Determining the presence of organic compounds in foundry waste leachates.
A field study, discussion of laboratory and field data, as well as significant conclusions about organics in foundry wastes are presented in this review of AFS sponsored research.
Part one of this report on research sponsored by the American Foundrymen's Society (modern casting, July 1989, p 27) described a laboratory study to identify organic contaminants that may be released to the environment from ferrous foundry wastes after disposal in a landfill. A field study to monitor groundwater quality adjacent to foundry waste landfills was conducted to determine whether any organic contamination of groundwater could be attributed to leaching of organics from the waste.
Groundwater quality was measured at four ferrous foundry waste landfills, located in Wisconsin, which contained only foundry waste. At three of the four sites, there was enough background information from previous investigations to determine groundwater flow patterns for well placement. Flow directions for the fourth site were estimated using a U.S. Geological and Natural History survey, private well data and topographical relief observations.
Three downgradient wells were placed at the waste boundary of each site, adjacent to the oldest portion of the site whenever possible. This minimized the impact of a number of problems. It eliminated any channeling of pure leachate along the interstitial space between the well casing and the bore hole that may have occurred had the wells been placed through the fill material.
It also minimized the distance from the fill to the well screen. This reduces the effects of transport time, which tends to increase mixing within the water column, exaggerate vertical transport and increase the effects of attenuative processes. All of these tend to reduce the concentration of the organics reaching the wells, making detection more difficult. Finally, this placement method used the oldest portions of the sites, which allowed adequate travel time for any organic plumes to reach the wells.
An upgradient well was placed at each site far enough from the waste to ensure an accurate representation of background water quality, but close enough to intercept possible contamination from facilities directly upgradient of the site. Particular attention was paid to the effect of any groundwater mounding often found during certain times of the year beneath landfills.
Monitoring well depths were set at both the top few feet of groundwater and at a lower depth determined, when possible, from vertical gradients established from past data.
Wells are specified in all sites as follows: Wells numbered 1 are upgradient; 2 and 3 are downgradient; and those with a letter designation are nested, with A being the shallower and B the deeper.
Care was taken during well construction and groundwater sampling and analysis to use approved procedures. The wells were constructed of stainless steel; appropriate materials were used during construction; multiple bailings were used to prepare for each sampling; proper sampling methods were used; etc. Analysis consisted of multiple gas chromatograph-mass spectrometer scans on different extracts of the samples, followed by gas chromatograph confirmation of any compounds observed. This is in accordance with the laboratory study analytical procedures.
The initial field work was accomplished in 1986. As a follow-up, the monitoring wells at the foundry landfills were resampled in 1988. At three of the four landfills studied earlier, a sufficient number of the monitoring wells were intact and could be resampled. However, at one of these, the wells were all dry and no samples could be collected. The remaining two landfills are identified as A and D, consistent with the previous work.
Groundwater at the landfills was resampled in August 1988. Water levels, pH, specific conductance, chloride, fluoride and organics were measured. Previous foundry landfill studies showed that total dissolved solids (represented by specific conductance), chloride and fluoride often are elevated in leachates from foundry wastes. These parameters were used to determine whether the water sampled was actually impacted by the foundry landfill.
Initial sampling of groundwater around the four landfills in the spring of 1986 indicated no compounds above the reported detection limit of 1 ppb. However, several volatile compounds were tentatively identified at trace levels. All wells were sampled and analyzed for organics four times over a four month period.
At foundry A, napthalene was detected in one of four samples from well 2A. From well 2B, traces of tetrachloroethene and dichloroethene (isomer not specified) were found in one sample each and a trace of trichloroethane (again, the isomer was not specified) was found in three of the four samples.
At foundry B's landfill, no organics were identified in any samples from any of the wells.
In samples from the foundry C landfill, a trace of 1, 1, 1-trichloroethane was found in one sample from well 1 and traces of 1,2-dichloroethane and trichloroethane were found in one sample each from well 2A.
Finally, at foundry D, well 2A had a trace of chloroform in one of four samples and a trace of 1, 1, 1-trichloroethane in three of the four. A trace of 1, 1, 1-trichloroethane also was found in one sample from well 2B and in one sample from well 3.
Since very few semi-volatile compounds were found in the laboratory leach tests and none were found in the initial groundwater samples, only volatile analyses were performed on the groundwater samples from the foundry landfills in the resampling. A trace of dichloromethane was found in samples from wells 2A and 1 at landfill A and from well 2A at landfill D. These were only found in trace amounts and could have come from the laboratory air during analysis. Well 2A from landfill D also indicated a trace ([is less than] 1 ppb) of 1, 1, 1-trichloroethane.
The landfill A results of the groundwater monitoring for water levels and inorganic parameters at the foundry landfills are inconclusive. Neither the gradient nor the indicator parameters indicate the direction of groundwater movement, or whether it has been impacted by the foundry wastes.
At landfill D, however, the data strongly support the placement of the wells. Groundwater flow is from west (well 1) to east (wells 2A, 2B, 3). The shallow downgradient well has specific conductance and chloride levels, indicating that leachate from the landfill is passing that point.
Discussion of Results
Laboratory Study--An appropriate question to raise is whether those compounds and the concentrations of those compounds found in the waste leachates represent serious potential pollution problems. For purposes of discussion here, inclusion in one of EPA's lists will be assumed to indicate that the compound is of concern.
One such list is the Priority Pollutant List, which is meant to identify those compounds that are an environmental hazard and are present in water. It does not specify any concentrations. Excluding pesticides and PCBs, there are 88 priority pollutant organic compounds, of which ten were detected in the sample leachates.
Another list is to be used together with the proposed toxicity characteristic leach procedure (TCLP) to determine whether a waste is hazardous by the toxicity characteristic. The proposed list, excluding pesticides, contains 38 organic compounds and a regulatory level for each. The regulatory levels are the concentrations in the waste leachate at which the waste will be considered hazardous. EPA will continue to add compounds to this list as the information needed to propose regulatory levels is obtained. From the currently proposed list, six compounds were detected at measurable concentration in the foundry waste leachates.
The drinking water standards provide another list for comparison. Currently, six volatile organic compounds are identified in the standards, along with maximum contaminant levels (MCLs). MCLs are the maximum concentration of contaminants allowed in water used for drinking. This list is being expanded, as required under the 1986 Amendments to the Safe Drinking Water Act. From the current list, two compounds were measured in the foundry waste leachates.
One final set of lists is included with the proposed Solid Waste Disposal Facility Criteria. These regulations are proposed by EPA under authority of Subtitle D of the Resource Conservation and Recovery Act. The regulations are meant to control the release of hazardous constituents from nonhazardous waste disposal facilities. The proposed rules relate only to municipal solid waste landfills at this time, but EPA will be preparing similar regulations for industrial waste landfills (U.S. EPA, Aug 1988).
The lists that are included as appendices to the proposed rules are groundwater monitoring parameters and data for determining trigger levels for the parameters. Appendix I identifies the volatile organic compounds which must be included as part of Phase I groundwater monitoring program at a landfill site. If any of these compounds are found significantly above background as determined by statistical analysis, then a Phase II monitoring program must be implemented.
The parameters for this monitoring program are listed in Appendix II. For the compounds listed in Appendix II, trigger levels are set as the MCL, where one is established or is calculated based on the data provided in Appendix III. Of the compounds found at measurable concentration in the foundry waste leachates, six are included in Appendix I, 15 are in Appendix II and nine are in Appendix III.
Table 1 identifies the compounds found on any of the lists mentioned above and on which list(s) they are. The levels, whether proposed or final, are given where appropriate. The number of samples in which the compound was found at a quantifiable level and the maximum concentration observed in a composite sample complete the Table. For reference, Table 2 gives the concentrations of the listed contaminants in all of the composited samples.
As can be seen from Table 1, all listed compounds detected were at concentrations below one ppm and well below the TCLP regulatory level, where one is proposed. Therefore, none of the composite samples would be hazardous by characteristic based on the organic contaminants and levels proposed. As will be shown later, this is also the case for the volatile compounds in the individual waste streams tested.
The only compounds found at or above any of the levels given in Table 1 are benzene and tetrachloroethene. Benzene is above the MCL in the leachates from three binder systems. Tetrachloroethene is at the trigger level in one leachate when a risk factor of [10.sup.-5] is used to calculate the trigger level. These levels, however, are meant to be applied to drinking water or groundwater and not leachates.
Attenuation processes, such as volatilization, degradation or adsorption, occurring after leachate begins to migrate through the ground, will tend to reduce the concentration of the contaminants. It is, therefore, very unlikely that concentrations in the groundwater would reach any of these levels.
Though none of the leachates would present a great environmental threat according to present criteria, upon reviewing Table 2 it can be seen that two binder systems are more likely to be of concern than the others, both from the number of compounds leached and the concentrations of the compounds. These binder systems are represented by Samples 2 & 6, phenolic urethane and core oil, respectively.
On the other hand, all of the other samples appear to leach relatively little of the listed compounds, with the exception of benzene and toluene in Samples 4 & 5, the furan nobake and phenolic ester systems.
When individual waste streams were tested, the contribution of each component to the composite results can be analyzed. Table 3 gives the results of the volatile analyses on the individual waste components of four binder systems.
The core room sweepings contributed the most to the organic content of the composite leachates in the phenolic urethane and the phenolic isocyanate systems. Efforts to reduce organics may want to focus on this source. In the core oil systems, system sand contributed the most to organics in the leachate; and in the alkyd isocyanate system, the three sources of organic compounds seem to be about equal.
These results are reasonable given the system sand and core binder systems used in each case. The phenolic urethane and phenolic isocyanate foundries used a green sand system. This sand mixture is mostly inorganic in nature (sand, clay and water) with only a small amount of seacoal added. It does seem reasonable that little would leach from this sand.
The alkyd isocyanate foundry, on the other hand, also uses a synthetic organic binder (nobake) in the system sand. It would be expected to leach compounds and concentrations from the system sand similar to that leached from the core butts and core room wastes.
The core oil system leached the largest number of compounds at the highest concentrations from the system sand. Oil sand was analyzed as the system sand. This material was a sticky sand from which almost the entire coating was removed during the leach test. It is not surprising that it leached a large number of compounds.
In three of the four samples for which individual components were analyzed, the core butts leached fewer compounds and at lower concentrations than the corresponding core room wastes. Thus, it appears that simply using the core sand reduces the level of contaminants.
In the alkyd isocyanate system, this was not the case. This could be because the system sand is also a synthetic binder system and in separating core butts, the two become mixed. Another possibility is that contact with the hot metal degrades the binders in cores made from this resin and renders them more leachable. There are other possible explanations, not the least of which is the sampling error, which is discussed below.
The point of sampling could be the cause of the high concentrations in the leachate from the core room floor sweepings from the phenolic urethane core system. The core room waste sample was not actually waste, but rather freshly mixed sand and binder. This sample would contain unreacted resin since the sand had not been exposed to the gas which catalyzes the reaction of the binders to form rigid polymers. It also would most likely contain solvents needed to allow the resin flow.
The point of sampling can affect all of the results. Therefore, it should be noted that the system sand of the phenolic urethane and the phenolic isocyanate core systems is green sand, however, the system sand sample was not waste sand but rather new sand.
The seacoal is the most likely source of organics in this sand. If the sand, and therefore the seacoal, has not had contact with the hot metal, the leaching properties could be very different from those of waste sand. It is not known whether this would increase or decrease the number and concentration of leached compounds.
One final point to discuss regarding this study is the reliability of the results. The results of replicates indicated that there is some uncertainty in the leaching and analyzing of the waste samples. However, this error was not very large and the analyses should be expected to produce meaningful results.
One source of error, however, was not considered in that discussion. That is, the sampling error which occurred when the waste samples were collected at the foundry. The samples were intended to represent the waste streams leaving the foundry. However, the point of sampling, changes in processes, changes in waste handling techniques and the age of the waste when collected, can lead to variation in the composition of the wastes and the leachates produced from them.
To help determine how significant these variations were, waste samples were collected from two of the foundries at the beginning of the sampling period and again at the end. Both samples from each of these foundries were leached and analyzed as separate samples; however, only the first sample collected from each foundry was analyzed by GC/MS.
Results indicated, for the same foundry, that there are large differences in concentrations, as much as 300 ppb absolute difference for one compound and a 13-fold difference for another. There also are differences in the compounds present at quantifiable levels between samples from the same foundry. There is variability in the organic compounds in the solid wastes generated within a foundry. Care must be taken in obtaining, interpreting and applying organics leaching data.
Field Data--The results from both the initial sampling and the resampling of the groundwater monitoring wells are summarized in Table 4. The most significant result of the field study is that no detectable (quantifiable) concentrations of any organic compound in the 45,000 compound library searched was found in any well at any of the four landfills studied.
The organics searched are those listed by AFS at the start of the study as known to be in the various organic agents used in the industry or found by study elsewhere, plus organics from the Wisconsin State Laboratory of Hygiene Library, which includes those known to be of concern for groundwater contamination. The 45,000 compound library as applied to this study ignored naturally occurring organic compounds, such as humic acid and acetic acid. The lists used included priority pollutants, VOCs, carcinogenic compounds, etc.
The organic compounds listed in Table 4 were found at trace concentrations. These were at levels below quantitation limits for the State Lab's equipment; however, since the GC/MS unit was still able to match its scans with scans produced by the library used, they were identified but not quantified other than being present in trace amounts.
EPA has established drinking water criteria for VOC content and published these limitations in the Federal Register. Following their regulations, none of the compounds observed at trace levels were found at or above drinking water limitations.
Table 4 also lists the binder systems used by each foundry. A comparison of contaminants to binder systems is not as clear-cut for the groundwater monitoring as for the leach tests. The difficulty arises because the wastes going into the landfills come from the use of many different binders. Also, these binders were not always identified by the same name as in the present study. Foundry D, for example, uses six types of core binders along with a green sand molding system.
Of the compounds found in groundwater, naphthalene, tetrachloroethene and 1,1,1-trichloroethane also were found in the leach tests. Dichloroethene and dichloroethane could be degradation products of tetrachloroethene and trichloroethane and, in fact, were found only at sites where one of the latter also was found. Further, except for the single occurrences of naphthalene and chloroform, all other compounds found at trace levels could have been contaminants found in the air of the laboratory, as they are all common solvents used in organic analyses.
Consideration of the field and laboratory test results indicates that 1,1,1-trichloroethane could have leached from the waste into the groundwater. Landfill D did have trace levels of this compound in four out of five samples of well 2A, one out of four samples each for wells 2B and 3 and was not found in the upgradient well 1.
This compound was found in the leach tests of two binder systems as well. Based on this information and the elevated levels of the indicator parameters in well 2A in particular, indicating that this well was impacted by the landfill, it is possible that this compound is being leached from the foundry wastes into groundwater at very low but detectable trace concentrations.
The work on inorganic parameters in foundry waste leachates found very poor correlation between leach test results and field measurements. The leach tests did not accurately predict the concentrations of inorganic parameters in unsaturated or saturated zone water in the landfills or in groundwater under and downgradient from the landfill.
Therefore, it would be inappropriate in this study to claim to be able to accurately extrapolate the leach test results to predict what will happen in the field with regard to organic parameters. Groundwater monitoring is the best way to evaluate actual impacts on the environment.
The lack of detectable quantities of organic compounds in the downgradient well groundwater samples appears to be caused by the low levels of organic compounds leached from most wastes, as indicated by the laboratory study; the low percentage of waste materials in the landfill leaching these organics; and the attenuation and dilution effects as these materials travel to and in the groundwater to the wells.
This study does not provide an estimate of the importance of dilution or attenuation at these landfills, but these mechanisms have been well documented elsewhere and undoubtedly occur here as well. Groundwater at one of the sites was found to have been impacted by inorganic parameters arising from the landfill, so any contamination by organics should have been observable in these groundwater samples.
One other factor affecting the low organic concentrations not studied here is the timing and rate of release from the landfills. The rate could have been so low in itself as to result in nondetectable concentrations in the groundwater and it is possible, but not likely considering all of the results, that the timing of sampling did not coincide with any plume reaching the wells.
Foundry process solid wastes do leach a number of organic compounds. The number, type and concentrations of compounds varies from foundry to foundry according to the binder system(s) used at the foundry and when and where samples are collected.
The results of this study must be considered in light of the uncertainty as to whether the samples were representative of the waste streams as intended and whether the four landfills are representative of other sites. A further difficulty is the problem of extrapolating leach test results to field conditions. Subject to these limitations, the following conclusions are given.
1. The leach tests showed that none of the ferrous foundry wastes tested, either as composites or individual components, would be classified as hazardous wastes by EPA's proposed toxicity characteristic because of organic contamination.
2. The wastes are not likely to compromise groundwater quality due to the leaching of organics, as shown by the generally low levels (concentrations) of organics in the leachates and the virtual absence of organic compounds in the groundwater at foundry landfills sampled in this study.
3. None of the four ferrous foundry landfills investigated generated groundwater samples with detectable levels of any of the 45,000 organic compounds that the Wisconsin State Laboratory of Hygiene has in its GC/MS library.
4. Trace concentrations of certain organics did appear inconsistently. These compounds appeared at levels below quantitation limits for the State Lab's equipment.
5. Xylenes were the most common compounds leached in the laboratory procedure; some xylene was found in the leachate from all nine binder systems sampled. Toluene was present in seven of the nine binder systems sampled and naphthalene in six out of nine binder systems.
6. The compounds leached in the laboratory at the highest concentrations were benzoic acid, naphthalene, methylenebisphenol, diethylphenol and 3-methylbutanoic acid. All were present at over 300 ppb in one or more of the leachate samples.
7. The core oil system leached the largest number of compounds and, for the most part, also the highest concentrations. The phenolic urethane system also leached a large number of compounds at relatively high concentrations. The remaining binder systems leached relatively few compounds. In particular, the furan hotbox, alkyd isocyanate and furan warmbox samples leached very few organic compounds, none at concentrations over 100 ppb.
8. Green sand appears to be a low leaching molding sand. The organics in the composite wastes of a green sand system come from the core sands, primarily from the core room wastes.
9. Though these results indicate no environmental problems due to organics based on current and proposed regulations, it is still advisable that the foundry industry continue to stay abreast with and participate in changes in the regulations as it has in the past; to continue to study waste reduction/minimization processes, to look for the binder systems with less leaching potential; and to develop sand reclamation processes that address the leaching of organic compounds. [Tabular Data 1, 2, 3, 4, Omitted]
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||part 2|
|Date:||Aug 1, 1989|
|Previous Article:||Casting aluminum/ceramic composites at Progress Castings.|
|Next Article:||56th World Foundry Congress metalcasters examine international issues, technology.|