Document dajYXwgMgqgqozQM4kbyvRk0B

PLAINTIFF'S EXHIBIT A/C Pipe Producers Association Public Affairs Committee ro FROM' J. F. Welch, Director, Public Affairs Internal Correspondence October 1, 1981 DATE susject Study - "Chrysotile Asbestos Fibers in Drinking Water From Asbestos Cement Pipe" ACTION REQUIRED: Review for information Background During the August 5,1981 deposition of Marty Steven Kanarek, Ph.D. (University of Wisconsin, Department of Preventive Medicine) it was revealed that Kanarek would soon publish a report on the corrosion of A/C pipe. Staff requested that Lewis & Gilman contact Environmental Science and Technology to obtain a pre-publication copy. This was received on August 18, 1981. Current Status Enclosed is the study as published in the August, 1981 issue of Environmental Science and Technology, a peer reviewed publication. It is not known who reviewed the study, although the author acknowledges the help of J. (Jim) Millette (U.S. EPA - Health Effects Research Laboratory), W. (Warren) Winkelstein, Jr. (Dean, University of California - Berkley, School of Public Health) and others unknown to Staff. It is noteworthy that Kanarek does not acknowledge the assistance of Robert C. Cooper, Ph.D. (University of California - Berkley, School of Public Health) or Jack Murchio (electron microscopist) who respectively co-authored and performed asbestos counts for the 1978 EPA-Kanarek study of cancer incidence and asbestos in drinking water in the San Francisco Bay area. Staff understands that both Cooper and Murchio opposed publication of this study because there was insufficient data to support the conclusions drawn. In the authors' words, the report " ... raises questions as to the validity of aggressiveness indexes for prediction of fiber release from asbestos cement pipes." These "questions" are based on the alleged observance of a 115-fold increase in mean chrysotile asbestos counts in water conveyed through a certain A/C pipe system. Moreover, the authors claim that the Aggressiveness Indexes (AI) of these waters were very near or above the non-aggressive level (AI=12.0). The authors claim: To our knowledge, this is the first documentation of such a large release of asbestos fibers from A/C pipe in a system of nonaggressive drinking water. They dismiss out of hand, without any data to support the dismissal, the possibility that the alleged increases in asbestos counts may have resulted from sampling or counting technique, natural variations in source waters, tapping, "dead-end" accumulation and other variables. They go on to conclude: Thus, it appears that the aggressiveness index cannot be relied on as a sole predictor of (asbestos) release... ii CAPCO JEN 0033174 A/C Pipe Producers Association Public Affairs Committee TO International Affairs Committee $ -- FROM J. F. Welch, Director, Public Affairs State of California - Toxic Chemical Alert Plan ACTION REQUIRED: Review for information Internal Correspondence DATE October 22, 1981 Enclosed is a news article forwarded by Winner/Wagner & Associates, Inc., the Association's legislative and regulatory counsel in California. It describes a "right to know" program to advise the general public of chemicals known or suspected to cause cancer. Details of the plan are not known at this time, but this development clearly has the potential to heighten consumer awareness about asbestos in drinking water and A/C pipe. Staff has requested special counsel to closely monitor the development of this toxics program. If you have any questions, please do not hesitate to call. JFW/ajb Enclosures cc: A. Kahn, Esq. N. Rahn, Esq. B. Garcia (Winner/Wagner) w/o enclosures B. Pigg (AIA/NA) Sir Neville Stack (AIA) copies to: Public Affairs Committee International Affairs Committee H. Olson I. Adams W. Perrell J. Woods J. Baker T. Dougherty D. Stinson W. McCallie B. Collier R. Dorner B. Giboin C. Walters R. Hobbs A. Saoulis R. Jalan V. Pattabhi H. Hudson C. Barton S. Al-Tarkait E. van der Rest HEGA/1 Chrono CAPCO JEN 0033175 Thus., October 8,1981 San Sranrisro CDjrenidf 15 Gov. Brown Seeks Low Toxic-Chemical Alert Plan ' By Dale Champion Governor Brown proposed yesterday a statewide pro gram for alerting people to toxic chemicals in their homes, communities, food and water. As a beginning, the governor told a San Francisco news confer ence, his office plans to make public a list of all chemicals known or suspected to cause cancer. Brown said the new "right to know" toxic substance program for the general public would be an Extension of a program that was started recently to warn workers of on-the-job chemical hazards. J "1 think It's time to end the chemical coverup and let the peo ple know what is going on,"-Brown $aid. `There are powerful economic interests working in the opposite direction," he added, without offer ing specifics. * Brown, flanked by other state officials, appeared at the slate De partment of Industrial Relations, .where he accepted stacks of-peti tions asking for a full state and federal inquiry into the long-term effects of pesticides on human health and the economic impacts of a reduction in pesticide use.' -1 The petitions, containing 120.000 signatures, were presented by members by a Palo Alto group ^.called Creative Initiative. Brown called the petitions *fvery impressive" and promised Tcareful review of the questions 'raised in them. "You are bringing [to public attention something that is of deep concern," the governor Itold leaders of Creative Initiative. * Matching petitions were presented to federal officials in Washington on September 29. Brown, who has made control and cleanup of dangerous chemi- ,, <*Shta ' By Sieve Rtxgmnn . 'It's time to end the chemical coverup/ Brown said cals a major* theme of his adminis-" tration in recent years, said he will present a public "right to know" toxics program in a spatial message to the Legislature next January. "it's time to act." he said. "A rigbt-to-know' program for all Cali fornians can be achieved through legislation or an initiative, by state regulation and through local ordi nances." "There are potential acute and chronic health risks of toxic sub stances to which people are ex posed on a daily basis." the gover nor said, "it is time to extend and broaden the right of individuals everywhere in California to know the exact nature of toxic substances that affect their daily existence in the workptace. at home, - in the outdoor environment and in the food and water they consume." ii CAPCO JEN 0033176 -2- In sum, the report is the most formal and independent perpetuation of the misconception that (1) the AI was developed specifically to predict asbestos release from A/C pipe and (2) there exists a quantitative or even a semi-quantitative relationship between the AI and asbestos release from A/C pipe. Moreover, it clearly suggests to those concerned about fiber release that there are no suitable or safe conditions of use for A/C pipe, even in non-aggressive waters. The market implications of this claim are obvious. AACPP Action Plan 1. On August 18,1981, Jim Leineweber (J-M Health Safety and Environment) volunteered to revew the original asbestos counts and water chemistry, and to critique the validity of the authors' conclusions. Hal Olson (J-M) will assist in this effort. 2. Based on Leineweber's analysis, a rebuttal (probably a letter to the editor) will be co-authored by AACPP Staff and submitted for publication in Environmental Science and Technology. 3. An industry position statement (a modification of the rebuttal) will be prepared for distribution via "A/C Advisory." It will be particularly important to get this to state health officials and sanitary engineers as soon as possible. Field sales will be responsible for distribution to utilities. Staff plans to execute this plan by October 23, 1981. If you have any further suggestions, please do not hesitate to call. JFW/ajb Enclosure cc: A. Kahn, Esq. N. Rahn, Esq. N. Battle J. Leineweber (J-M) W. E. Smith, M.D. Brian Commins, Ph.D. Special Counsel (7) B. J. Pigg (AIA/NA) J. Stack (AIA) K. Robock ii CAPCO JEN 0033177 -3- copies to: Public Affairs Committee H. Olson I. Adams W. Perrell J. Woods J. Baker T. Dougherty D. Stinson W. McCallie B. Collier HEGA/6 Chrono a CAPCO JEN 0033178 (12) Unauthored,Seu Tcchnol. 1974,116,10. (13) " U.S. Department of the Interior "Gulf of Mexico Outer Conti nental Shelf Monthly Report"; U.S. Geological Survey Conserva tion Division: Metairie, LA, 1974-1980. .(14) Guinasso, N. C.,Jr.; Schink, D. R. "A Simple Physiochemical Acoustical Model of Methane Bubble Rising in the Sea", Technical Report 73-15T; Texas A &. M University: College Station, TX, 1973. (Hi) Corwin, J. F. In "Organic Matter in Natural Waters"; Hood, D. V\\, Ed.; University of Alaska, Institute of Marine Science Occ. Publ. No. 1,1970, pp 170-82. (16) Schwarzenbach, R. P.; Bromund, R. H.; Gschwend, P. M.;Zaf- iriou, 0. C. Org. Geochem. 1978, J, 93. (17) Sauer, T.C., Jr. Org. Geoc/rem. 1981, in press. (18) U.S. Department of the Interior "Draft Environment Statement, Outer Continental Shelf, Gulfof Mexico, OCS Sale No. 44", Bureau of Land Management, 1976. (19) Schluntz, E. K., U.S. Department of the Interior, U.S. Geological Survey Conservation Division, Metairie, LA, personal communi cation, 1980. (20) Sauer, T. C., Jr. Ph.D. Dissertation, Texas A & M University, College Station, TX, 1978, p 346. Received for review May 23,1980. Accepted March 25,1931. Support for this work was provided by NOAA Contract No. 03-78-D080043. Chrysotile Asbestos Fibers in Drinking Water from Asbestos-Cement Pipe Marly S. Kanarek* Department of Preventive Medicine and Institute for Environmental Studies, University of Wisconsin, Madison, Wisconsin 53706 Paul M. Conforti Department of Biomedical and Environmental Health Sciences, University of California, Berkeley, California 94720 Lorene A. Jackson Association of Bay Area Governments, Hotel Claremont, Berkeley, California 94705 Chrysotile asbestos fiber counts were compared in drinking water samples taken before and after flow through asbestoscement pipes in the San Francisco Bay area. The drinking water supplies in that area are chemically nonaggressive by the standard measures. In one major distribution system, mean chrysotile counts were 115 times greater in drinking water after flow through the pipe. This demonstration of substantial increases in asbestos fiber counts in nonaggressive drinking water raises questions as to the validity of aggres siveness indexes for prediction of fiber release from asbes tos-cement pipes. Many issues have been raised as to the possible health risks to the general population from exposure to asbestos in the environment (1). A key question is whether fibers are released to drinking water passing through asbestos-cement (A-C) pipe (2). At least 1.5 X 10s mi of such pipe are in use world wide, with more than 200 000 mi in U.S. drinking water sys tems (3). The need to answer this question is now more im portant because of the results of our recent epidemiological investigation in the San Francisco Bay area which revealed statistical associations between the presence of chrysotile asbestos fibers in drinking water systems and the incidence of digestive and other cancers (4,5). Thirty-nine drinking water samples paired before and after passing through asbestos-cement pipe were taken for asbestos fiber counting as part of the 372 drinking water samples col lected during the effort to assess levels within the water dis tribution systems of the San Francisco Bay area for the epi demiology study. Undoubtedly, the primary sources of as bestos fibers in some ofthe drinking water supplies in this area are watersheds and reservoirs consisting largely of serpentine rock, the parent material for chrysotile asbestos. Mean chry sotile asbestos fiber counts (in fibers/L) for the major water systems were 2.7 X 10' for Pardee, 2.5 X 10s for Contra Costa Canal, 2.G X 10n for Hetch Hetchy, 4.1 X 10T for Crystal Springs, 5.1 X10 for San Andreas, and 1.9 X 10 for San Geronimo/Bon Tentpe. This report describes the field data re lating to the possible additional release of fibers from asbes tos-cement pipe. There are limited data on the question of release of fibers from asbestos-cement pipe. A 1974 study revealed that as bestos fibers can be transferred from the pipe wall to water circulating through a closed A-C pipe loop, not buried in the soil. Field samples from Malvern, PA, and Glendale, AZ, showed possible release of fibers from pipe walls, and the amount of release appeared not to be significantly influenced by the existing fiber level in the water before flow through the pipes (6). Studies of private wells and municipal waters in Rio Grande Valley, NM, and Illinois showed no evidence of sig nificant increases in fiber counts attributable to pipe (7,8). It has been assumed that there is a positive correlation be tween fiber release from the walls of A-C pipe and the "ag gressiveness" or corrosiveness of water. An index of aggres siveness has been formulated as follows (9): aggressiveness index = pH + log {AH) where A = total alkalinity in mg/L CaC(>3 and H = calcium hardness in mg/L CaC03. The definitions for the degrees of aggressiveness for the index are as follows: highly aggressive, pH + log {AH) < 10.0; moderately aggressive, pH + log {AH) = 10.0-11.9; nonaggressive, pH + log (AH) > 12.0. A 1979 paper by EPA personnel detailed the results of their pipe study in 10 public water supplies in the U.S. (10). There appeared to be significant releases of fibers from A-C pipes in only the five systems with highly aggressive water. A1978 epidemiologic study in the state of Connecticut of the use of A-C pipe in public water supplies and the incidence of gas trointestinal cancers showed no obvious correlations (11). The supplies were not exposed to significant amounts of naturally occurring asbestos. Asbestos fiber counts in water systems in Connecticut were rarely over 1 X 105 fibers/L. In comparison to West Coast water supplies exposed to asbestos-bearing rock, or the drinking water of Duluth, these values are rela tively low. A 1980 reanalysis (12) of the Connecticut data utilizing more sophisticated statistical techniques also showed no consistent associations between asbestos fiber content of town drinking water and cancer incidence. It must be re membered that asbestos-cement pipe was not installed in any parts of Connecticut until 1955, and the possible maximum AAtO OQCV'OI IAA4C m . fS\ j ** . > - I' .... -' *I CAPCO JEN 0033179 latent period for asbestos-related tumors is up to 40-50 yr. Thus it may be too soon for definitive conclusions in Con necticut. Experimental Sectiun . The water sampling program for the San Francisco Bay epidemiulogy study was designed to assess the levels of as bestos in the drinking water of the area's different water dis tribution systems. The collection of water samples before and after flow through varying lengths of A-C pipe was only one of several factors involved in this overall assessment of as bestos in drinking water supplies. Most "before" A-C pipe samples were collected from treatment-plant effluents or at the main turnout going into a community. Samples were then taken on the same day from various taps within the distribu tion system to determine continuity ofvalues throughout the area. Since the primary concern of the study was to determine asbestos fibers per liter of drinking water received by the population, the "after" A-C pipe samples were often taken from a tap near the extremity of a system. The sampling period for the pipe data was July 1974 to March 1978, with most of the data gathered after August 1976. Turbidity and pH were measured on all samples collected after August 1976. Total alkalinity and total hardness (both recorded in mg/L CaCOj) were measured on all samples taken after March 1977. Aggressiveness indexes were derived for all samples. Calcium-hardness values were calculated from av erage calcium hardness to total hardness ratios for each sys tem. Samples taken before March 1977 were assigned ag gressiveness values calculated from measured pH and from estimated alkalinity and hardness averages from each dis trict's historical water quality records. The length, the di ameter, and the age of A-C pipe involved in the before-after samples were determined by reviewing maps of districts and consulting with water purveyors. The characteristics of the A-C pipe and the drinking water sampled are presented in Table I. All of the drinking waters of the Bay area rale as nonaggressive or moderately aggressive according to the standard scale. Even the moderately aggressive waters tend toward nonaggressiveness on the scale. Water samples were collected in 0.5-gal glass bottles cleaned with hot detergent water and rinsed with glass-distilled water. The asbestos content of the distilled water was below de tectable limits of the analysis. Before each water sample was Table l. General Characteristics of Asbestos-Cement Pipes and the Drinking Waters Sampled watac ayatam A e c lhara pipe median age (yr) median length (m) median diameter (cm) 18.5 541 20.32 20.25 1986 30.48 22 9 1951 2050 20.32 25.4 water mean pH mean total alkalinity (mg/L CaCOj) mean calcium hardness (mg/L Ca CO3) mean before aggressiveness Index mean after aggressiveness index mean aggressiveness Index of all samples overall aggressiveness 8.96 8.03 9.20 7.79 82.37 83.20 53.20 129.98 42.6 79.18 23.05 82.35 12.28 11.83 11.93 11.30 12.52 12.05 12.54 12.07 12.4 11.94 12.24 11.95 none moderate none moderate collected, the glass bottle was rinsed several.times at the sample site with the water to be analyzed. Water samples were filtered through 0.45-pm size Millipore filters. The Millipore filter material, which is a cellulose ester, was removed from the residue by acetone vapor in a cold-finger Soxhlet extractor. This was accomplished by first covering collodion-coated 200-mesh nickel grids with a thin layer of carbon in a highvacuum optica) coaterand then placing the grids on the cold finger. Disks of the Millipore filter, 3.3 mm in diameter, were placed particulate side down on the prepared grids. The cold finger was then placed into the extractor cold and the acetone reservoir heated to just below the boiling point of acetone, a temperature sufficient to produce a very slight tremor. The Millipore filter was dissolved in 6-12 h, leaving any particulate matter from the filtered water sample on the grid. The grids were examined and fibers counted by using a Sieman Elmiskop electron microscope at 10 000X optical magnification. One grid opening of the 200-mesh grid is con sidered one field. Twenty fields were counted in grids which had a moderate or small number of fibers or where there were no fibers detected. In cases of samples heavily loaded with asbestos fibers, a sufficient number of fields were counted to get a total of 100 fibers. The number of fibers counted were related to the grid area counted, to the whole grid, to the filter area, and to the volume of sample filtered. This count was recorded ns the number of fibers of either chrysotile or un identified fibers per liter ofwater. The lowest detectable limit of asbestos concentrations in tap water was between 1 X 104 and 1 X 10s fibers/L, depending on the turbidity and the subsequent amount of water sample filtered. The diameter and the length of each fiber were measured and recorded per field. Chrysotile asbestos fibers were identified and distin guished from other fibers by a combination ofmorphology and selected area diffraction patterns using the transmission electron microscope. Several split samples were analyzed by another laboratory with almost identical results. In many of the water samples, unidentified fibers were observed. Un identified fibers could have been amphibole asbestos fibers, fibrous talc, or halloysite (a type of fibrous clay). Results Table II presents the mean chrysotile fiber counts before and after sections of A-C pipe for each system. When the sample result was below the detectable limit of the analysis (BDL), the lowest detectable limit (LDL) value of the analysis was used as the'sample count in taking the mean. The nonparametric Sign test was used for an overall view of increase or decrease in chrysotile counts on the 39 pairs of before-after observations (73). There were 19 pairs in which increases were observed; 9 pairs showed decreases; and 11 pairs exhibited no change in fiber count between before and after observations. The Sign test included a correction for continuity and treated each no-change observation (tie) as '/> increase and V> decrease. The Sign test yielded a p value of 0.039. This indicates a moderate statistical significance, falling between the usual a levels of 0.05 and 0.01. Each system was then analyzed sepa rately, yielding p values of 0.79 for system A, 0.50 for system B, and 0.01 for system C. The number of pairs recorded on Table II. Mean Chrysotile Asbestos Fiber Counts before and after Asbestos-Cement Pipe tyitm' to.ct batora-attar palra chrysotila fibara/L bate** attar A 14 2.25 X 10s 1.26 X 10s B 12 1.88 X 105 3.11 X 10s C 9: 7.50 X 104 8.63 X 10s others 4 2.88 X 10* 5.35 X 10s CAPCO JEN 0033180 oilier systems was insuffic*ifrnl to carry out the Sign test. Since the Sign test does not account for the amount of increase or decrease in filler count, i paired Student's t test for the dif ference in number of fibers counted hetween each of the 39 before and after sample pairs was used. Again, the increase in fibers after transport through A-C pipe was moderately significant, p = 0.02, with a t statistic of 2.0-1 and 38 degrees of freedom. Discussion Eight of the nine system C before-after pairs showed an increase after flow through A-C pipe. These are listed in Table III. The mean of the before samples was 7.500 X 104 fibers/L and the mean of the after samples was 8.628 X 10* fibers/L. This is an increase of 115 times the number of fibers, or ap parently, on the average, 8.553 X 10 fibers/L were released from the pipe into the drinking water. To our knowledge, this is the first documentation of such a large release of asbestos fibers from A-C pipe in a system of nonaggressive drinking water. A possible explanatiun of the increase in fiber counts after flow through A-C pipes is that fibers already present in the water from natural sources break up, thus yielding higher counts. Analyses of fiber size distributions have been carried out on data generated in the overall sampling program for the San Francisco Bay area (14). These analyses pooled system A and system C data to obtain sample sizes large enough to produce statistically reliable conclusions. The after A-C pipe samples have a larger proportion of longer fibers than the before A-C pipe samples. Thus, the theory of fiber breakup seems to be invalidated in that the naturally occurring fibers were generally shorter than those from segments of A-C pipe. Other factors that might affect the release of fibers from A-C pipe include size of pipe, flow rate of water, alternate wetting and drying of surfaces, quality ofjoint materials used, corro sive cleaners in use to flush sections of pipe, and tapping of pipe. Hydrants and other "dead ends" in a water distribution system often act as a reservoir for fibers released by tapping at other places in the system. None of these factors accounts for the releases recorded in system C. The chemistry of system C water seems unusual as com pared to the other systems. Even though the aggressiveness index is high, indicating nonaggressive water, the calcium content of the water is low. It appears that the aggressiveness index may be unduly weighted in this case by the high pH of these waters, which results from lime addition by the water utility for this system. As was found in the study by Buelow et al.(10), the aggressiveness index appeared to rise after flow in the pipes in system C. However, as can be seen from Table I, the aggressiveness indexes rose after flow through the pipes in all of the systems, even though there appeared to be only substantial release of fibers in system C. Thus, it appears that the aggressiveness index cannot be relied on as a sole predictor of release, and further research is needed on water chemistry factors and fiber release. Further sampling is recommended to examine the consistency of release in those areas with A-C Table 111. Individual Chrysotile Asbestos Fiber Counts before and after Asbestos-Cement Pipe in System C system C location beloro LOt-'* UJL* i 5.0 X 10* 2 5.0 X 10` 3 5.0 X 104 4 5.0 X 10* 5 2.0 X 10s 6 2.0 X 10s 7 BDL6 8 BDL*' 9 2.5 X 10* 5.0 X 10* 5.0 X 10* 5.0 X 10* 5.0 X 10* 1.0 X 10s 1.0 X 105 2.5 X 10* 2.5 X 10* 2.5 X 10* 6.2 X 10 3.4 X 107 2.5 X 10s BOL4 6.0 X 10s 6.4 X 10s 2.0 X 107 1.5 X 10 5.0 X 10* 1.0 X 10s 2.5 X 105 6.7 X 10* 1.0 X 10 1.0 X 10s 8.4 X 10* 2.0 X 10s 5.0 X 10* 2.5 X 10* * Lowest detectable limit ol the analysis. 4 Below detectable limits of the analysis. pipe and to give insight into the mechanisms and predictors of such release. Acknowledgment We thank J. Millette, C. L. Chiang, W. Winkelstein, Jr., and S. Selvin for reviewing an earlier draft of this work and Arthur Sacks and Frances Culwell for technical editing. Literature Cited (1) Selikoff, I. J.; Lee, D. H. K. "Asbestos and Disease"; Academic Press: New York, 1978. (2) Hatlenbeck, W.; Hesse, D. In "Dusts and Disease"; Lenten, R., Dement,J., Eds.; Pathotox: Park Forest South, IL, 1979; p 171. (3) Olson, H.J. Am. Water Works Assoc. 1971,65,515. (4) Kanarek, M. S.; Conforti, P. M.; Jackson, L. A.; Cooper, R. C.; Murchio, J. C. Am. J. Epidemiol. 1980,112,54. (5) Conforti, P. M.; Kanarek, M. S.; Jackson, L. A.; Cooper, R. C.; Murchio, J. C. J. Chronic Din., in press. (6) American Water Works Association Research Foundation. J. Am. Water Works Assoc. 1974,66, Sept, part 2. (7) Oliver, T.;Murr,L.J. Am. Water Works Assoc. 1977,69,428. (8) Hallenbeck, W.; Chen, E.; Hesse, C.; Patel-Manklik, K.; Wolfe, A. J. Am. Water Works Assoc. 1978,70,97. (9) American Water Works Association. "Standard tor Asbestos- Cement Distribution Pipe, 4 in. through 16 in., for Water and Other Liquids," AWWA: Denver, CO, 1977; AWWA C400-77. (10) Buelow, R. W.; Millette, J. R.; McFarren, E. F.; Symons, J. M. J. Am. Water Works Assoc. 1980,72,91. (11) Harrington, J. M.; Crann, G. F.; Meigs, J. W.; Landrigan, P. J.; Flannery, J. T.; Woodhull, R. S. Am. J. Epidemiol. 1978, 107, 96. (12) Meigs, J. W.; Walter, S. D.; Heston, J. F.; Millette, J. R.; Craun, G. F.; Woodhull, R. S.; Flannery, J. T. J. Environ. Health. 1980, 42,187. (13) Conover, W. "Practical Nonparametric Statistics"; Wiley: New York, 1978; p 121. (14) Tarter, M. "Data Analysis of Drinking Water Asbestos Fiber Size"; USEPA: Cincinnati. OH, 1979; EPA-600/1-79-020. Received torreview June 16,1930. Accepted April 27,19S1. This work teas supported 6v Environmental Protection Agency Grant Ho. 604366. CAPCO JEN 0033181