SERA United States Environmental Protection Agency Municipal Environmental Research^ Laboratory Cincinnati OH 45268 Research and Development EPA-600/S2-82-048 August 1982 Project Summary Stormwater Management to Improve Lake Water Quality Martin P. Wanielista, Yousef A. Yousef, and James S. Taylor Many investigators have identified the urban environments as those pro- ducing high levels of water pollutants relative to other land uses. In a 59 hectare (146 acre) urban watershed in Orlando, Florida, the Stormwater sys- tem discharges to an 11 hectare (27 acre) lake. The lake water quality is characterized by frequent algal blooms, odor, and in general, reduced recreational activities. The lake is one of the focal points of the city's recre- ational/social programs. Stormwater management proce- dures were established based on the runoff sampling program and a target mass pollutant reduction of storm- water runoff with special emphasis on phosphorus. Management of storm- water for the removal of phosphorus was accomplished by diversion for retention (filtration) of the first flush of pollutants. Lake water quality will be managed by the reduction of Stormwater mass, bottom mud inactivation, and littoral zone plantings. Approximately half of the water column phosphorus concen- tration was estimated to originate in the bottom muds and return to the water column primarily during anoxic conditions. Thus, a coagulant cover- age of the bottom muds will be used. The coagulant possibilities are alum sludge or alum. This Project Summary was devel- oped by EPA's Municipal Environmen- tal Research Laboratory. Cincinnati. OH, to announce key findings of a research project that is fully docu- mented in a separate report of the same title (see Project Report ordering information at back). Introduction Stormwater may be a significant source of surface water pollution in urban areas. Lake impacts have been and continue to be studied on an international level by the National Eutro- phication Research Program in the United States and an international program with United States participa- tion. This summary presents Stormwater impacts on an urban lake and recom- mendations for Stormwater and lake management. The impact was first defined by visual observation and storm- water was the major pollution source. There was no point sources of industrial or domestic wastewaters. Samples were taken of Stormwater, lake water, and bottom muds and estimates made of mass loading and impacts to determine a combination of management practices. Some elevations of best management practices (BMP's) for urban Stormwater had been completed before this work. The critical relationship between a BMP and receiving water quality, however, had not been well documented, except for some dissolved oxygen responses in rivers. Watershed and Lake Characteristics The drainage area is the Lake Eola watershed in Orlando, Florida. The Stormwater system, which is separate from the sewage system, drains a ------- watershed of approximately 59 hectares (146 acres), composed of 33.7 hectares (83.2 acres) of commercial and 25.3 hectares (62.8 acres) of residential areas discharging to an 11-hectare (27-acre) lake. In addition, 4.5 hectares (11.2 acres) of parkland surrounds the lake and are not considered part of the watershed because of infrequent runoff to the lake. Streets and parking lots comprise approximately 19.7 hectares (49 acres) of the watershed within a total of about 49.3 hectares (122 acres) of impervious lands. The pervious area is only 9.7 hectares (24 acres), most of which is in the residential areas. The land use is illustrated in Figure 1. Visual observation of the watershed indicates a total of 35 parking lots with an area of about 12 hectares (29 acres) discharging stormwater. Parking areas were identified to select possible areas for management of stormwater dis- charged to the land-locked lake. One of the parking lots was designated a samp- ling location for runoff waters and is usually maintained between 26.5 meters (87.0 ft) and 27 meters (88.5 ft) above sea level by well discharge and pumping. It is a shallow lake with a mean depth of approximately 3 meters (9.84 ft) and about 73 percent of the total volume located within the 0 to 3 meter (0 to 9.84 ft) frustrum layers. Most urban lakes in central Florida (more than 5,000) have similar physical characteristics. Benefits The benefits of the lake and its surroundings are evident but difficult to quantify. The lake is a focal point of Orlando with frequent music concerts, arts and crafts shows, tourist visits, children's park, and relaxation areas, and is one of the main reasons the downtown area continues its economic security. The property surrounding the lake brings top value because of its location. The dollar benefits from lake activity are estimated at $2.3 million per year. Lake Impacts From visual observation and analytical data. Lake Eola has persistent algal blooms virtually year round. Populations of the macroscopic algae, Chara, and the filamentous green algae, Spirogyra, covered up to 30 percent of the lake surface during the summer rainy sea- son. Lake Eola water was found to be somewhat alkaline with pH ranging from 8.4 to 9.5 (indicative of the rate of algal production). The average annual value and the measured range of values for pH, chlorophyll-a, inorganic and organic carbon, and Secchi disc trans- parency are shown by Table 1. The average values shown in Table 1 are similar to values reported in the literature for eutrophic lakes. L_ 1 Legend Commercial Residential Parkland Sampling locations Sub- watershed Figure 1. Lake Eola watershed land uses and sampling locations. 2 ------- Table 1. Values for Selected Parameters Measured in Lake Eola, Florida, Between July 1978 and August 1979 Parameter Chlorophyll-a Organic carbon Inorganic carbon pH Secchi disc No. of Samples 64 67 68 57 32 Average Value 25.4 10.9 18.8 8.85 106 Standard Units Deviation mg/m3 mg/L mg/L cm 8.8 6.7 6.4 — 13.0 Range of Values 9.0 - 3.0 - 13.8 - 8.40- 90 36.4 29.1 40.6 9.5 120 Concentrations of dissolved oxygen in Lake Eola, although usually at or above saturation near the surface, drop period- ically during the spring and summer months to less than 1 mg/L in deep areas of 4 meters (13.1 ft) or greater. Phosphorus from the bottom sediments is released up to a level of 250 mg/m2 after 2 months of anoxic conditions. This anaerobically released phosphorus has the potential for increasing water column phosphorus by 11.6 /ug P04~3- P/L, about 50 percent of the average orthophosphorus concentration in the lake(23//g/L). Extensive bioassay experiments indi- cate that when the concentration of orthophosphorus in Lake Eola was less than 0.10 mg/L, algal production was regulated by adding orthophosphorus alone. Above this concentration, appar- ently an excess of phosphorus was available, and algal growth was regu- lated by the N/P ratio. In most cases, however, the concentration of orthophos- phorus was below0.04 mg/L, and most likely algal production was limited by the concentrations of added phosphorus alone. In addition to the enhanced algal growth conditions experienced during the summer rainy months, runoff enter- ing the lake after prolonged periods of drought may produce severe toxic effects on aquatic life in Lake Eola. Contami- nants are allowed to accumulate within the watershed, and when a storm event occurs, the mass loading to the lake is many times larger than experienced during frequent rainfall periods of time. This influx of toxic and oxygen demand- ing wastes can be lethal to many forms of aquatic life. Evidence of such a phenomenon was recorded in March 1979 when a rain event followed a dry period of 6 weeks. oncentrations of organic carbon as .ligh as 400 mg/L were measured in stormwater runoff entering the lake during this event. Two days later, dis- solved oxygen concentrations had been reduced from saturation near the sur- face to 4 mg/L at a depth of 1 meter and to near zero below 2 meters. Numerous large-mouth bass averaging 1 to 2 kg were found floating in the water, and large masses of dead filamentous algae had accumulated in thick mats over much of the lake's surface. During 1979, a total of six fish kills were reported. One dead bass (about 2 kg) floating at the surface was brought to the laboratory, processed and analyzed for metal con- centrations in selected organs (heart, gall bladder, liver, stomach) and flesh. It appeared that nickel and lead concen- trated in the gills, iron concentrated in the heart, and zinc and copper concen- trated in the liver. At this time, however, it is not known that these metals were directly responsible for the fish kill. A pathogen isolation study was con- ducted for 1 year with the collection of 129 water and sediment samples. Four- teen were composites of runoff, 32 were bottom samples, and 83 were lake water. Clostridium was isolated from the bottom sediments of the lake and Salmonella was isolated from the lake water samples. Domestic ducks in the Lake Eola waters and park areas number approxi- mately 20, with decreasing populations noted over the past 5 years. Microbiol- ogists at the Orange County Pollution Control Agency have speculated during site visits, that gas production from the anaerobic sediments is increased in the summer months. This anaerobiosis pro- motes growth of the botulism organism that produces a toxin that, in turn, concentrates in the small insect larvae of the sediments. When ducks eat the larvae they can die. After the autopsy of two dead ducks, botulism was the reported cause of death. Stormwater Stormwater pollutants and flowrates were first estimated by sampling storm- water relative to the hydrograph. In this hydrograph-related sampling program, eight rainfall/runoff events were quan- tified. Next, a composite sampling pro- gram was completed with seven rain- fall/runoff events. One major question was the percentage of dissolved pollu- tant materials present in the runoff. The sampling program indicated that the dissolved nutrients and organics were approximately 50 percent or more of the total, while the dissolved fraction of lead was 20 percent. From the 15 runoff samples, estimates were made for aver- age mass loading (as discharged) and average concentrations. Priority pollu- tants were measured. Estimates of loading rates from both commercial and residential areas were calculated from the runoff studies. The Lake Eola site loading data were compared with the loadings of SWMM/ level I analysis and other national data. The suspended solids (SS) and BOD data (Table 2) appear to agree, but total nitrogen loadings are higher in the Lake Eola watershed. Possible reasons are that the residential areas should be classified as commercial when consider- ing loading rate data, the landscaping maintenance places an additional nitro- gen load, and the heavy rainfall (130 cm) is greater than the national average. Most likely, a combination of these reasons caused the increase. The commercial and residential land use pollution contribution to the total was estimated to be 98 percent for SS, 96 for BOD, 95 for total organic carbon, 94 for total Kjeldahl nitrogen (TKN), and 91 for total phosphorus (TP). Total contribution was defined as the sum of the contributions from stormwater, at- mosphere, and ducks resident on the lake. The sampling program and the lake impact work led to the following conclu- sions: (1) stormwater is the major external source of lake related pollution; (2) phosphorus and other stormwater pollutants require removal; and(3) sedi- mentation was possibly not the choice method for stormwater management because of the large percentage of dissolved pollutants. Target Phosphorus Reduction The major question is to what degree should the bottom sediment and storm- ------- Table 2. Loading Rate Comparisons fkg/ha-yrj Data Source SS 5005 roc TN TKN POt-P TP Lake Eola Commercial Residential 1076 827 196 87 1167 757 32.0 40.5 27.8 36.1 1.7 3.1 3.5 6.2 SWMM/Levell Commercial Residential 1255 922 181 45 16.7 7.4 4.3 1.9 National A verages Commercial Residential 941 470 97 39 14.5 6.6 3.0 2.0 water be treated to economically reduce nutrient enrichment, fish and duck kills, and algal activity to an acceptable level? Using the trophic state models, a target reduction level of phosphorus loadings into the oligotrophic/mesotrophic level may reduce algal blooms. In addition, a chlorophyll-a mean concentration of 7 /ug/L may indicate a mesotrophic state. Table 3 illustrates the target level and need for an approximate 90 percent reduction in phosphorus load and con- centration. In the National Eutrophication Study, total phosphorus concentration of less than 10/ug/L in the water column was noted as a target reduction to classify lakes as oligotrophic. A combination of stormwater treatment and bottom sedi- ment inactivation may produce a water column concentration of less than 10 fjg/L. The bottom sediments were esti- mated to contribute 11.6 yi/g/L of the average water column concentration of 23//g/L Stormwater Management Selection Each stormwater management prac- tice that could be defined in terms of cost and efficiency and was practical for the watershed was evaluated for storm- water control. Selecting the best combi- nation of practices that met cost and efficiency constraints (least cost) could be aided by a computer analysis. Cost-efficiency curves (present value dollars versus removal quantities) were developed for each subwatershed of Lake Eola. Removal efficiencies were used from the literature and nonpoint urban runoff programs. These efforts in the central Florida area had defined the efficiencies and costs for diversion/ percolation basins, swales, underdrains, and vacuum sweeping nonpoint source management methods. In the highly impervious urban areas, the cost of land is expensive, and land intensive activ- ities (detention and retention basins) are sometimes not aesthetically pleas- ing. Thus, street sweeping diversion with retention underground, and catch- basin cleaning appeared probable for the Lake Eola area. Dutch drains, rooftop storage, coagulation, filtration, and solids concentrators were other manage- ment methods being investigated. These methods formed the basis for determin- ing optimal combinations of practices. A linear programming network routing model was incorporated into a computer program, and the cost-efficiency curves were estimated by "piecewise" linear approximation. Table 3. Target Reductions Models Before Vollenweider 2.33 g-P/m2/yr Dillon 0.49g-P/m2 Larsen-Mercier 0.48 mg/L OECD/ chlorophyll1 269 mg-P/m3 Target Reduction Levels 0.2 g-P/m2/yr 0.05g-P/m2 0.05 mg/L 70 mg-P/m3 'Reduction corresponding to a chlorophyll-a of 7 fjg/L. 4 One limitation on stormwater control was the use of private property. Thus, it was decided to do all management within the city right-of-way. The alter- natives considered for management of the stormwater were: 1. diversion of stormwater to the sanitary sewer system for treat- ment; 2. street cleaning by both broom and vacuum sweepers; 3. diversion of stormwater into perco- lation basins; 4. conversion of inlets to catchbasins; 5. coagulant addition with sedimen- tation (detention); 6. silt removal from lake, and draw- down every 5 years; 7. natural "living filter" treatment through littoral zones; 8. fabric bag filters; 9. detention systems (on-line pass through); 10. "best" combination of any or all of the above alternatives; and 11. diversion of stormwater into infil- tration trenches near the lake. The first alternative was eliminated because it was not considered as a general solution for other areas and it required the replacing of over 7.00C meters (22,950 ft) of sanitary sewei lines, thus the capital cost of pipe anc pumping stations was over $600,000. A "column study" was done to exam ine detention and detention after coagu- lation as a method for improving watei quality. Because a significant amount o the pollutants in stormwater were in the dissolved or colloidal state, coagulatior with detention may be necessary t< remove these types of pollutants. Deten tion with significant holding times wil remove suspended materials. The reduc tion reported ranged from a high of 5: percent for TSS to a low of 20 percen for lead. Alum coagulation improvei stormwater quality more than detentioi alone. Reductions of TP and TSS afte alum coagulation were in excess of 91 percent. Standard isoconcentration line and regression equations relating per cent removal to time, settling velocity and the logarithm of time were devel oped. The regression equations are eas to use and were developed within predictive error of less than 5 percen The detention systems were judged to expensive because of land area require and the chemical costs. Detention afte coagulation, however, may be feasibl for other locations. ------- Before the percolation alternatives could be considered, the infiltrative capacity of the soils was estimated by defining the type of soils and the loca- tion of the water table. The water table is at least 2 meters (6.56 ft) below the ground surface for a ground elevation of 29 meters (95 ft) or higher. Borings close to the lake indicate the water table is near elevation 27 meters (88 ft). In addition, sandy soil is available to about 6.5 meters (20 ft) below ground level. Percolation of stormwater is possible for parking lot and street drainage for those areas whose ground elevation is above 29 meters (95 ft). All alternatives were evaluated for estimated cost and yearly pollutant removal efficiencies. The cost for the natural "Living Filter" areas was esti- mated from local contractors and the city of Orlando records. Native vegeta- tion was selected and has been used in other lakes. All other cost data were obtained from recently bid sewer pro- jects. The selected solution was based on minimum present value cost and maxi- mum removal efficiencies. The fabric bag alternative had a lower capital cost, but poor removals relative to other alternatives. The location of the best management practices was near parking lots and immediately before lake dis- charge. Parking lot and street diversion were designed to percolate the first 0.6 to 1.25 cm (0.25 to 0.5 in.) of every storm chosen, resulting in a removal efficiency of 90 percent on a yearly basis. The resulting capital cost for stormwater management and lake res- toration was approximately $6,2507 hectare ($2,5007 acre) of watershed. Results and Conclusions Presented here are the results and conclusions separated into the areas of stormwater, lake impacts, sediments, and selection of management practices. Stormwater The stormwater pollutants and quanti- ties were first estimated by sampling stormwater relative to the hydrograph. Eight rainfall/runoff events were quan- tified in this manner. Next, a composite sampling program was initiated meas- uring seven rainfall/runoff events. From the 15 runoff samples, estimates were made for mass loading (as discharged) and average concentrations. Loading rate and average concentration data were estimated for 19 water quality parameters, including data for 10 metals. Results of the sampling program are: 1. The dissolved portion of storm- water pollutants was equal to or greater than 50 percent of the total except for zinc, lead, and chromium. 2. Loading rates of SS, BOD5, and dissolved phosphorus were com- parable to national averages. Load- ing rate data for nitrogen are higher than national averages. 3. The relative loading rates for each external (not including lake recy- cle) source of pollution were esti- mated. Stormwater produced the greatest percentage of the total loadings. Lake Impacts During this research, monthly water quality analyses were performed in Lake Eola and bioassay experiments were conducted to determine the effect of nutrients and stormwater additions on algal productivity in this lake system. Coagulation of stormwater to remove nutrients and limit algal production was also studied. From the results obtained in this research, the following conclu- sions were reached; 1. The input of stormwater in Lake Eola has damaged this aquatic system. Persistent algal blooms exist virtually year round. Bottom sediments have become covered with a layer of loose flocculant material and anoxic conditions exist in areas more than 4 meters (13.12 ft) deep during the spring and summer. 2. When the concentration of ortho- phosphorus in Lake Eola is less than 0.10 mg/L, algal production is regulated by the addition of orthophosphorus alone. Above this concentration, it appears that an excess of phosphorus is avail- able, and algal growth is regulated by the N:P ratio. In most cases, however, the concentration of orthophosphorus is below 0.04 mg/L, and algal production is limited by the concentration of added phosphorus alone. 3. Although nitrogen was able to stimulate algal production in lim- ited bioassay experiments, it will probably not be a limiting resource in Lake Eola when considered over a period of years because of the large numbers of nitrogen-fixing blue-green algae that are charac- teristic of eutrophic systems. 4. Adding stormwater runoff to Lake Eola water in any concentration increases algal production. A mix- ture of 25 percent stormwater runoff produces the largest stand- ing crop. 5. Coagulating stormwater runoff with alum reduces concentrations of both orthophosphorus and ni- trate nitrogen by 80 to 95 percent and re moves certain heavy metals. "3. Coagulating stormwater runoff significantly reduces phosphorus and most metals. Thus, the treated stormwater is less toxic to aquatic organisms and algae growth is reduced. 7. Dry weather storm sewer flow has a negligible effect on algal produc- tion in Lake Eola. 8. Maximum algal growth in bioassay experiments generally occurred after 6 to 11 days of incubation. When stormwater runoff is added to Lake Eola water, however, a growth lag is often experienced that may extend the time required to reach maximum yield to as many as 18 days. 9. Continuous stormwater inputs into Lake Eola during the rainy season greatly enhances algal growth because of the constant input of nutrients and dilution of toxic components. Inputs of storm- water after a long dry spell may inflict serious toxic effects on aquatic life. 10. Concentrations of copper and zinc in Lake Eola are sufficient to cause incipient inhibition of algal production. Average stormwater runoff concentrations of copper, zinc, and lead are sufficient to produce complete inhibition or algicidal effects. 11. Human pathogens were identified in the lake. Clostridium was iso- lated from bottom samples and Salmonella from water samples. Sediment During this study, sediment phos- phorus was characterized and the mag- nitude and effect of anaerobically re- leased phosphorus were studied to determine the effects on the productivity of Lake Eola and the related eutrophica- ------- tion. Based on the results of this re- search, the following conclusions were reached: 1. The major source of phosphorus entering Lake Eola is from storm- water runoff. The major phos- phorus sink is the sediments. Both components greatly exceed the quantity of phosphorus stored in the water column at any time. 2. The input of stormwater runoff has severely affected the natural decomposition cycle in the Lake Eola ecosystem, resulting in an accumulation of partially decom- posed organic matter with a con- stantly high oxygen demand. 3. Flocculent sediment particles of smaller size and density contain considerably more phosphorus than larger and more dense parti- cles. In Lake Eola, flocculent, deep water particles contained about 61 percent more phosphorus than discrete, dense sand particles from the shallow areas. 4. A thick mat of decomposing algal cells covers the sand placed over the sediments in the Lake Eola restoration project of 1973 be- cause the source of phosphorus was not controlled. 5. Aerobic phosphorus release be- cause of turbulent agitation of the sediments is most likely an insig- nificant phosphorus source, but may serve to maintain lake water phosphorus levels when productiv- ity has depleted the phosphorus in the water column. 6. The hypolimnion of Lake Eola remains anaerobic from mid-May to mid-August, corresponding to the wet season and the associated increase of stormwater runoff during this period. This anaerobic condition enhances phosphorus release from the bottom sedi- ments. 7. Anaerobically released phos- phorus, when mixed out of the hypolimnion, has the potential for increasing water column phos- phorus by 11.6 fjg P04~3-P/L, which amounts to 50 percent of the average orthophosphorus con- centration in the lake (23 /jg/L). Selection of Management Practices Many stormwater practices are avail- able and tested in many geographical regions. Unfortunately, few of these practices have been documented rela- tive to their removal efficiencies for selected water quality measures. Diver- sion of the first-flush of stormwater has been documented. If the first-flush is percolated through soil, direct surface discharge of pollutants is eliminated. If soil conditions and water table permit percolation, then diversion for storage and treatment by percolation is possible. There appears, however, to be a need for more documentation on other prac- tices. The following conclusions expand upon our existing knowledge. 1. Percolation is possible after storm- water diversion in urban areas. Perforated pipe under sidewalks and parking lots within municipal right-of-way should be examined. 2. Diversion of the first 0.64 cm (0.25 in.) of every stormwater runoff event will produce long-term re- moval efficiencies of at least 80 percent. 3. Sedimentation will remove sus- pended materials. The reduction reported herein ranged from a high of 52 percent for TSS to a low of 20 percent for lead. 4. Alum coagulation plus detention improved stormwater quality more than detention only. Of the water quality parameters investigated, only the nitrogen forms and Ca were not reduced by alum coagula- tion. The reduction of TP and TSS after alum coagulation and deten- tion was in excess of 90 percent. 5. Standard isoconcentration lines and regression equations relating the percent removal of any of the water quality variables were devel- oped as linear or semi-log func- tions of settling velocity or deten- tion time. The regression equations were developed using an alpha error of less than 5 percent. The regression equations are easy to use. 6. Phosphorus release experiments using alum stormwater sludge, lakewater, and lake sediment indi- cated that the phosphorus re- moved by alum coagulation of stormwater would be retained in the solid form in a simulated lake environment for aerobic or anoxic conditions with or without gentle mixing. After defining the stormwater man- agement practices for the watershed, a computer program "MANAGE" was used as a methodology for the choice of the least cost combination. As a general methodology, however, more work was needed to estimate localized cost- efficiencies for stormwater manage- ment practices. Incorporating mathe- matical programming methods to this work allows one to select a combination of practices that will remove a maximum amount of pollutants at least cost. Implementation Steps Based on citizen concern and histori- cal water quality data on fish and duck kills, oxygen depletion, and algal blooms, it was evident that the factor causing the water quality impact had to be identified. Trophic state analysis indi- cated that the lake was estimated as eutrophic. In laboratory tests, algal productivity was related to stormwater. Also, the bottom sediments were shown to contribute to the phosphorus concen- tration in the water column. Based on the runoff quality and quantity data along with lake limnologi- cal data, an implementation plan for stormwater management was devel- oped. Since phosphorus is most likely the limiting nutrient, it will be controlled. The two major sources of phosphorus are stormwater and lake bottom mud recycling. By reducing stormwater phos- phorus mass inputs, littoral zone plant- ing, and coagulant coverage of bottom muds, it is predicted that the effects of stratified conditions (anaerobic) will be minimized and algal blooms will be reduced. The stormwater management will be done by diversion/percolation of parking lot runoff and limited street runoff approximately 24.0 hectares (60 acres) In addition, most of the areas no managed with this method will be diverted for filtration before discharge t( the lake, approximately 27.0 hectares (66 acres). The full report was submitted ir fulfillment of Grant No. R-805580 b' the University of Central Florida unde sponsorship of the U.S. Environmenta Protection Agency. ------- Martin P. Wanielista, Yousef A. Yousef, and James S. Taylor are with the University of Central Florida, Orlando, FL 32816. Richard P. Traver is the EPA Project Officer (see below). The complete report, entitled "Stormwater Management to Improve Lake Water Quality," (Order No. PB 82-227 711; Cost: $21.00, 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: Oil and Hazardous Materials Spills Branch Municipal Environmental Research Laboratory-Cincinnati U.S. Environmental Protection Agency Edison, NJ 08837 •&U. S. GOVERNMENT PRINTING OFFICE: 1982/559-092/0459 ------- United States Center for Environmental Research pees pg|d Environmental Protection Information Environmental Agency Cincinnati OH 45268 Protection Agency EPA 335 Official Business Penalty for Private Use $300 RETURN POSTAGE GUARANTEED _.. . _, Third-Class Bulk Rate MERL0063240 LOU H TILLEY REGION V £PA LIBRARIAN 230 S DEARBORN ST CHICAGO IL 60604 ------- |