
8.
ND
50.
7.
2200.
3.
11300.
4170.
1770.
810.
48.
ND
103.
ND
ND
6.
400.
55.1
390.
509.
75.
ND
10100.
NIB 07-S
Unit 1
duplicate
of NIB 06-S
2910.
ND
ND
8.
ND
50.
6.
2100.
3.
8010.
4960.
1570.
730.
40.
ND
56.
ND
ND
3.
300.
52.4
350.
404.
76.
ND
9100.
LOD
20.
10.
30.
1.
10.
10.
1.
100.
1.
9.
20.
20.
20.
5.
3.
9.
300.
40.
2.
300.
0.5
90.
3.
2.
2.
10.
a) LOD is the calculated limit of detection; ND means not detected.
-------
22
TABLE 1 (cont.)
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
NIBCO - Nacogdoches, TX
Concentrations in mg/Kg (wet wt.)
Aqua-Regia Digestions
SAMPLE NUMBER
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
NIB 08-R
post-casting
resin sand
55.
ND
3.
0.2
ND
2.
0.7
200.
ND
14.7
176.
13.
13.
1.1
ND
ND
ND
ND
ND
ND
0.98
41.
0.9
1.2
ND
96.
NIB 09-R
pre-casting
resin sand
41.
ND
ND
ND
ND
2.
ND
120.
0.2
ND
104.
ND
8.
ND
ND
ND
ND
ND
ND
ND
0.47
18.
ND
0.6
ND
2.
NIB 10-R
duplicate
of NIB 08-R
294.
ND
ND
1.7
ND
ND
1.6
350.
ND
1200.
371.
208.
77.
3.1
ND
6.1
40.
ND
0.5
50.
6.32
73.
51.6
9.0
0.3
1060.
NIB 11-D
baghouse
dust c)
864.
ND
ND
3.7
109.
37.
1650.
800.
ND
1880.
1650.
49500.
236.
12.2
ND
14.8
310.
ND
2.5
ND
20.2
1820.
369.
24.0
3.0
571000.
NIB 16-S
Unit 2 c)
waste sand
2830.
ND
ND
7.6
23.
ND
6.4
1850.
ND
35100.
2580.
4020.
725.
24.9
ND
190.
ND
ND
12.7
450.
55.3
392.
1590.
67.9
2.4
12800.
LOD
2.
1.
3.
0.1
1.
1.
0.1
10.
0.1
0.9
2.
2.
2.
0.5
0.3
0.9
30.
4.
0.2
30.
0.05
9.
0.3
0.2
0.2
1.
a) LOD is the calculated limit of detection; ND means not detected.
c) LOD ten times higher due to analysis of diluted sample.
-------
23
TABLE 1 (cont.)
ELEMENTAL CONSTITUENT ANALYSIS BY ICAP
NIBCO - Nacogdoches, TX
Concentrations in mg/Kg (wet wt.) Aqua-Regia Digestions
SAMPLE NAME
ELEMENT
Aluminum
Antimony
Arsenic
Barium
Bismuth
Boron
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Sulfur
Tin
Titanium
Vanadium
Zinc
Green Sand
Unit 2
2390.
ND
NO
7.2
ND
ND
5.2
1690.
ND
15600.
2070.
2200.
642.
22.0
ND
83.0
ND
ND
5.1
520.
54.5
386.
666.
55.6
ND
9250.
Unit 1 c)
duplicate
of NIB 06-S
2830.
ND
ND
9.5
ND
28.
7.2
2210.
2.2
8050.
4050.
1750.
763.
36.5
ND
53.6
ND
ND
3.8
330.
54.3
425.
392.
70.9
ND
9640.
Unit 2 c)
duplicate
of NIB 05-S
3010.
ND
ND
8.8
20.
ND
5.4
1940.
44.8
17000.
50800.
1880.
835.
394.
10.4
716.
ND
ND
4.0
230.
59.6
367.
782.
80.3
5.1
7240.
New Sand
silica
sand
47.
ND
ND
0.3
1.
ND
ND
120.
ND
1.1
224.
ND
64.
1.5
ND
ND
ND
ND
ND
ND
0.26
ND
ND
1.9
ND
4.
New Core
Sand
resin sand
36.
ND
ND
0.2
ND
ND
ND
190.
ND
ND
90.
ND
7.
0.7
ND
ND
ND
ND
ND
ND
0.64
17.
0.4
0.5
ND
3.
LOD
2.
1.
3.
0.1
1.
1.
0.1
10.
0.1
0.9
2.
2.
2.
0.5
0.3
0.9
30.
4.
0.2
30.
0.05
9.
0.3
0.2
0.2
1.
a) LOD is the calculated limit of detection; ND means not detected.
c) LOD ten times higher due to analysis of diluted sample.
-------
24
TABLE 2
Lead and Iron Measurements
Samples from Nacogdoches Landfill (NLF)
Sample
NLF A-01
NLF A- 02
NLF A- 03
NLF A- 04
NLF A- 05
NLF A02-G
NLF 01-01
NLF 01-02
NLF 01-03
NLF 01-04
NLF 01-05
NLF 01-06
NLF 01-07
NLF 01-08
NLF 01-09
NLF 01-10
NLF 01-11
NLF 01-12
NLF 01-13
NLF 01-14
NLF 01-15
NLF 01-16
NLF 01-17
Total Lead
(mg/kg)
5800.
5030.
1420.
2540.
2030.
NA
1550.
2400.
2120.
1630.
1470.
NA
2200.
2630.
1460.
2430.
2100.
2170.
2330.
2080.
2290.
1910.
NA
TCLP Lead
(mg/L)
0.52
1.6
4.9
8.2
4.7
220.
5.9
<0.22
0.75
2.4
5.7
65.
11.
1.6
1.3
5.7
11.
3.
2.8
2.4
1.5
3.3
45.
Total Iron
(mg/kg)
155000.
99900.
102000.
112000.
120000.
NA
151000.
75900.
51900.
109000.
165000.
NA
88100.
65300.
84100.
77900.
106000.
90600.
40300.
55400.
77300.
42200.
NA
NA = Not Analyzed; TCLP results from Environmental Science
and Engineering, Inc.; Total results from NEIC aqua regia
digestions.
-------
25
TABLE 2 (cont.)
Lead and Iron Measurements
Samples from NIBCO, Inc.
Sample
NIB 01-S
NIB 02-S
NIB 03-S
NIB 05-S
NIB 06-S
NIB 07-S
NIB 08-R
NIB 09-R
NIB 10-R
NIB 11-D
NIB 12-D
NIB 13-D
NIB 16-S
GREEN SAND
UNIT 1 BUCKET
UNIT 2 BUCKET
VIRGIN SAND
CORE SAND
Total Lead
1670.
2650.
3140.
1120.
1770.
1570.
13.
<2.
208.
49500.
NA
NA
4020.
2200.
1750.
1880.
<2.
<2.
TCLP Lead
23.
66.
27.
21.
43.
45.
3.6
0.22
9.3
790.
78.
110.
61.
NA
NA
NA
NA
NA
Total Iron
3890.
1620.
2840.
2970.
4170.
4960.
176.
104.
371.
1650.
NA
NA
2580.
2070.
4050.
50800.
224.
90.
NA = Not Analyzed; TCLP results from Environmental Science
and Engineering, Inc.; Total results from NEIC aqua regia
digestions.
-------
TABLE 3
Treatment of TCLP Extract from Unit 1
Metal Concentrations after Treatment
Concentrations in mg/L
26
Experiment
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
Sample &
Date Treated
Untreated
1/9
Untreated
3/12
Untreated
3/20
ferric oxide
1/9
Baked rust
3/20
Rust 3'/12
Rust 3/20
Iron Chips
1/9
Rusted iron
1/9
Iron from
Lufkin 3/20
Iron from
Lufkin 1/9
Iron Filings
1/9
Zinc Metal
1/9
Lead
(mg/L)
52.1
45.8
47.6
42.7
8.43
4.5
3.49
ND
ND
1.30
0.61
ND
ND
Copper
(mg/L)
82.9
80.3
78.3
77.0
58.4
55.7
49.8
ND
ND
ND
ND
ND
ND
Iron
(mg/L)
4.21
0.96
0.95
ND
ND
ND
ND
296.
485.
436.
513.
638.
4.09
Zinc
(mg/L)
171.
164.
148.
151.
144.
158.
143.
147.
142.
143.
157.
144.
1020.
ND = Not Detected; Detection limits varied due to carryover.
-------
27
TABLE 4
TCLP of Samples from Nacogdoches Landfill
Concentrations in mg/L
Sample
No.
1
2
3
4
5
Sample
07
(Cell 01)
composite
of
01, 04, 05
(Cell 01)
10
(Cell 01)
Composite
of
04 & 05
(Cell A)
Composite
of
12 & 16
(Cell 01)
Lead
(mg/L)
10.8
4.23
11.6
5.27
2.81
Copper
(mg/L)
164.
36.5
12.9
9.89
0.830
Iron
(mg/L)
14.5
118.
212.
184.
489.
Zinc
(mg/L)
146.
75.0
117.
69.4
113.
Final pH
5.19
5.12
5.29
5.28
5.52
TABLE 5
46 Hour TCLP Extraction of Waste Sand from Unit 1
Concentrations in mg/L
Sample
Blank
TCLP fluid 1
Fluid 1 with
ferric oxide
prepared
from iron
metal
Fluid 1 with
Lufkin iron
metal
NIBCO waste
sand
(Unit 1)
NIBCO waste
sand with
10% iron
metal
NIBCO waste
sand with
10% iron as
hydrous
ferric oxide
Lead
(mg/L)
<0.5
<0.5
<0.5
46.7
<0.5
<0.5
Copper
(mg/L)
<0 5
1.60
<0.5
119.
<0 5
12.1
Iron
(mg/L)
<0.5
0.50
>600.
0.76
558.
<0.5
Zinc
(mg/L)
<0.5
<0.5
<0.5
156.1
118.
100.
Final
pH
4.91
4.89
5.95
5.20
6.01
5.51
-------
12
10 -
8 -
ts
UJ
Q.
P 4 H
CO
2 -
0 -
0
Figure 1 TCLP Lead vs. Total Iron
Samples from Nacogdoches Landfill
i
2
6 8 10 12 14
Total Iron (percent)
16 18
20
-------
Figure 2 Treatment of TCLP Extract
100
Experiment Number
-------
Figure 3 Lead and Copper in TCLP Extracts
180
160 -
140 -
^ 120 -
l3>
100
2
+•<
x
LJJ
c
o
80 -
re
£ 60
o
u
c
o
O 40
20 -
0
0
2
Copper
Lead
mma
3
5
6
Sample Number
------- |