PHYSICAL - CHEMICAL TREATMENT PLANT DESIGN Presented by Richard L. Woodward Vice President Camp Dresser & McKee Inc. Boston, Massachusetts Prepared for Environmental Protection Agency Technology Transfer Seminar Atlanta, Georgia May 1973 ------- Physical - Chemical Treatment Plant Design by Richard L. Woodward Camp Dresser & McKee Inc. INTRODUCTION Mr. Cohen has discussed the various unit processes used in physical - chemical treatment of wastewaters. It seems most useful that I use my allotted time to tell of two projects in which we have been involved where a physical - chemical process was selected. One project at Fitchburg, Mass., is now under construction and the other at Niagara Falls, N.Y. should soon be started, after some frustrating delays. Some preliminary results from a third project which is still in the study stage will also be discussed. NIAGARA FALLS. N. Y. Geographic Situation This city is a famous tourist center but it is also the center of a large chemical complex with such well known names as Hooker, Union Carbide Dupont, Goodyear, National Lead, 01 in, Carborundum and others. The Niagara River, which flows by the city, is an international boundary water connecting 3 Lake Erie and Lake Ontario and has a flow of about 200,000 cfs (5,700 m /sec). The city has a population of about 86,000. We were retained by the city in January, 1970 to study its water pollution control problems, to submit a preliminary design report by February, 1971 and to have the project under construction by March 1, 1972. Financing problems have delayed construction. Nature of Wastes. A sampling and gaging program of all major industrial wastewater discharges in early 1970 showed a total flow of 164 mgd ------- -2- (7.2 m /sec) a suspended solids load of 274,000 pounds per day (124,000 kg/day) and a COD load of 242,000 pounds per day (110,000 kg/day). Only 63 mgd 2 (2.8 m /sec) went directly to the municipal sewers but it contained some 70 percent of the suspended solids, about 55 percent of the COD load. This, of course, completely overshadows the domestic waste load. The waste reaching the existing treatment plant in 1970 frequently had a pH of close to 11, but acid dumps would occasionally drop this to below 3. Lime waste was evident and caused much of the discoloration of the river which was noticeable below the treatment plant. Volatile vapors from chlorinated hydrocarbons were occasionally so irritating to the eyes and lungs of workers at the treatment plant that they had to leave. Occasionally chlorine residuals could be measured at the plant. BOD measurements on this wastewater were meaningless. During pilot plant work it would sometimes be found that the BOD increased in going through physical - chemical treatment. A few coliform tests on the raw wastewater showed concentrations of only 100 to 400 coliforms per 100 ml. Clearly this wastewater was not a suitable candidate for biological treatment. Effluent Quality Requirements. All wastewater discharges to the Great Lakes and Niagara River in New York are required to remove phosphorus. A limit of 1 mg/1 of phosphorus is the general requirement for wastewater dis- charged to these waters. The usual requirement for secondary treatment was not defined in terms of BOD, but a limit of 112 mg/1 of COD has been prescribed by the New York Department of Environmental Conservation along with a limit of 35 mg/1 of suspended solids and 0.23 mg/1 of phenol. Pilot Plant Operations. A small pilot plant was constructed at the existing wastewater treatment plant and operated for approximately three months in mid-1970. Figure 1 shows schematically the facilities provided. Because of ------- -3- the high pH of the influent waste the chemical used for phosphate removal was lime,operating at a pH of 11.2 - 11.6. An average dose of 90 mg/1 of lime was required to maintain this pH. The average removal of COD in clarification was only about 40 percent, but the suspended solids removal was close to 80 percent. Phosphorus removal averaged only about 65 percent in the clarification stage. The carbon columns increased the overall removal of COD to about 75 percent, of suspended solids to over 90 percent, of phosphorus to 70 percent. Phenol removal was about 95 percent. The carbon columns were operated at a 30 minute empty bed contact time. Acid was used to reduce the pH to around 7 prior to application to the carbon columns to avoid precipitation of calcium carbonate on the carbon. The clarifier sludge in filter leaf tests dewatered readily to a solids content of more than 30 percent. The cake produced was not putrescible. The indicated carbon exhaustion rate was about 250 pounds per million gallons. The pilot plant work was discontinued after about 3 months because it was known that the waste characteristics would change markedly by the time the plant was built because of reduced contributions of waste from industry. The only specific organic compound analyzed for was phenol and, as indicated above, it was effectively removed. The nature of organics that failed to be adsorbed are not known but substantial amounts of organic chlorine were found in the effluent and samples analyzed by mass spectrometry showed the presence of mono- and dichlorobenzene in both influent and effluent samples. Negotiations with Industry. From the start of the project close contact was maintained with the local industries through a committee on which all important liquid waste contributors were represented. The entire pollution ------- -4- abatement program was to be financed by user charges and each major industry knew that it would be obliged to sign a long-term contract to pay its share of the cost. Rough preliminary costs figures were given to the industries as soon as possible to enable them to determine what in-plant changes could be made to reduce their treatment charges. A three-part rate based on flow, suspended solids and COD was used with the 18 larger industries, each of which was contributing more than 100,000 gallons per day of liquid waste, 250 pounds per day of suspended solids or 250 pounds per day of COD. As the work progressed it was possible to refine the cost estimates as better estimates of flow and waste load were received from the industries. In the end the industrial waste flow provided for in the final design was 28 mgd 3 (1.2 m /sec), the contribution of suspended solids was reduced to less than one-third and of COD to less than half of what was going to the municipal sewers in 1970. Plant Design. With such uncertainties as to the changes in the amounts and strength of the waste it was necessary to provide for considerable flexibility in designing the treatment plant. Figure 2 shows a schematic diagram of the plant as it will be built. Figure 3 shows an architect's rendering of the proposed plant. Final design on the initial construction program began in February, 1971 and was concluded in April, 1973. The treatment process will be similar to that used in the pilot plant, and a schematic process diagram is shown as Figure 2. Some of the basic design criteria include 48 mgd average flow with 86 mgd peak flow (estimated 2020 dry weather flow), 100,000 pounds per day suspended solids and 145,000 pounds per day COD. A composite population equivalent based on these criteria is about 500,000. ------- -5- The wastewater from the east side of the city will enter the main pump wet well in the new treatment plant from the Southside Interceptor. The main pumping station will include four, 15,000 gpm mixed-flow pumps, any three of which will have the capacity to pump the design flow from the east side, with one standby. The discharge from the two sources (the main pumping station and the new pumping station on the existing treatment plant site), will flow through mechanically cleaned bar screens and into two rapid mix basins with a combined detention time of about 1 min at 86 mgd. The materials which may be fed to the wastewater in the rapid mix basins include lime, a metal salt (ferric chloride, alum, or most likely chlorinated ferrous sulfate), a coagulant aid, return sludge from the primary sedimentation basins, acid (if the treatment process includes depressing the pH rather than elevating it), and the spent backwash water from the activated carbon beds. • The change from the lime system to a metal salt system could be made in a matter of minutes, and the choice would be dependent on the pH of the incoming waste, the cost and availability of chemicals, the ease of sludge dewatering, and other factors. There is a strong possibility that the lime discharges to the sewers will be eliminated to reduce user charges and that either aluminum or iron may be more economical chemicals. From the rapid mix basin the wastewater will be split hydraulically into five flocculation basins, each followed by a sedimentation basin. The flocculation basins will be 60 ft square and each will have nine flocculators. The flocculators will give decreasing "G" values in the direction of flow, with 100 sec" possible at the influent end and as little as 10 sec" at the effluent end of the flocculation basin. The detention time in the basins will be about 20 min at the peak flow. ------- -6- Sedimentation basins will be 60 ft by 250 ft long with an average depth of 12 ft. The overflow rate at the peak flow will be 1,400 gpd/sf and about 800 gpd/sf at the average flow. The detention time at peak flow will be 1.4 hr. Sludge will be drawn off from the sedimentation basins to one of two 70 ft diameter gravity thickeners, although some of the sludge may be returned to the rapid mix tank to aid in floe forming. The flow from the primary basins will go through an acid mix tank to adjust the pH to between 6 and 7 (if the high lime process is used), and from there the liquid will be pumped to the activated carbon beds. The pumping facility will include four, 27,000 gpm mixed-flow pumps, any three of which will pump the peak flow plus the spent backwash water, with one standby. The granular activated carbon beds will be downflow gravity beds. There will be 28 beds, each 17 ft by 42 ft with a 9 ft depth of activated carbon. For a plant this size gravity concrete beds are considerably more economical than steel pressure vessels. The underdrain system will consist of precast concrete slabs with plastic nozzles placed 8-in on center. The rate controllers on the effluent piping of each of the beds will split the flow equally among all of the beds on line. The superficial velocity through the beds will be 3.3 gpm/sf at the peak flow. The treated effluent will flow to a chlorine contact chamber with a detention time of 15 min at 86 mgd. Each of the carbon beds will be backwashed at least once each day with an air wash followed by a water wash. The maximum air wash rate will be 3,000 scfm, and the water wash will be up to 20,000 gpm from pumps taking suction from the plant effluent line. This 20,000 gpm corresponds to a rise rate of 44 in/min in the bed. This high rate would be needed only when removing carbon for regeneration. The spent backwash water could flow either to the ------- -7- rapld mix basin, where it would be mixed with the incoming flow and treated, or it could be directed to its own rapid mix basin, where it would be mixed with the incoming flow and treated, or it could be directed to its own rapid mix basin and into an end flocculation basin. It is likely that even plain sedimentation may improve the spent backwash water quality to the point where it could be discharged to the river, with chlorination, without having to go through the activated carbon beds again. Carbon in beds which have reached their adsorptive capacities will be taken out of the beds hydraulically for regeneration. The regeneration facility will include a six-hearth furnace, in which carbon will be applied to the top hearth and slowly pushed from one hearth to the next lower one. About three days will be required to regenerate one bed. The regeneration will be done in the absence of oxygen, and steam will be applied as a source oxygen for the adsorbed organic carbon to be driven off as CO.. Above the top hearth, the gases from the regeneration will be burned at about 1800°F to break down any chlorinated hydrocarbons that exist in the gas into CO and HC1. The gases will go through a wet scrubber which will dissolve the HC1 and remove the particulate matter from the stack gas. The scrubber effluent will be recycled to the front end of the plant. Sludge Handling and Disposal. About 80 tons of dry sludge solids per day is the expected sludge production, if lime is used. Smaller amounts of sludge are expected if iron or aluminum is used. The sludge will be pumped from the gravity thickeners and will be dewatered on four, 500 sf vacuum filters, which will operate about 120 hours per week. Ferric chloride, a ^ polymer and lime will be available for conditioning the sludge. The dewatered sludge will be hauled to landfill for burial. Incineration of the sludge was considered, during both the preliminary and the final design ------- -8- phases, but it would be more costly than burial. The plant influent suspended solids contain a low percentage of volatile material, and this, together with the inorganic chemical additions, would make incineration little more than a drying process with most of the heat having to be applied externally. The reduction in volume and dry solids weight of the evaporated sludge would not be significant over the vacuum filtered cake. Similarly, the large amounts of inert suspended solid in the sludge made lime recovery appear unprofitable. Changes in the waste characteristics as industries reduce their waste loads may call for a review of this decision. Costs. The estimated cost of the treatment plant is $37 M and various necessary sewers, a pumping station and force main will add another $16 M. Annual operating costs are estimated at $1.8 M. The capital cost for the treatment plant is amortized in 20 years at 6 percent interest the total 3 cost of treatment alone based on a flow of 48 mgd (2.1 m /sec) is $290. FITCHBUR6. MASSACHUSETTS Geographic Situation. Fitchburg, Massachusetts is an industrial city located in central Massachusetts and has a population of 43,000. It is located at the headwaters of the Nashua River. Near the upstream end of the city two paper mills, the Weyerhaeuser Co. and the Fitchburg Paper Co. (a division of Litton Industries) use all of the flow of the stream during low flow periods. In fact, Fitchburg Paper Co. depends on Weyerhauser's waste flow for its raw water supply. Thus it was not feasible to intercept the waste from these mills and treat it at a downstream plant which will serve most of the city. Figure 4 shows the Nashua River as it flows through Fitchburg, the location of the mills and of the two proposed wastewater treatment plants. The west plant will serve primarily the two paper mills along with some domestic waste from Fitchburg and an adjoining community. The east plant will serve the remainder of the area. ------- -9- Nature of Wastes. The wastewater from the two mills is primarily white water containing paper fiber along with chemicals used in paper making such as alum, starch, pigments and fillers. The total waste flow from the two industries averages about 13.8 mgd (0.6 m /sec). The waste has an average BOD of about 50 mg/1 and suspended solids of 360 mg/1. Effluent Quality Requirements. Because the effluent from the west treatment plant will constitute the entire stream flow during low flow periods a BOD of 5 - 7 mg/1 was needed. To avoid aesthetically objectionable color from the pigments used in making colored paper good removal of color is necessary. Suspended solids should also be kept to low concentrations. Pilot Plant Operations. When the pollution abatement problems of Fitchburg were studied in 1968, biological treatment was proposed for both the east and west Fitchburg plants and in 1970 pilot plant studies were made for both types of waste. The paper mill wastewater contains substantial amounts of alum and by adding an alkali to increase the pH to about 6.7, a good floe is formed which settles readily and removes some 50 percent of the BOD and more than 90 percent of the suspended solids. Although the activated sludge pilot plant did a satisfactory job of removing BOD, color was not consistently removed and there was some carry-over of the biological floe in the effluent. It was necessary to add nitrogen and phosphorus to the waste as it was deficient in both nutrients. In addition, some problems were anticipated in operation because both mills operate on a five-day week and shut down for a two-week period each summer. This would complicate the operation of a biological treatment plant. A small pilot plant facility was established at the Weyerhaeuser mill where waste was available from one of its primary clarifiers. Wastes from the ------- -10- Fitchburg Paper Co. mill were trucked to the site and mixed in suitable proportion to Weyerhauser waste. The combined wastes were settled in a tank truck and then pumped to four carbon columns in series. About 90 per- cent of the settled BOD and COD were removed by the carbon columns operated at a rate which would provide an empty bed contact time of 23 minutes. Very good color removal was obtained. The exhaustion rate was such that the carbon usage would be about 1,600 pounds of carbon per million gallons. This was much higher than we found at Niagara Falls and about double the rate commonly reported in treating domestic wastewater. Negotiations with Industries. The West Fitchburg plant is to be paid for primarily by the industries which will provide most of the flow. A thorough economic analysis of the relative costs of the physical-chemical system is compared with activated sludge. This showed that the PC system would be about 6 percent more expensive, but nevertheless the industries as well as the city and the state regulatory agency chose the more expensive system because of its better effluent quality and operating stability in the face of major changes in flows. Plant Design. Figure 5 is a schematic of the West Fitchburg plant. Municipal wastewater will be settled and chlorinated heavily before being combined with the industrial wastewaters. The primary sludge from the municipal clarifiers will be processed at the East Fitchburg treatment plant. The combined wastes will be coagulated, flocculated and settled. The chemicals normally used will be lime and a polymer. The industrial wastewater normally has a pH in the range 4.2 - 4.7 and upon adjustment of pH to 6.7 the alum in the waste forms a good floe. At times of high turbidity when large amounts of titanium are being used in paper making it will be necessary to add alum. ------- -11- Two circular municipal clarlflers will be provided Initially. These are designed for an overflow rate of 1,000 gallons per square foot per day. Two rapid mix and two two-stage flocculation basins will provide 5-minute and 30-minute detention periods, respectively, at maximum day flow rates. Two industrial clarlflers, each 130 ft in diameter will provide an overflow rate of 840 gallons per square foot per day at maximum days flow rate. Sludge will be pumped off site to four sludge lagoons. Vertical turbine pumps will deliver the settled wastewater to the activated carbon columns. Four variable speed pumps will be provided each with a maximum capacity of 8,500 gpm at a design head of 145 ft. These have been designed to provide for 10 ft of head loss through prefiltration facilities if these should be needed. Twelve activated carbon columns, each 20 ft in diameter with a 15.5 ft bed depth, are provided. They will be operated in parallel. The columns and piping allow for two vessels to be out of operation, one for backwashing and the other for regeneration. Ten active vessels were selected to permit maximum utilization of the carbon. The surface loading and empty bed contact 2 time for the 10 active vessels will be about 8 (gal/min/ft ) and 15 min, respectively, at peak hourly flow, and the flow through the vessels will be divided uniformly and controlled by individual rate controllers. The vessels have been designed to operate at a maximum pressure of 50 psi. The overall vessel height from dished end to dished end is approximately 33 ft, which will allow for a 40 percent expansion of the carbon bed on backwashing. The filter bottom will consist of 1 ft of gravel over Leopold tiles. The columns will be backwashed with plant effluent utilizing two variable speed, vertical turbine pumps, one of which will serve as a standby. A maximum wash water rate of 9,100 gal/min will insure 40 percent expansion ------- -12- of the filter at 75°F and the water will be discharged to a backwash lagoon. It is anticipated that each column will be washed for 15 min daily, with one column backwashed every 2 hours on a time cycle, and will also use a surface wash. A clear well with a volume equal to approximately 1.0 times a single backwash is provided, but the effective volume is greater since there is a continuous flow into the clear well at all times except at shutdown. A backwash basin (two 65 ft diam. tanks) will be provided to uniformly return the backwash water together with the carbon transfer motive water to the head end of the plant. Under the most adverse conditions, the maximum rate of return to the head end of the plant will be 2.2 million gallons per day or about 12 percent of the plant throughout. Because the dissolved oxygen will be depleted in passing through the carbon columns, the effluent will be aerated to maintain a level of 5 mg/liter of dissolved oxygen. Approximately once a week it will be necessary to replace the carbon in one of the vessels. Spent carbon will be transferred to a spent carbon storage vessel through water eductors which are sized to empty or refill one vessel in less than 6 hours. The volume of the spent carbon and regenerated carbon storage vessels will be equal to 2 times the carbon in any one filter vessel. The carbon will be regenerated in a 10 ft - 9-in diameter six hearth furnace utilizing steam. It has been sized based on continuous operation with a 2 loading rate of 100 pounds of carbon/day/ft of hearth. An impingement scrubber will be used to remove particulates from the gas stream. In transferring the carbon, two water streams are required: fluidizing water for maintaining the carbon-to-water ratio (1-2 Ib of carbon/gal of slurry) and motive water to operate the eductors. ------- -13- Fluidizing water will be pumped against essentially the static head and motive water at the most efficient combination of pressure and volume. Fluidizing water and motive water to and from the carbon vessels will be provided by separate constant speed pumps which will operate at 100 psi, with suction from the process water header. Figure 6 shows the plant layout and Figure 7 is an architect's rendering of the proposed plant. Estimated Costs. Bids were taken for construction of this plant and a contract was let in January, 1973 for $10,700,000. It should go into operation in early 1975. The estimated first year operating cost is $515,000, based on 1971 prices. Amortizing the plant construction cost of 6 percent over 20 years, this gives a treatment cost, including operation and mainte- nance, of $280 per million gallons. WESTERN CITY Geographic Situation. We are currently involved in a study of water pollution abatement facilities for a western city of about 200,000 population which is located on a well regulated stream with a minimum flow of about 2,000 cfs. Nature of Wastes. The wastewater is rather representative of normal municipal wastes although there is a substantial contribution of industrial wastes from food processing plants and local service industries. There are no problems with toxic industrial wastes. Effluent Quality Requirements. Some 30 miles downstream from the existing primary treatment plant a 150 ft high dam has been built forming a reservoir some twenty miles long. During the summer this becomes thermally stratified and the bottom waters become depleted of dissolved oxygen. There have been increasing problems with algae growths in the reservoir. ------- -14- The city has been ordered to provide facilities for phosphate removal from its wastewater and to provide secondary treatment. Pilot Plant Operations. Two principal schemes have been studied for providing this treatment, namely chemical precipitation of phosphates in the primary facilities followed either by carbon adsorption or by biological treatment. Both have been studied on a pilot plant scale using lime, alum and ferric chloride as the precipitating chemicals. Approximately 300 mg/1 of lime, 175 mg/1 of alum and 90 mg/1 of ferric chloride appear to be necessary to achieve about 90 percent removal of phosphates as is required. Because chemical costs are high with lime at about $40, alum at $70 and ferric chloride at $160 per ton, lime, together with recalcination, seems to be the most economical chemical. The activated carbon columns have not shown notably promising performance. Early results indicate about 65 percent removal of both COD and BOD by the carbon columns with effluent BOD's and COD's of about 40 and 63, respectively, with an empty bed contact time of 30 minutes. The indicated exhaustion rate is about 1,100 pounds of carbon per million gallons and a COD loading of about 0.45 pounds of COD per pound of carbon. Although these are only preliminary results it would appear that the biological treatment will be the more economical in this situation. SUMMARY At Niagara Falls a physical - chemical treatment process was selected largely because the wastes to be treated would have been toxic to biological processes and many of the objectionable constituents would not have been removed by biological processes. At Fitchburg the process was chosen because of the better quality effluent produced and the greater stability in operation as compared with biological treatment. In each case the choice was a clear one. In a third city with no unusual industrial waste problems preliminary ------- -15- results of pilot plant work indicate that a chemical - biological treatment process will probably be preferable to physical - chemical treatment. At the moment there has been no full-scale experience with independent physical - chemical treatment. I think we all await such results to enable us to improve designs and cost estimates for such plants. The delays which have held up construction and operation of the several plants of this type have been frustrating. We hope that they are about at an end. ------- RAW SEWAGE CHEMICAL PUMP ADDITION COAGULATION FLOCCULATION TANKS CLARIFIER pH ADJUST- COAL/SAND ME NT FILTER LIME POLYMER ACTIVATED CARBON COLUMNS PH DISSOLVED ADJUST- AIR FLOTATION MENT ACTIVATED CARBON COLUMNS ACID- DISCHARGE DISCHARGE FIG. 1 PILOT PLANT FLOW DIAGRAM NIAGARA FALLS, NEW YORK ------- BAR RACK RAPID FLOCCULATION MIXER BASIN CLARIFIERS CARBON CONTACT BEDS INFLUENT SLUDGE TO. DISPOSAL VACUUM SLUDGE FILTERS THICKENERS CARBON REGENERATION EFFLUENT TO RIVER CHLORINE CONTACT CHAMBERS FIG. 2 SCHEMATIC OF WASTEWATER TREATMENT FACILITY NIAGARA FALLS, NEW YORK ------- FIG. 3 ARCHITECTS RENDERING NIAGARA FALLS, NEW YORK ------- WESTMINSTER FIG. 4 FITCHBURG WASTEWATER ABATEMENT FACILITIES ------- I. TOWN OF WESTMINSTER (FUTURE) 2. MONTACHUSETT REGIONAL VOCA- TIONAL TECHNICAL HIGH SCHOOL 3. CITY OF FITCHBURG PRIMARY SEDI- MENTATION COMMINU- TION FLOW MEASURE- MENT CHLORINA- TION FLOW MEASURE- MENT r I. WEYERHAUSER CO. 2. FITCHBURG PAPER CO. CHEMICAL ADDITION EAST FITCHBURG STP PRIMARY SLUDGE SLUDGE LAGOONS (OFF SITE) COAGULATION FLOCCULATION CARBON REGENER- ATION T I I I I I I t CARBON | FILTER I \ FILTER BACKWASH LJ ID U_ U. UJ FIG. 5 SCHEMATIC OF WASTEWATER TREATMENT PLANT WEST FITCHBURG, MASSACHUSETTS ------- SLUDGE AND CLARIFIERS / CHEMICAL BUILDING TREATMENT PLANT BUILDING BOSTON AND MAINE RR •-J I- SCALE IN FEET FIG. 6 WEST FITCHBURG WASTEWATER TREATMENT FACILITY SITE PLAN ------- FIG. 7 ARCHITECTS RENDERING WEST WASTEWATER TREATMENT FACILITY FITCHBURG, MASSACHUSETTS CAMP DRESSER & MCKEE ------- |