United States Environmental Protection Agency Risk Reduction Engineering Laboratory Cincinnati OH 45268 Research and Development EPA/600/S2-88/066 Mar. 1989 x°/EPA Project Summary Proposed Test Protocol to Determine Toxicant Leaching into Potable Water Ronald Rossi, Craig R. Turner, and Dipak K. Basu A simple apparatus was con- structed from Teflon* to test teachability of contaminants. Plates coated with coal-tar-based mate- rial were placed In the Teflon test chamber of the apparatus, and water of controlled parameters was con- tinuously passed through the chamber at a velocity of 2 L/min for a period of 24 hr. The test apparatus was unique in its ability to perform an accelerated leaching test under flowing water conditions. These tests showed migrations of various component polynuclear aromatic hydrocarbons (PNA) with no detectable aging effect in three successive 24-hr leachates in results that fell in the range of 30 to 400 pg/L. The possibility of PNA migration from the coal-tar-based material lining pipes in the field was tested at three utilities. No quantitative corre- lation could be established between the leaching observed under labora- tory and field tests. However, the leaching pattern from the field pipes can be quantitatively explained from the results of laboratory leaching tests. This Project Summary was devel- oped by EPA's Risk Reduction Engi- neering Laboratory, Cincinnati, OH, to announce key findings of the research project that Is fully docu- * Mention of trade names or commercial products does not constitute endorsement or recom- mendation for use. mented In a separate report of the same title (see Project Report ordering information at back). Introduction Potable water used by large segments of the U.S. population is exposed to direct and indirect additives. The direct additives are chemicals that are deliberately added during the treatment of raw water for coagulation, softening, corrosion control, disinfection, fluorida- tion, and other purposes. As a result, finished water may contain both intended and unintended residuals from direct additives. Indirect additives are defined as contaminants that are inadvertently introduced into the potable water through paints, coatings, liners, sealants, pumps, and other items used during storage and distribution of potable water to con- sumers. The change in potable water quality as a result of the presence of direct and indirect additives necessitates an evaluation of the possible health hazards arising from these additives. Through a memorandum of under- standing signed by the U.S. Food and Drug Administration and the U.S. Environmental Protection Agency (EPA) in 1979, the responsibility for monitoring and controlling these additives was vested in EPA (44FR42775, July 20, 1979). As one of the initial steps towards meeting this responsibility, EPA, in cooperation with the National Research Council, produced a Water Treatment Chemicals Codex for direct additives only. The Codex recommends a mini- mum acceptable purity specification as it ------- related to health for about 25 commonly present additives in potable water. Indirect additives, on the other hand, have been monitored through a voluntary program and through the issuance of advisory opinions by EPA or its predecessor agencies. Obstacles to the development of a safety evaluation program for indirect additives include a lack of established maximum contami- nant levels in several instances, absence of suitable laboratory simulation data and field studies, and the lack of a general consensus regarding the target param- eters to be monitored as indicators of indirect additives. The present investi- gation was undertaken to provide support for Codex development methodology for indirect additives. The purpose of this research was (1) to develop a compre- hensive and realistic laboratory test protocol that would simulate contaminant migration from coal-tar-based mate- rials similar to those that line potable water pipes and tanks in the field, and (2) to correlate the laboratory test results with actual field studies. Coal-tar- based materials were used to represent sources of secondary additives for developing the test protocol because leachates from these materials are known to contain a large number of compounds, some of which are sus- pected to be carcinogenic. Experimental Laboratory Apparatus The laboratory leaching apparatus was fabricated and assembled inhouse. It consisted of four parts: (1) a leaching compartment of Teflon having a well 20-1/2 by 1-1/2 by 9/20 in. in the midsection and a removable Teflon top plate for the insertion of test plates; (2) Teflon connecting fittings and tubes; (3) a variable speed circulating pump with all Teflon wetable parts; and (4) a flowmeter. A diagram of the laboratory apparatus is shown in Figure 1. Test plates were fabricated from a stainless steel sheet 0.05 in. thick. Each plate had a dimension of 20 x 1.5 in. These plates were sand-blasted before being coated. Each plate was cleansed with purified deionized distilled water (DDW) and acetone and was dried before the application of the coating material. A one-coat Bitumastic Super Tank Solution (Type I) was used in the experiments as a coating system. The test plates were coated with the suction-feed spraying system at a delivery pressure adjusted to about 60 psi at the gun, as recommended by the manufacturer of the coating material. Two coats of the coating material were evenly applied on both sides of the test plates. The test plates were air dried for 10 days and those plates with total dry thickness ranging from 0.040 to 0.060 in. were selected for further leaching tests. The leaching test with coated plates began by pre-exposing the plates in a solution of sodium hypochlorite con- taining 50 ppm of free chlorine at a pH of 10.5. The plates were allowed to stand in this solution overnight and were subsequently washed with purified DDW. Two plates were placed inside the leaching compartment, and it was sealed. The compartment was filled with test water of controlled parameters, and the flow rate of the circulating water was maintained at the desired value. The entire leaching apparatus was transferred to an environmental chamber where the temperature at which the test was conducted could be controlled. It was experimentally determined that for a flow rate of 2 L/min, the temperature of the environmental chamber had to be set at 12.8°C for the circulating water to attain an equilibrium temperature of 21.5°C and at 25.6 °C to attain a temperature of 28.5°C. The system was allowed to for the desired length of time. At the end of the run, all of the w from the leaching apparatus was drai into a measuring cylinder. An aliquc the water ( = 70 mL) was kept sepa for performing alkalinity, hardness, and residual chlorine tests. The resii leachate was solvent extracted \ methylene chloride (6 mL of methyl chloride for every 100 mL of water). extract was concentrated to a f volume of 1 mL using a Kuderna-Dai apparatus and subsequent blowdi using prepurified N2 gas at 30 °C. Field Sampler The field sampler consisted of following three main components: (1) containing a two-stage resin coli system; (2) variable water pumping consisting of a Masterflex pump i appropriate pumpheads and f controller; and (3) flowmeter. Chroma extender-type columns of 150 x 25 x mm were used to hold the resin bed. resin bed consisted of equal volumes XAD-2 and XE-348 resins separa by glass wool. The length of the s< resin bed was set at 13 cm. The end the Chromaflex columns were plug Flowmeter Pyrex®/Teflon® _ Ajf B,egd Stopcocks,{. Teflon® Tubing Teflon® Teflon® . Teflon® Male Fitting Male Fitting Steel Plate Leaching Compartment Figure 1. Schematic diagram of leaching apparatus (not to scale). ------- with clean glass wool and were connected to two tapered column adaptors by "0" rings and clamps. Two such resin columns were connected in parallel to the water to be sampled through a Pyrex glass Y-tube. The other end of the column was connected individually to a variable Masterflex pump. The pump units with their flow controllers allowed water to pass through the resin bed at the desired rate. The outlets of the pumps were connected to calibrated flowmeters to measure the flow rate that was maintained at 40 mL/min. The effluent water was collected in calibrated collapsible plastic carboys. Measurement of the total water collected in each carboy over a known period of time permitted the gross water flow rate through the resin beds to be estimated. All connections along the different components were made with the custom-made 8-mm Pyrex glass tub- ing of convenient shape and length, and minimum lengths of Tygon tubing were used for interconnecting the Pyrex glass tubing. Field Sampling Analyses Potable water samples from three water supply systems on the West Coast of the United States were analyzed as ield samples. The rationale for selecting these systems was that they represented three large public water utilities and all contained transmission pipes lined with coal-tar-based material. The expected occurrence of secondary additives in this water make it well suited to test whether or not the designed laboratory leaching apparatus, upon which the proposed test protocol is based, could be successfully used. At each water supply, water samples were collected at three points. The first point was the water at the treatment plant before it was exposed to lined trans- mission pipes or tanks. The second and third sampling points were near the beginning and near the end of a transmission system that had pipes lined with coal-tar-based materials. The sampling unit was transported to each sampling location by packing the individual components in suitcases. For the convenience of transportation, the Masterflex pumps with the flow- controllers and pumpheads were carried in a separate suitcase provided by the manufacturer. A total of 20 L of water was collected from each sampling point. At the end of the sampling, the ends of 'he resin columns were sealed with .jolyurethane foam plugs, parafilm, and masking tape. The columns were wrapped in aluminum foil, cooled with non-wetable ice packs, and then transported to the laboratory. In the laboratory the resin beds were warmed and separated, and only the XAD-2 portions were spiked with a known amount of C14-fluorene. Only the XAD-2 resins were subjected to the elution method. The XE-348 beds containing more polar compounds were not further analyzed. The XAD-2 resin bed was washed with about 20 to 25 mL of acetone. The vacuum from an aspirator removed excess acetone from the resin bed. The dry XAD-2 bed was removed to a clean thimble, and the resin was Soxhlet- extracted for 24 hr with methylene chloride. The acetone wash was diluted with purified DDW and extracted with methylene chloride. The two methylene chloride layers were combined and concentrated to 1 mL for radioactive counting and analysis using a gas chromatography-mass spectrometry- data system combination (GC-MS- DS). Before the radioactive counting, a solvent exchange of methylene chloride to toluene was performed on 500 pL of the above concentrated extract. The collection efficiency of seven PNA's with this sampling apparatus averaged 84% in the laboratory. To increase the sensitivity of the GC- MS-DS analysis, several modifications were made in the original system. The injection port of the original packed column GC was replaced according to the manufacturer's instructions with an on-column (capillary) injection system. The interface of the exit end of the column with the MS was also altered; the jet separator and its associated accessories were completely removed and the exit end of the capillary column was introduced directly into the ion source. The conditions used for the operation of the GC-MS were as follows: Analyzer temperature: MS delay: Scan: Scan time: 200°C 4 min following sample injection 50 to up to 500 amn 1.2 sec/scan Column: Column program: 30 m x 0.25 mm DB-1 fused silica capillary 50°C for 4 min; 8°C/min to 270°C; Hold at 270 °C for 20 min Carrier gas He linear velocity: 35 to 45 cm/sec Source temperature: 170°C In the specific ion mode, the parent ion and two other fragment ions with the highest intensities were monitored for a period of 150 millisec each. The performance characteristics of the MS were verified by frequently injecting 50 ng decafluorotriphenylphosphine (DFTPP) into the GC-MS system. When the performance characteristics fell below the recommended levels, the source of the unacceptable performance charac- teristics were corrected either by cleaning the ion source, plugging pos- sible leaks, or changing the sorbent traps for the carrier gas. The GC-MS system produced a total ion chromatogram of each sample. Since the chromatograms contained large numbers of peaks, the MS data system was used to identify the peaks and locate the individual peaks when necessary. The tentative identification of each peak was by NBS Spectral Library search. The final identification of a compound was made by matching the relative retention time (with respect to anthracene-d-io) and the MS fragmentation pattern with an authentic standard. A peak was quantified by comparing its area with that of the anthracene-dio internal standard. This method of quantification assumes a linear response of peak area with the concentration. Results and Discussion The 24-hr laboratory leaching experiments conducted with plates coated with a coal-tar based material and water of controlled parameters (aggressive index, 8.6; free residual chlorine, 3.3 ± 0.03; and temperature 21.5 ± 0.5°C) showed the migration of the following components at the specified concentrations (iig/L); indene, 73; naphthalene, 36; quinoline, 213; indole, 45; acenaphthalene, 77; fluorene, 79; phenanthrene/anthracene, 298; carba- zole, 399; pyrene, 117; and triphenylene/ chrysene, 81. The results of three consecutive 24-hr washings, each per- formed to determined the aging effect of the coated plates, failed to show any difference in the concentrations of the individual PNA's in the successive leachates. The increase of residual chlorine level in water from 0.93 to 3.3 mg/L, resulted in decreased concen- ------- trations of some of the PNA's, particularly the levels of phenanthrene/ anthracene, purene, and fluorene in the leachates, possibly because of the formation of more chlorinated PNA's at higher chlorine level. No clear trend resulting from the teachability of PNA's was observed by changing the water temperature from 21.5°C to 28.4°C. Field test results of water collected from one utility showed that the concentrations of a few PNA's noticeably increased as a result of finished water passing through transmission pipes/tank lined with coal-tar-based material. For example, the concentration of indene, fluoranthene, and pyrene increased from 0.12 pg/L, 0.06 pg/L, and none detected to 0.17, 0.11, and <0.05 jig/L, respectively. No noticeable difference in the level of PNA's in water originating from pipes lined with coal-tar-based material was observed, however, when compared with levels from the other two utilities. This is probably because the transmission pipe used for sampling water in one utility was relatively new ( = 5 yr) and the transmission pipes in the other two utilities were older ( = 10 years old). Probably a difference of PNA concentrations in the two other systems would have been observed if the detection limit for the PNA's had been lower (<0.05 ng/L). No quantitative correlation could be established between the leaching observed under laboratory and field tests. This is not surprising considering that, among the many differences between the two cases, laboratory plates leach at about 4 orders of magnitude higher than do the field pipes. However, the leaching pattern from the field pipes can be qualitatively explained from the results of laboratory leaching tests. Summary and Conclusions The Laboratory leaching apparatus and the test protocol developed in the present study have been successfully used to determine the teachings of major individual components from plates lined with coal-tar-based material. The test protocol was developed to identify and quantify the major individual contam- inants in the leachates. This is an important improvement over the pre- viously available protocols because it will permit the assessment of possible health hazards arising from the individual leached components. The developed test protocol accelerates the leaching of components from coated surfaces and permits measurement of the level of major leachables in short-term leaching tests. The leaching study shows that PNA's are the major contaminants that will leach into water from surfaces lined with coal- tar-based material. Evidence is pro- vided that PNA concentrations in leach- ates depend on the free residual chlorine in the water and that an increase in chlorine level will decrease the level of some PNA's in the leaching water. The study also demonstrates that short-term laboratory leaching tests are not suitable for studying the aging effect in the field of pipes lined with coal-tar-based materials. Results of a few field tests with potable water after their passage through pipes lined with coal-tar-based material demonstrate that no quantitative corre- lation can be made between the laboratory and field leaching tests. The leaching of components from lined pipes in the field will be about four orders of magnitude lower than laboratory leaching from coated plates. To establish the possible PNA leaching in the field from transmission pipes lined with coal-tar- based material older than 5 yr, therefore, the detection limit for PNA quantification should be s50 ng/thousand liters. The results of laboratory leaching tests are useful in field tests aimed at predicting and rationalizing the observed leaching of contaminants from transmission pipes lined with coal-tar-based material. Recommendations Based on our experience with the present project, we recommend that the following ideas be considered for imple- mentation of further research: 1. Proper interlaboratory verification I undertaken so that the test apparat and the test protocol developed in tl study can be used by EPA as standard method for the approval materials intended to be used contact with drinking water. 2. The verified test protocol be used identify and quantify the maj individual components for subseque toxicological testing to determii which of those identified are causati factors in leachates found to be toxic 3. The verified test protocol be used further establish the effect of wal quality parameters on the migration contaminants into the water. 4. A guideline based on the toxicity the individual contaminants I established to specify the minimi. acceptable detection limit of t contaminants for the verified te protocol. 5. Further research be conducted verify the applicability of the prese test protocol to other materials used the transmission of potable water. 6. Further research be conducted establish a possible correlate between the leaching observed unc laboratory conditions and those unc field conditions. The full report was submitted in f fillment of Cooperative Agreement f> CR-811083 by Syracuse Resear Corporation, Syracuse, NY 13210-40* under the sponsorship of the U. Environmental Protection Agency. ------- Ronald Rossi, Craig R. Turner, and Dipak K. Basu are with Syra. — Corporation, Syracuse, NY 13210-4080. Alan A. Stevens is the EPA Project Officer (see below). —" The complete report, entitled "Proposed Test Protocol to Dete> I R;'K-J ^ /;'"!">jsi ~ fj 7 C !* Leaching into Potable Water," (Order No. PB 89-125 9591 AS; "" v /: *' ~" " '"'" subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Risk Reduction Engineering Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 United States Center for Environmental Research BULK RATE Environmental Protection Information POSTAGE & FEES PAID Agency Cincinnati OH 45268 EPA PERMIT No. G-35 Official Business Penalty for Private Use $300 EPA/600/S2-88/066 0000329 PS M ------- |